W I N A N I K M U LT I M E D I A A M P L I T U B E 2 G U I TA R A M P AND EFFECTS MODELING POWERHOUSE! IT'S MUNGO. www.VirtualInstrumentsMag.com J U N E / J U LY 2 0 0 6 - V O L . 2 N O . 3
THE WORLD OF SOFTSYNTHS AND SAMPLERS
C A N T H I S S A M P L E P L AY E R READ YOUR MIND?
www.magesy.ru VSL’s new Vienna Instruments player harnesses their vast orchestral library under real-time control
A BREAKTHROUGH TWEAK: RUN 75% MORE ON THE SAME MACHINE. USA $4.95 CANADA $6.50
REVIEWED: Audeon UFO softsynth • Cycling ’74 Cycles vol. 4 • IK Multimedia Classic Studio Reverb • PSPaudioware Neon HR precision mastering EQ • Linplug Octopus synth • Ueberschall Urban Jointz sound library
F R E E A L G O R I T H M I C R H Y T H M G E N E R AT O R . D O W N L O A D I T F R O M W W W. V I R T U A L I N S T R U M E N T S M A G . C O M .
www.magesy.ru
www.magesy.ru
www.magesy.ru
www.magesy.ru
www.magesy.ru
From the
Editor T
his magazine would be nothing without the great writers who’ve been contributing to it, and I’m happy to welcome a few new ones this issue. First but not least, it’s a real thrill having a second writer who was on Keyboard magazine’s staff back in the ’80s heyday (the first being Jim Aikin; that incarnation of Keyboard was what inspired me to get into this field). Michael Marans is one of the most knowledgeable people around when it comes to synthesizer programming, and his first article “Control Voltage” is great for beginners and more advanced programmers alike. We look forward to many more. Next, everyone on the internet forums knows composer Thomas Bergersen’s work. There are some very talented people around, but absolutely nobody I’ve heard can do what Thomas J, as he’s usually called, does using orchestral sample libraries. Check out “From Sketch to Score,” and be sure to download and listen to the audio example from our website. Thomas lives in Trondheim, Norway, but he works in the U.S. and all over. Also in the international spirit, thousands of TASCAM GigaStudio users around the world have benefitted from Mattias Henningson’s Windows registry tweaks. Mattias is from Sweden; Scandinavia is a hotbed for hardcore sample library users. What Mattias did is figure out how to access a good 25% more installed memory in GigaStudio 2, and TASCAM incorporated his
tweaks into GigaStudio 3 (which incidentally can often load a further 20%, but without installing more RAM). This time Mattias explains how to use the Windows 3GB switch, allowing programs other than GigaStudio to access a good 75% more RAM. There used to be no reason to install more than 2GB in a Windows machine, but now it makes sense to put in 3 or 4GB. (Why is RAM access so important? Because it allows you to have more programs cued up and ready to play in your sampler, even if you don’t end up using all of them.) The other new writers I want to introduce are Jason Scott Alexander, who among other things does a great job with plugins…something we’re now getting around to covering more, as promised in the inaugural issue; and Orren Merton. Part 1 of Orren’s Logic Pro Ultrabeat tutorial was in our last issue, so hopefully it’s not too serious a violation for this acknowledgement to accompany Part 2. Orren wrote an excellent book on Apple Logic Pro (“Logic 6 Power”) for Thomson Course Technology. Next issue we’ll announce the winners of the Mungo Giveaways we’ve been having. You can enter the current one by going to our website, www.VirtualInstrumentsMag.com. While you’re there, perhaps you’d like to subscribe. This might also be a great time to renew your subscription; the rates have been artificially low while we launched the magazine, but we’re going to raise them in the near future. Enjoy the issue. VI
Editor/publisher: Nick Batzdorf
Virtual Instruments is published bi-monthly for $12.95/year, $20/two years (with occasional add-on one-year promotions for $9.95) by Virtual Instruments, Inc., 3849 Ventura Canyon, Sherman Oaks, CA 91423-4710. 818/905-5434, 1-877/ViMagzn.
[email protected]. Application to Mail at Periodicals Postage Rates is Pending at Van Nuys, CA, and at additional mailing offices. POSTMASTER: please send address changes to VIRTUAL INSTRUMENTS, 3849 VENTURA CANYON, SHERMAN OAKS, CA 914234710.
Art director: Lachlan Westfall/Quiet Earth Design Ad/marketing manager: Carl Marinoff Web designer: Denise Young/DMY Studios Contributors: Jim Aikin, Peter Buick, David Das, Doyle Donehoo, Jerry Gerber, David Govett, Ashif “King Idiot” Hakik, Michael Marans, Monte McGuire, Orren Merton, Chris Meyer, Dave Moulton, Zack Price, Frederick Russ, Bruce Richardson, Craig Sharmat, Lee Sherman, Dietz Tinhof, Jesse White. Advertising contact: Carl Marinoff 818/590-0018.
[email protected] Subscriptions/Address changes: 818/905-5434, 1-877/ViMagzn,
[email protected]. The best method is to subscribe via our website: www.VirtualInstrumentsMag.com. Letters to the editor:
[email protected] Writing for Virtual Instruments Magazine: query
[email protected] or call 818/905-9101
Distributor: Rider Circulation Services, 3700 Eagle Rock Blvd., Los Angeles, CA 90065. 323/344-1200. Bipad: 05792, UPC: 0 744 70 05792 5 05 Standard disclaimer: Virtual Instruments Magazine and its staff can’t be held legally responsible for the magazine’s contents or guarantee the return of articles and graphics submitted. Reasonable care is taken to ensure accuracy. All trademarks belong to their owners. Everything in here is subject to international copyright protection, and you may not copy or imitate anything without permission. © 2006 Virtual Instruments, Inc.
www.magesy.ru VIRTUAL
INSTRUMENTS
5
www.magesy.ru
www.magesy.ru
June/July 2006
V2.N3
30 Jerry Gerber interview by Nick Batzdorf
10
Letters
12
Launch
A very different kind of poster boy for the V.I. medium.
Introductions, updates, news
44 From Sketch to Score by Thomas J. Bergersen
14
Voltage Control by Michael Marans
A new series on programming synthesizers from the ground up
26 3GB Switch by Mattias Henningson
The superhero responsible for increasing GigaSampler’s memory access by a third explains how to tweak Windows XP so other programs can access about 75% more memory for loading samples.
8
VIRTUAL
INSTRUMENTS
Translating sequenced music to a score for live players presents some unique challenges. The highly skilled and talented young composer Thomas J—who’s also unparalleled at orchestral MIDI programming—goes over his process for doing that, offering pointers along the way
Pro Ultrabeat 54 Logic pt. 2 by Orren Merton
Part 2 of a 2-part tutorial on the built-in groove machine.
www.magesy.ru
contents VI
June/July 2006
V2.N3
random
40
tip
Random Tips Getting Steinberg V-Stack to wake up
58
Buick Sessions: Sound design in music production
64Trends – scripting 64
The craze over customizing Native Instruments Kontakt 2
by Peter Buick
Part 2 of our look at inspiration and production: a free algorithmic rhythm generator program.
22
Vienna Symphonic Library Symphonic Cube
42
by Nick Batzdorf
The Vienna Instruments player is not your father’s sampler, nor is VSL his sampled orchestra
36
IK Multimedia Classik Reverb by Lee Sherman
Real-time control meets the character and flexibility from the glory days of outboard
38
Linplug Octopus by Jim Aikin
A boatload of distinctive and highy programmable electronic tones
reviews VI 52
PSPaudioware Neon HR Linear Phase Precision Mastering Equalizer
Audeon UFO synth by Jim Aikin
Something old, something new…
Review by Jason Scott Alexander
Sometimes a guest that takes over your machine is welcome
50
Loop Librarian: Uberschall Urban Jointz; and cycles vol. 04: momentary incursions by Chris Meyer
A collection of useful “modern hihop and Rnb flavas,” and the latest volume in a unique series of sound design elements
VI
www.magesy.ru VIRTUAL
INSTRUMENTS
9
l e t t e r s VI
Letters write to:
[email protected]
Studio Farms article I was very interested to read Jesse White’s article “Studio Farms” in the April/May edition. I have been interested in creating a server-based system for some time now and appreciate the helpful info Jesse included in his article. I am running a Windows DAW with Steinberg Cubase SX3 as my main recording and sequencing environment. I have delved fairly heavily into plug-ins and V.I.s for some time now and realize how these can tax my system, so a studio farm set up would be perfect for me. I also run multiple accelerator cards in my setup (Universal Audio UAD-1 Studio Paks and TC PowerCore FireWire). My question is, in the studio farm setup, where would these units be placed? I assume that my DAW would be able to access the cards if placed in the server. Please clarify. Gary Janzen Vancouver, Canada If you’re using FX-Teleport (www.FXMax.com) to run audio over ethernet, you’d want them on your sequencing DAW rather than the server. The reasons for this aren’t philosophical—that’s just what works.
Hi Jessie: First of all, the “Studio Farms” article you wrote was an epiphany for me on the way to set up my studio. Between the benefits of reliability (if something goes down) and workflow (only using 1 comp), I 10
VIRTUAL
was so thrilled by the theory behind it all. I have currently been using five PCs and one Mac all slaved to a main Mac running MOTU Digital Performer. Each of the comps, however, has a dedicated set of V.I.s that it runs and if one goes down.... Well you know. This concept of using Fx-Teleport as a distribution system, having everything fed from a central server and working off one box is so cool! My question: immediately after reading your article, I went right out and bought a RAID card and some drives and began reconfiguring one of my PCs as a server. The problem I ran into, however, was that whenever I would request a V.I. be on a “farm” machine, it wouldn’t work because the “farm” machine wasn’t authorized for that plug-in. Do you have to have a license for every plug-in for every “farm” machine? If so, that’s just going to be too cost prohibitive for me. Is there something I’m missing? I installed the plug-ins on my main computer, but installed the *.dll file and the libraries onto my server. It works fine if I’m using the plug-in on my main comp, but as soon as I farm it out to one of the other comps, it won’t come up and says (on the farm comp screen) that it’s not authorized. If licenses end up being a problem, then we need to start bugging the manufactures to provide some type of “farm” licenses. The way this works is in no way a violation of any existing license because this system acts as one computer with one user. Patrick Rose via email Jesse White responds: I’m glad my article is helping people understand how to streamline their studios better. As far as authorizations, a lot of the time you can just use the same serial number on multiple machines. There isn’t really an answer for dongles. The reality that these companies don’t understand is that their software works great for networking applications—but they don’t offer an easy way to set it up. The authorizing can be a
INSTRUMENTS
little bit of a pain with these applications. I’m hoping that bringing this reality to the table with articles like this and composers wanting to migrate to this system that companies will realize their need for a solution. NB adds: This is likely to become an increasingly important issue as more and more musicians start using multiple computers. Copy protection schemes need to make it possible to run software on an entire subnet, i.e. a “studio farm” network. Perhaps dongle manufacturers could offer developers “sub-dongles” that require the presence of the main dongle on the same subnet, and challenge-response schemes could do something similar. They’d have to figure out how to ensure that multi-user facilities still purchase site licenses, but that shouldn’t be an insurmountable detail.
Back issues and subscriptions I am thoroughly enjoying Virtual Instruments since I subscribed. What a great magazine! I’m certain that you will have much success and I’ll do my part to ensure that by continuing to subscribe. Any decisions yet about offering hard copies of previous issues? David Noll Harlingen, TX Thanks very much David. Yes, back issues are now available in the US for $5 each plus $5 shipping/handling per order (regardless of the number of issues—a policy that will obviously have to change when we’ve been around a little longer and there are more back issues). The exception is that the premiere issue is $20 + $5; we have only a handful of them left. International shipping/handling depends on the country, but it’s going to be in the $9 range. Please visit www.VirtualInstrumentsMag.com. While we’re on the subject of back issues, there’s some confusion about the PDF downloads of all our issues. The PDFs are exactly the same as the print edition, page for page, ads and all. We have no immediate plans to remove our back issues from the download account area, so if you subscribe you can read all the issues you’ve missed on your computer (or you can print them out). Now, we offer download-only subscriptions to overseas readers for the same price as a US sub (because overseas delivery is expensive). What’s confusing people is that every subscriber to the print magazine can also sign up for a download account as part of the service. If you’ve subscribed and haven’t signed up for a download account, please go to our website and click on Download Subscription. We have to approve the accounts manually, so please be patient if it’s the middle of the night when we’re out of the office. VI
www.magesy.ru
www.magesy.ru
l a u n c h VI
Launch
Introductions, updates, news
Ableton Live Lite Enhanced: free upgrade for Digidesign and M-Audio users The new Enhanced edition includes many features from Ableton’s Live 5. Live is a unique digital audio sequencer that lets you string together layers of loops and pattern arrangements in real time, drag and drop effects and instruments, and much more. Some Digidesign and M-Audio products are bundled with a light version that has been upgraded. New features include plug-in delay compensation, MIDI remote control, full ReWire support, hi-def recording and rendering (up to 32-bit), Mackie Control support, Track Freeze, and their Complex warp mode for time stretching. http://www.ableton.com/free-m-audio-upgrade http://www.ableton.com/free-digidesign-upgrade
ILIO Ethno Techno S.A.G.E,. Xpander for Spectrasonics Stylus RMX Steinberg HALion Symphonic Orchestra Based on Steinberg’s HALion sampler, this new orchestral library ($499, crossgrades and upgrades available) is over 27GB and includes 1250 patches. A Crescendo controller feature lets you blend samples with different dynamics in real time, there are keyswitch instruments, you can adjust the recorded ambience, and the library is designed to be playable and efficient. The cross-platform instrument works in VST, AU, DXi, ReWire, and stand-alone formats. www.steinberg.net
12
VIRTUAL
INSTRUMENTS
Ethno Techno ($99) is the first in a series of four S.A.G.E. expanders for Spectrasonics’ popular Stylus RMX Realtime Groove Module, featuring the playing of Bashiri Johnson. Produced by Spectrasonics’ Eric Persing, the library features rare instruments and unique rhythms and textures, “from broad and spacious ambient explorations to radical heart-pounding beats.” These new sounds can be used on their own or as a complement to other RMX libraries. www.ilio.com
www.magesy.ru
l a u n c h VI
PACE Anti-piracy InterLok 5.4 for Intel Macs and PowerPC PACE copy-protection, whether software-based or using their iLok USB dongle, now works on Intel Macs. www.paceap.com
Submersible Music DrumCore 2 Submersible DrumCore 2 adds new audio engine features, and more A-list drummers and styles to its loop librarian/player (reviewed in our premiere issue). New features include continuously variable tempos down to three decimal places, tempo sync with ReWire hosts, separate outs with pitch and pan controls for its MIDI drumsounds, and REX export capability. It can also work as a librarian for REX2 and ACID format libraries intermingled with its own. The neew loops, fills, and drum kits, include country grooves from Lonnie Wilson; odd meters (5/4, 7/8, 9/8) from Alan White; doublekick pop/rock/progressive grooves from Terrio Bozzio; more jazz brush content from Jon Bishop; and pop grooves from Ben Smith. www.drumcore.com
Toontrack EZDrummer This $179 acoustic drum library was recorded at Avatar Studios in NY by Nir Z and produced by Neil Dorfsman, Mattias Eklund, and Henrik Kmellberg. It has the same layer depth as Toontrack’s dfh Superior, with 7000 16bit/44.1kHz sound files, but it’s designed to reduce system requirements to a minimum, while providing entry-level usability and professional flexibility. Features include multiple mic controls, an internal mixer for stereo and multitrack routing into the host from one plug-in, preset mix modes for quick sound changes, a visual interface that combines kit construction and auditioning, instant access to a drum library of patterns you can drag and drop, and instant loading into RAM to cut down subsequent loading time. Toontrack has announced three expansion packs: Latin Percussion EZX, Drumkit From Hell EAH, and Vintage Brushes EZX. www.soundsonline.com
BIAS Peak Pro 5.2 The latest upgrade ($179) to BIAS top-line audio editing, processing, and mastering program features 100% native compatibility on Intel-based and PowerPC Macs. It also offers a new authorization method that doesn’t require a USB dongle. Watch for a full review of Peak 5 in our next issue, but the major new features include an advanced playlist with Red Book premastering features (CD Text, ISRC, track indexing, custom gaps, cross-fades). Peak also hosts V.I.s, which makes the program especially interesting in our context. www.BIAS-inc.com
Izotope XRB Expansion for iDrum and Garageband XRB ($29 until 7/23, $49 thereafter) is a collection of 2057 drum grooves triggering 256 playable kits. iZotope also offers a free set of 16 iDrum kits and 74 Apple loops. www.izotope.com
www.magesy.ru VIRTUAL
INSTRUMENTS
13
f e a t u r e VI
more
online
www.virtualinstrumentsmag.com
Our ‘First DAW’ beginner’s series is on temporary hiatus, since this article fills its role. But it shall return.
Voltage Control A new series on programming synthesizers from the ground up
Not everyone has to (or should) go out and sample a grand piano every time they put a new track together. But what about your synth sounds? Do you really want to use “Toto Horns III” and “Stevies’s Moog Bass IX” on your track? Great sounds to be sure, but 30 years later, are they original, compelling, and—most important of all—yours? by Michael Marans
14
VIRTUAL
INSTRUMENTS
www.magesy.ru
f e a t u r e VI We strive for innovative creativity in our compositions, so why do we settle for using factory presets to realize them? Lots of reasons, to be sure: synths these days can be enormously complicated; their interfaces— most often accessed via a mouse—sometimes don’t exactly invite creative experimentation; and lets face it, the majority of factory presets are ridiculously good, often created by the top sound designers in our industry. Then again, maybe we’ve all become just a little bit lazy. (Sorry.) So our exercise today is to poke, prod, and cajole you into digging into your virtuFig. 1: Pop quiz: Can you spot the assigned parameter in the oscillator section that has absolutely no effect on al synths and creating the sound? (Answer below). your own unique While some virtual instrument interfaces may appear cryptic on the surface, Synth1’s front panel actually reveals sounds. Not just any a wealth of information about what is going on in the patch. In the patch Solo 1 shown here, key elements of the sounds, mind you, but sound are immediately obvious: the oscillator waveform shapes (both sawtooth), LFO1 routed to create vibrato killer sounds on par (notice it’s assigned to both oscillators and engaged via the LFO1 Wheel Sens control), and the use of a 12dB per with the best of ’em. octave (2-pole) lowpass filter. Clicking on a control reveals a pop-up with the actual parameter value—useful for We’ll start by explorexamining in detail how the patch was created. (Here the Filter Decay has a value of 36.) ing classic subtractive Pop quiz answer: The M. ENV parameter LED is lit, indicating that the dedicated oscillator modulation envelope analog synth programis in use. However, as the red LED next to “p/w” reveals, the modulation destination is oscillator pulse width. Since ming. Before your eyes both oscillators are assigned to sawtooth waveforms, the output from this modulation source is simply going off glaze over and you into the ether. break into a big yawn, remember that some of the greatest synth players ever made their marks with signature sounds created using Knowing which knob to turn critical for finding the most direct path to the some very simple analog tools—and in fact There’s an old joke where a repairman fixes desired result. these players made their unique sounds using a broken fridge by kicking it, then presents In the old days, analog synthesizers consistthe same tools. the owner of the fridge with a bill for $100. ed of electronic circuits connected together News flash: you can make your mark too. When the owner complains about paying manually in various ways to produce the And we’re going to prove the point by show$100 for a kick, the repairman scrawls on the desired sounds. Those circuits, despite being specialized for sound generation, were really nothing more than user-controllable voltage generators. In today’s virtual instruments, the voltages Our exercise today is to poke, prod, and cajole have been replaced by bits and bytes, but the concepts remain the same. Voltages generatyou into digging into your virtual synths and ed by one circuit are applied to another to some type of effect: apply voltage to a creating your own unique sounds. create filter, and the filter opens or closes accordingly. Apply voltage to an oscillator and the pitch rises or falls accordingly. Apply voltage to an amplifier and the volume raises or lowers ing you how to create and customize sounds bill: Kicking the refrigerator, $5.00. Knowing accordingly. using a pretty basic downloadable freeware where to kick, $95.00. So in the spirit of getting a handle on prosynth, called, appropriately, Synth1. In the synth world, knowing where to kick gramming your virtual instruments, a brief You can use any analog subtractive synth means knowing which (onscreen) knob to visit to the old days of voltage control is in to follow along in this article. But using turn. Yeah, that sounds so basic, but if someorder. Synth1 can help you realize how even the one says to you, “I want a brighter sound,” The basics components of subtractive anamost basic tools—when properly applied— that result can be realized in dozens of differlog synthesis are: can yield great results. ent ways. So knowing which knob to turn is Sound generation (oscillators)
www.magesy.ru VIRTUAL
INSTRUMENTS
15
f e a t u r e VI reverb, and so on. (Knobs too numerous to mention at this point, but almost sure to be found in your synth’s Effects section.) Now we’ve laid out a very basic roadmap for what knobs you might potentially turn to achieve a particular result. But what if you grab the right knob, twist it madly, and don’t hear any effect on your sound? Welcome to one of the biggest frustrations in synth programming.
Fig. 2: Some users (including this author) may find Synth1’s default mustard green color a bit off-putting. It’s easily changed in the Option screen (accessed by clicking the Opt button at the bottom center of the front panel). You can enter your own hex code to pick virtually any color; we choose red (990000), since Synth1 is reportedly modeled on the Clavia Nord Lead 2 red synth; to go with that, we changed the text from black to white. (Do an internet search for “hex color number 990000, and you’ll see how this works.) We also set the panel size to 200% to make on-screen programming a bit easier.
Harmonic/tonal shaping (filters) Amplitude/volume shaping (amplifier) Dynamic/real-time control (assorted modulators, such as envelopes and LFOs, and realtime controls such as mod wheels and keyboard velocity) Effects processing (reverb, delay, chorus, etc.). So back to our hypothetical: “I want my sound brighter.” If we work our way through the above list, we can accomplish that desired effect at every step of the way using different techniques. Please check out the sidebar Synth Filters 101. With an oscillator, we would choose a waveform (the Waveform Selector knob) with greater high harmonic content, e.g. a sawtooth or narrow pulse wave versus a square or sine wave. At the filter level, we would open up the filter (the Filter Cutoff knob) to allow more of the oscillator’s high harmonic content to pass through, or use the Filter Select knob to choose a filter with a gradual roll-off (e.g. 12dB per octave, or two-pole) rather than a steep slope (24dB, or four-pole). At the amplifier level, we would open the amplifier (the VCA Gain or Amount knob) to allow the full waveform to be heard (as opposed to keeping it soft, which would affect its perceived brightness). At the dynam16
VIRTUAL
ic control level, we would open our filter and amplitude envelopes (Filter and Amplitude Amount knobs) to allow the desired harmonic content to come through, and also adjust our real-time controls, such as keyboard velocity, to similarly open the filter and amplifier. Effects processing could include EQ-ing the sound, adding some high-end “glistening”
Nothing happens This common roadblock illustrates why it’s critical to understand the “voltage control” parameters of your synth. Let’s start with perhaps the most fundamental analog synth parameter of all, filter cutoff. In basic analog subtractive synthesis, the filter—most commonly a four-pole 24dB peroctave lowpass design—is used to remove harmonic content from the oscillator’s waveform. In other words, you start with a wave with all of its inherent harmonics intact and playing unfiltered at 100% of their amplitude, and then subtract harmonic content to create the desired sound. Dynamic control over the filter, i.e. changing the waveform’s harmonic content over time, is generally accomplished using the filter envelope. This is most commonly implemented in an ADSR (Attack, Decay, Sustain, Release) format, which allows for fairly sophisticated control, despite its simple design. Let’s say you want to create a simple blown horn sound. If you imagine a real horn, you can “hear” that in the initial blowing the sound is softer and less rich harmonically. You can approximate that blown sound in your virtual instrument by adjusting the filter envelope to have a somewhat prolonged attack time; in other words the filter will open up over time, allowing the sound to become more harmonically rich (brighter) after noteon. So you turn up the filter envelope’s attack time parameter. And your turn it up more. And you turn it up more—now to maxi-
Downloading the freeware Synth1 and the Patches Used in This Article: Synth1: Synth1 is compatible with Cubase SX, Fruity Loops 3.5, and Sonar (v1.0 and above). It can be found at: http://www.geocities.jp/daichi1969/softsynth/index.html#down.Select synth1107beta.zip at the top of the page. The site is in Japanese, but installation instructions—and an English language operator’s manual—can be found at the “Installation(english)” link on the same download page. We highly recommend printing the manual. Patches: Go to www.VirtualInstrumentsmag.com and click on the More Online tab, then download the Synth1 Patch Banks. Be sure to load What Knob to Turn.fxb so you can follow along with the examples used in this article (and get some cool sounds too!). The custom patches are located in Bank 2 of the download.
INSTRUMENTS
www.magesy.ru
www.magesy.ru
f e a t u r e VI secret of analog synthesizer programming officially revealed! If you want to hear the effect of your programming, you need to turn the appropriate “receiving” control to its minimum value so that the incoming voltage can “work.” A practical example: In the custom Synth1 patches we’ve provided at www.virtualinstrumentsmag.com, call up Bank 1, Preset 10, French Horn. Play a few notes to familiarize yourself with the sound we are going to create. Now call up Bank 1, Preset 11, Voltage Experiment. In this preset the filter is wide open and the Attack Time parameter is set to its minimum value. Play a note or two, and you’ll hear a basic percussive pad sound. Using this pad as a starting point, we’ll work toward creating our Preset 10 French Horn sound. Start by turning up the Filter attack time knob (A) to maximum. Notice that’s there’s no effect on the sound. Now turn the Filter amount knob (amt) to its maximum position; this routes the full output voltage of the filter envelope to the Fig. 3: Patches saved in Synth1 can be color-coded, providing a useful way to organize groups of sounds. Shown here is the filter. mini-bank of 15 patches created for this article. Still nothing! So we’ve proven the concept outlined a couple of paragraphs earlier: once the filter mum—and still hear absolutely no effect on “voltage” to your filter. If the filter is wide is wide open, no amount of voltage we throw your sound. open, that is, if the filter control (usually at it is going to make a difference. You look at your synth’s front panel and labeled as Filter Cutoff, FC, or in the case of With the Attack and Amount parameters notice that the Filter Envelope Amount knob Synth1, “frq”) is set at maximum, the filter is set to maximum, let’s use our minimum value (“amt” in Synth1), which controls how much already open 100% and passing full, unfiltechnique and turn the Filter Cutoff (frq) convoltage is sent to the filter from the filter tered audio. You could give it a bona fide trol fully counterclockwise, effectively setting envelope, is set to minimum. Aha! So you electric shock and it’s not going to open any the filter cutoff frequency to 0Hz. Now play a crank it to 100%…but still nothing! How is more. So whatever voltage you’re sending to this possible? it from the filter’s envelope generator is simGoing back to the all-important concept of ply being ignored. voltage control, the filter envelope is sending And there, my friends, you have it: the
Kicking the refrigerator, $5.00. Knowing where
Turn it up to 11
to kick, $95.00.
I
n our main story we’ve explored the importance of setting a “receiving” parameter to its minimum value so that incoming control voltage can yield the desired effect. But an equally important programming technique is selecting a parameter and “turning it up to 11.” The idea behind this approach is to exaggerate a parameter grossly so that you can home in on the effect you’re looking for precisely; once dialed in, you simply back down the parameter to the appropriate level. For example, let’s say you wanted to adjust the rate of LFO vibrato. By setting the LFO amount to 100% you would hear wild vibrato, and you’d have no problem hearing tempo of the vibrato as you adjusted it. Once the desired tempo was attained, you would simply back down the amount to the desired level. This technique is applicable for setting virtually any parameter, from envelope attack and decay times to pulse width modulation to tempo-based delays.
18
VIRTUAL
INSTRUMENTS
note and hold it (don’t give up—you’ll need to hold the note for quite a long time). Notice that the long attack time is fully realized, due to the fact that the filter was completely closed, so the voltage from the filter envelope is able to work to maximum effect. When the attack time comes to its end, the harmonic content of the note decreases rather quickly. The rate of decrease is con-
www.magesy.ru
www.magesy.ru
Playin’ around The mini sound bank developed for this article was designed to illustrate the basic concepts we’ve covered, and give you a starting place for developing your own customized sounds. Three groups of sounds in particular—Leads, Basses, and Wurlitzer pianos—demonstrate how subtle changes in parameters can often yield dramatic results. As you move from one patch to another within a group, take note of the sonic changes and the parameters that were adjusted to create them. (Clicking on a knob in Synth1 will yield a pop-up so you can see the exact parameter value.) The Wurlitzer electric pianos, for example, make use of simple changes in the oscillator waveforms (from soft triangle to harmonically rich sawtooth) and a gradual opening of the filter to create increasingly brighter and edgier sounds. The solo tones use the same technique, and also play with the filter decay control to lessen the percussive “knock” at the beginning of the sound gradually. We’re not going to tell what changes were made in the bass sounds to morph them from one to another. Hopefully, you’ll have gleaned enough knowledge from this article to figure that out for yourself! Keen-eyed synthesists will also notice that many of the patches make use of pulse width modulation, keyboard modes (e.g. unison, mono, poly), detuning, effects, and so on.
Synth filters 101 A
filter is the most basic sound-shaping component in a traditional analog synthesizer, since it fundamentally determines the harmonic content of the sound that will be produced. A filter shapes that content by attenuating (reducing the volume) of specific frequencies of the waveform being fed into it from the oscillator(s)— which is why this is called subtractive synthesis. Which frequencies are attenuated and by how much is determined by the type of filter used. Here’s a basic overview of the filters available in Synth1 and commonly implemented in other synths. Lowpass: As its name implies, this filter allows frequencies that are lower than the cutoff frequency to pass through unfiltered—set the cutoff frequency to 0Hz and no frequencies pass; set it to 20kHz and all sound passes through unfiltered. Lowpass filters are generally implemented in 2-pole and 4-pole designs, with the number of poles determining the steepness of the filter’s slope, i.e. how effective the filter’s performance will be. Each pole represents 6dB of attenuation per octave of frequency. So a 2-pole filter provides 12dB of attenuation per octave and 4-pole filter provides 24dB. The musical result is that 2-pole filters tend to favor sounds that have a relatively high degree of harmonic content such as bowed strings, while 4-pole filters are generally associated with fatter sounds such as basses. Highpass: A highpass filter operates in the exact opposite manner of a lowpass filter by allowing frequencies higher than the cutoff frequency to pass through unfiltered, while attenuating those below the cutoff frequency. It’s primarily useful for thinning out a sound and/or for getting rid of unwanted low end noise, such as subsonic frequencies (thumps, pops, and the like). Bandpass: Combine a lowpass and a highpass filter, and you get a bandpass filter. This filter attenuates frequencies both above and below the filter cutoff point, allowing a band of frequencies to pass through. In addition to being a sound sculpting tool, a bandpass filter can be very useful for homing in on the primary frequencies that define a sound, so when mixing, extraneous frequencies that might clash with other sounds can be eliminated or minimized. Resonance: A filter’s resonance control boosts the frequencies at the filter’s cutoff point—the equivalent of the “Q” control in an equalizer. We all know the classic “Rez Bass” sound; it’s created by boosting resonance, then sweeping the filter using the Filter Envelope Decay parameter. As with a bandpass filter, the resonance control can also be used to accentuate a sound’s primary frequencies to help create clean mixes.
trolled by the filter envelope’s Decay (D) parameter, and the level to which it decreases and stays for as long as the note is held is determined by the Sustain (S) parameter. Armed with this knowledge, we can quite easily transform our percussive pad sound into a blown horn simply by adjusting the attack time to a natural level (try 60 - 70 in Synth1). Further refinement of the sound is accomplished by adjusting the Filter amount (lower settings reduce the sound’s brightness,
but also lessen the effect of the envelope’s settings) and the Filter Cutoff. The balance between the Filter Cutoff and the Filter Amount will determine both the overall harmonic content of the sound and its dynamic harmonic content over time. There is no “right” combination of settings; it’s all up to your individual taste and the musical application. So experiment with various values to see their effect.
The voltage soundscape Synthesizer sound design using the voltage control concept is a lot like painting: there are thousands of ways to mix paint, and millions of ways to create trees, clouds, people, and the like. No one approach is correct; no one technique is absolute, and no one aesthetic is universally accepted. LFOs and envelopes and ramp generators and wheels and ribbons and keyboards are just voltage generators, and those voltages provide a nearly unlimited palette for creative sound design. We’ll explore more of this concept in the future. In the meantime…keep twisting those knobs! VI
Michael Marans fell in love with synthesizers immediately upon seeing (and hearing) his first one—a Buchla Series 100—back in 1966. He’s been programming them, writing about them, and consulting on their designs ever since.
20
VIRTUAL
INSTRUMENTS
www.magesy.ru
www.magesy.ru
r e v i e w VI
Vienna Symphonic Library Symphonic Cube The Vienna Instruments player is not your father’s sampler, nor is VSL his sampled orchestra.
Review by Nick Batzdorf
VSL Vienna Instruments, Standard/Extended/Both Libraries: Solo Strings, $415/535/950; Chamber strings, $595/715/1310; Orchestral Strings I, $595/715/1310; Orchestral Strings II, $535/655/1190; Woodwinds I, $595/715/1310. (Not yet reviewed but now available: Harps, Woodwinds II, Brass I & II, Percussion.)
www.VSL.co.at. US distributor: Ilio Entertainments, P.O. Box 6211, Malibu, CA 90265. 818/7077222, 800/747-4546, www.Ilio.com.
Platform: Mac OS X—AU, VST; Windows XP—VST. Can also operate stand-alone.
License: Uses a Syncrosoft USB dongle (sold separately for $23, or included with the complete Symphonic Cube) that limits one Collection library to one machine; licenses can be moved (if you purchase additional dongles) in order to divide the orchestra among multiple machines, and you can purchase additional dongles.
22
VIRTUAL
Vienna Instruments Control Edit screen. Each cell in the Matrix on the middle left of the interface contains a different solo cello Preset. The top row contains progressively faster legato transitions, called up seamlessly as you increase your playing speed (the Speed control at the upper right). You can’t tell from this screen, but the middle row has marcato bowings and the bottom row has spiccatos; this Matrix is using the mod wheel to choose which row (i.e. which playing technique) you want to play. Notice the “eye.” The inner meter displays MIDI velocity, the outer one shows the audio level.
he Symphonic Cube series, the flagship sampled orchestra product that Vienna Symphonic Library has been working toward since they released the first edition a few years ago, is here. That is, most of it’s here—we’re covering the initial release of the first five instrument Collections. The remaining five have since come out, and we hope to cover them soon. VSL’s Vienna Instruments Collections include a large amount of new material, and all the files are now in 24-bit format rather than the original versions’ 16; the difference in sound quality is definitely noticeable, espe-
T
INSTRUMENTS
cially when you add up an entire orchestra. But the most important addition is VSL’s remarkable new Vienna Instruments player, which we previewed in the December/January issue. This player is sure to have an impact on the playback features in all software samplers. Rather than being a tool for creating programs, its design is entirely focused on making VSL’s huge library manageable. It also makes it much faster to work with. Vienna Instruments takes the complete workflow into account, from organizing this gargantuan library to making its original
www.magesy.ru
Performance Tool features automatic: inserting various types of legato transition samples, and avoiding repeated samples of the same note. Above all, the player brings the vast quantity of VSL articulations under real-time control. Well, the last point may be overstating things just a little, since in the real world you’re still likely to do some tweaking to get the most out of VSL. But you could theoretically call up every articulation you need and play it live; the only practical limit is memory—the computer’s for loading programs, and your own for remembering which controllers or keyswitches bring up what programs. The other advance in Vienna Instruments is that instead of loading one articulation per MIDI channel and sequencer track, you can now just load everything you need for an instrument in a single instance of the player. So for example the entire solo violin can now be on one sequencer track and one MIDI channel, no matter how many articulations you’re using. But Vienna Instruments’ most dazzling trick is its Speed control—the ability to switch programs based on how fast you’re playing. Groundwork The Vienna Instruments Collections are available as ten separate libraries that can stand alone. So you could start with, say, the woodwinds (which are exceptional) or solo strings (also very good, and great for layering on top of other section strings to make use of VSL’s recorded legato transitions for added realism). Given the size of the investment required to purchase the entire Symphonic Cube—roughly $11,000—it makes a lot of sense to offer it this way. All the Collections come on Mac/PC DVDs, and the ones being reviewed range in size from about 30 to 55GB. That includes the Standard library for each, as well as an optional Extended library with a lot of additional material. You install both versions (they’re bundled together in the same files), and the Extended version—which you will almost certainly lust after—is unlocked for a 30-day trial period if you haven’t bought it. Please go to VSL’s site (www.vsl.co.at) for a list of what’s included in this library; suffice it to say that even the Standard libraries are remarkably comprehensive. We ran Vienna Instruments on a 2.8GHz Pentium 4 custom VisionDAW machine (www.visiondaw.com) with 2GB of RAM, and a dual 2.5GHz PowerMac G5 with 5GB of RAM; later we maxed out the G5 with 8GB of RAM, with some astounding results. The VisionDAW’s processor falls just short of the recommended 3GHz, but it’s well above the 2GHz minimum; according to the Macintosh specs you can run these libraries on a 1GHz G4, but VSL recommends a G5. The Symphonic Cube uses Syncrosoft USB dongles for copy protection (VSL calls the one they sell for $23 the ViennaKey). As part of the installation process you download one
license for each Collection you purchase. That license can go on any Syncrosoft dongle— you can buy extras if you don’t have them from other Syncrosoft-protected libraries— and you can transfer licenses from one dongle to another at any time. The content can be spread across multiple computers or you can access it from one location over a network, but only the machine with the dongle containing a given Collection’s license can play it. Of course you can move the dongles from machine to machine, but be careful when doing that, because Syncrosoft gets extremely unhappy if you attempt to start Vienna Instruments without the dongle attached. (Unlike some other Syncrosoft-protected programs, it doesn’t actually bring down the host sequencer if you forget the dongle, but it puts your computer into a stupor long enough that you will probably force-quit manually.) These libraries installed without any problems on both test machines. The Syncrosoft dongle is pretty much unobtrusive on the Windows computer, launching the Vienna Instruments player in about 15 seconds. The first time you run the program on the Mac, however, it takes about a minute and 40 seconds to check in with the dongle and launch; this delay is the same with the stand-alone program or the plug-in. But additional instances of the player come right up and samples load very quickly, considering the size of the programs. The player—overview VSL has always been a huge library. It’s known for having a staggering number of articulations that are pretty much uniform for every instrument, and it’s impeccably detailed.
The Symphonic Cube has far more articulations than even the earlier VSL editions. In the Vienna Instruments player these mapped articulations are called Patches, and they can still be loaded individually. The only real difference is that you can’t edit the actual sample files the way you could in the EXS24 and GigaStudio versions that preceded this one. To paint the picture of how refined this library is, take just one of the solo violin’s 18 main folders: Short and Long Notes. In addition to the different short bow playing techniques (detaché, spiccato…) this folder has a variety of vibrato choices for the sustained notes: with and without vibrato, progressive vibrato, marcato with vibrato, diminishing vibrato… The problem is how to access all those articulations. In the past, and with other libraries including VSL’s own Horizons and Opus series, you program performances by splitting each MIDI note/group of notes to a track assigned to the articulation you want it/them to play. That’s not necessarily a bad way of working, in fact MIDI programming is a unique skill. However, I think most people would agree that the way the Vienna Instruments player works is a big step forward even for musicians who prefer to program than to play. (Never mind that most of us use a combination of both.) What this player does is allow you to load up as many of these Presets as you need and arrange them into performance set-ups called Matrixes, where they’re ready to be summoned by a variety of MIDI commands. Lots of pre-defined Matrixes are included, you can create your own, and you can customize the commands for switching between Patches very easily. Finally, you can have several Matrixes
Keyswitching L
ike most modern sample libraries, VSL makes extensive use of keyswitches, which are notes on an unused area at the very top or bottom of the keyboard. Vienna Instruments Presets use keyswitches to call up different Matrixes. They’re an excellent control source, especially for scale runs—hit, say, an F# keyswitch and you bring up an F# scale and all its modes. Now, almost all sequencers since the ’80s have had a note-chasing feature that looks back to see which notes are sounding at the current location, and then sounds those notes as soon as you hit Play. The same goes for controller-chasing—they always know whether the sustain pedal is supposed to be up or down, that the mod wheel is supposed to be in the middle, and so on. Unfortunately, none of the MIDI sequencers on the market understands the difference between regular notes and keyswitch notes. Unless the keyswitch note happens to be “sounding” at the point you’ve located to, the sequencer doesn’t know to play it and tell (in this case) the Vienna Instruments player which cell—i.e. which Preset—it’s supposed to be on. Naturally, your music will sound all wrong. One workaround is to extend the length of every keyswitch note up to the beginning of the following one (many sequencers have commands for doing this). However, this is a nuisance, and it also doesn’t solve the other problem: keyswitch notes need to be kept off sequencers’ notation screens so they don’t appear in the score. In a moderately ideal world there would be a separate lane for keyswitch notes in the sequencer, and notes in that lane are excluded from the score. But in Nirvana the sequencer and sampler actually become one with each other: every keyswitch note is displayed with a name of the program assigned to it, making it blissfully easy to edit.
www.magesy.ru
VIRTUAL
INSTRUMENTS
23
The VSL library W
hen Vienna Symphonic Library came out with their First Edition for TASCAM GigaStudio and Apple Logic Pro’s EXS24 samplers a few years ago, it was hard to believe that anyone could undertake a sampling project of that scope. Since then two other major orchestral libraries have been released, making the whole concept somewhat more believable. But VSL’s custombuilt sampling studio, The Silent Stage, has been highly active all along, and VSL is still the largest sample library in the world. One of the first things you notice about VSL is that until you run it through a good reverb, most of it sounds totally wrong. That’s intentional. The Silent Stage isn’t dry, but it has a very short reverb time; the instruments aren’t close-miked, but they don’t sound like they were recorded from the audience perpective—and in fact they weren’t. This allows you to use VSL in many contexts, from contemporary to chamber to symphonic, depending on the artificial space you put it in. Each of the three major orchestral libraries has a general tendency. If East West Quantum Leap Symphonic Orchestra is bombastic and Sonic Implants Symphonic Collection is warm, I’d say that VSL leans toward the subtle and refined side. But of course it’s fully capable of getting bombastic or warm, just as the other two are capable of subtlety. As much as possible, VSL records the same set of articulations for every instrument. That helps make the library easy to learn, and you can generally switch or copy parts to different instruments very quickly. Two other interesting features of VSL are that in addition to
loaded, ready to be called up by a keyswitch. These big set-ups are called Presets. And that’s how you’re able to call up any Program in real time. Matrix reloaded One of the first things you notice about the Vienna Instruments player is that it displays meaningful information about the Patches, Matrixes, or Presets you’re going to load. That includes a description of the program and how much RAM it takes up. Just this seemingly basic librarian feature alone makes the library considerably easier to deal with, since knowing the programs you have to work with is half the battle (the other half is getting the sound of all the articulations in your head). The player doesn’t stop there, though—its interface is very clear, and it takes no time to learn your way around it. What does take some thought and (for me, anyway) constant refinement is how you set up your Matrixes and Presets. However, every instrument comes with a Universal Preset that has all the important articulations already set up for you. The Universal Presets don’t include every single articulation in the library, but you could be very happy using them and nothing else; they include a lot. There are also Universal Matrixes that contain legato, slightly accented, and fast playing techniques, each one in four different lengths. Consider that not very long ago, entire sample libraries weren’t nearly as extensive as the Universal Matrixes, never mind the much larger Universal Presets. But while the Matrixes are a manageable 24
VIRTUAL
the standard articulations, they recorded repetitions of each note so you don’t hear the same sample over and over; and they recorded legato transitions in the same breath or bow from and to every note of the chromatic scale over a wide range. To implement these samples they designed a program called the Performance Tool, which “listened” to your playing and sent out keyswitches to summon the appropriate articulations. Its Repetition and Alternation modes were sophisticated round-robin players; its Legato mode interpreted your playing. If you held one note while playing the next, it inserted a legato transition sample. But if you left a space between the notes, it played normal articulations. These features are built into the new Vienna Instruments player and no longer require a special program or any set-up. Subjectively, VSL’s best instruments are its solo woodwinds and solo strings. That doesn’t mean the other Collections reviewed here aren’t good—on the contrary, the strings are very flexible, and I especially like the new “harsh” playing techniques. But solo instruments done well are incredibly impressive; conversely, they can really stink up the place when they don’t work, because they’re so exposed. Still subjectively, VSL’s greatest strength—apart from the subtlety of expression you’re able to achieve with such a vast selection of articulations—is somewhat intangible: the notes all have a meaning. This is especially true of the short woodwinds; rather than just going “toot,” their possibilities are immediately obvious. That’s the great thing about sampling as a medium: it has an inherent life.
size, but there is one problem with the Universal Presets: they use RAM, RAM, and RAM. The Extended solo cello’s Universal Preset is 1013MB (the Standard library cello is 552MB); while they aren’t all quite that large, you can only run three or maybe four of these programs on a well-stocked machine. One way to reclaim RAM is to use the Optimize function—play Vienna Instruments your sequence, and it unloads all the samples that didn’t get triggered. That’s not something you’d want to do before you’re finished with a part, since clearly you can’t play samples that aren’t loaded, but there is a Reset button that loads the whole thing again. It can be shocking how much RAM you reclaim, even with complicated parts that use a lot of articulations. One solo violin part I Optimized caused the Vienna Instruments RAM usage to drop all the way to 15MB— down from over 900MB. Still, the real way to save RAM is to set up programs that aren’t quite as luxurious. That’s very easy to do, thanks again to the clever interface. Switching The Vienna Instruments player uses a system of “vertical” and “horizontal” cells, which are slots for Programs. There are up to 12 available in each direction; to load a program into one, you simply select it and drag the Preset you want onto the workspace area. It’s also possible to load two Presets into a single cell and either layer or crossfade them. Presets in a cell can be assigned independent level settings, start delays, 4-stage envelopes,
INSTRUMENTS
and release sample delays. The settings of these sliders are reflected in the “eye” of the selecting ring when you move them; normally the eye displays concentric MIDI velocity and audio level meters. There’s considerable flexibility in how you construct Matrixes and set up controllers to switch between Presets. In a typical set-up, you might have a few different types of articulations on the vertical axis—perhaps legato, sustained marcato, and staccato techniques. Or perhaps you want legato and portmento. However, there’s nothing to stop you from loading completely different instruments into adjacent cells (if you have more than one instrument Collection, of course), or layering two different instruments in a single one. That could be useful in live performance situations. Conversely, it’s common to use the same Preset in more than one cell if you want to keep coming back to it (it only gets loaded into RAM once). But we’re jumping ahead. Now, most of the “factory” Matrixes use the mod wheel to switch cells along the vertical axis. While the mod wheel’s physical location makes it a natural choice, you could just as easily switch Patches using any other MIDI continuous controller, the pitch wheel (so it automatically springs back to whatever articulation you have assigned to its middle value), velocity, keyswitches, or the ingenious Speed control. These switches are the same ones available along the horizontal axis, and you can edit (CONTINUED ON PAGE 62)
www.magesy.ru
www.magesy.ru
f e a t u r e VI
The 3GB Switch The superhero responsible for increasing GigaSampler’s memory access by a third explains how to tweak Windows XP so other programs can access about 75% more memory for loading samples. by Mattias Henningson
“640kb ought to be enough for anyone.” If Bill Gates really said those words or if it’s an urban legend is still held in darkness, but what we do know is that with the fast computers we have in our studios today the number crunching power is no longer necessarily the obvious bottleneck in our machines. The increasing need for RAM memory caused primarily by the modern huge sample libraries is taking over the role as our main enemy. As most of us still live in the 32-bit world with 32-bit Windows XP and 32-bit processors we’re stuck with limitations induced by the operating system that we simply can’t get around. That said, there are still things we can do to improve the situation quite a bit. A primer The memory space every application (process) in Windows XP has access to is 4GB. This is simply due to the fact that 4GB is the largest number that can be represented with 32 bits. Out of these 4GB the system is given 2GB for its own use. The end result is that a process can access a maximum of 2GB. This includes all things that the process loads such as plug-ins and virtual instruments. The 2GB given to the system is used by kernel-mode applications, operating system components and drivers. All this means that it doesn’t matter how much memory you install in your machine; every process will get a maximum of 2GB. As the system seldom needs 2GB for its own use, Microsoft added a way to switch from this even split between normal user-mode and kernel-mode applications to a more desirable 3-to-1 split in favor of usermode applications. This is done by setting a “boot switch” that 26
VIRTUAL
INSTRUMENTS
changes how XP configures itself on startup. When the switch is set, a user-mode application is allowed to allocate up to 3GB memory instead of the previous 2GB. Quite a substantial improvement and a healthy increase for all those memory hungry software samplers! Of course the switch doesn’t make much sense unless you really have 3GB or more physical memory installed. Giving it the boot The 3GB switch is an option you have to add in your boot.ini file. You can access this file by going into My Computer->Properties>Advanced->Startup and Recovery->Settings. In the following dialog press the Edit button. This should open a text file with content similar to this:
Before changing anything, please note that if you destroy this boot.ini file, your computer won’t be very happy when starting up. So take all precautions and medications. If you fail despite this warning, you can find a way out with a little help from our Microsoft friends by reading this article: http://support.microsoft.com/kb/330184/en-us.
www.magesy.ru
www.magesy.ru
f e a t u r e VI To maintain a safe route back to the standard configuration if your machine doesn’t like the 3GB switch for some reason, copy the last line and edit it by adding the /3GB switch. Also change the description on the changed line. In the example it could look like this:
Save the file and restart the machine. When you restart the machine, you will get an option to start either one of the two configurations you now have on your computer, your previous configuration or the 3GB-enabled configuration. If you select the 3GB and the machine starts up without problems, you’re done. There are cases when a driver, most probably a video driver, fails to load when setting the 3GB switch since the amount of kernel-mode memory in the system is too low. If that happens, it’s good to know that apart from going back to the even split, you actually have an option to specify the exact relationship between usermode and kernel-mode memory by adding another switch in the boot.ini file called /userva. Setting this to 2.5GB in the example above would make boot.ini look like this:
Nah, it can’t be that easy, right? No, of course not… There is one more catch. The host application (be it your main sequencer, another host, or maybe a stand-alone software synth or sampler) has to be Large Address Aware. Obey the LAA Large Address Aware, from here on denoted as LAA, is simply a single bit in the header of the application’s exe file that tells the operating system whether the application is able or not able to make use of addresses above 2GB if the system is configured to support them. What is really cool is that this bit can be set from the outside without having access to the source code if the application doesn’t support LAA already. The fact is that not many applications are delivered with this bit set even though the programmers seem to be of good breed and deliver code according to the rules that LAA applications must abide to. So far the only applications I know about with LAA support as delivered are the latest incarnations of Steinberg Nuendo 3 and Cubase SX3. Some of the other host developers such as Brainspawn with Forte and FX-Max with FXTeleport have confirmed that they are adding support in upcoming updates of their products. In any case it may be worth checking with the developer of the application before hacking it yourself. On the other hand, people say DIY is rewarding. So how do I add LAA support? Adding LAA The standard way of adding LAA support would be to use a command-line utility called editbin.exe, which is included with Microsoft Visual Studio and some other development packages. In this article, however, we will use an alternative Windowsbased utility I wrote specifically to back up this article called LaaTiDo.
A link to the utility can be found on the Virtual Instruments Magazine web site, and directly from me at http://www.musikbanken.se. Using LaaTiDo you’re able to: Check if an application has LAA support Backup the original file Add or remove LAA support The utility will guide you through this 5-step process: Browse for and select the application executable file. In this example we will add LAA support to Brainspawn Forte.
28
VIRTUAL
INSTRUMENTS
www.magesy.ru
f e a t u r e VI Check if the application already has LAA support by pressing the Check button under Step 2. Our Brainspawn forte case gives the following result:
As you can see the application does not have LAA support which means we have to add LAA support ourselves. First we really need to make a backup of the original file. When selecting the file in step 1 the application automatically proposed a location and file name for the backup.
Let’s say we’re happy with the name and location. Press the Backup button. If the backup operation was successful the “I have a backup…” checkbox is selected and the patching button for Step 4 is enabled. If we already had a backup we could just check the checkbox and continue. Press the Enable button and the exe file will be automatically changed to include LAA support. Upon completion it will look like this:
The increasing need for RAM memory caused primarily by the modern huge sample libraries is
Check the file once again with the Check button to make sure it reports LAA support. Test the application and enjoy the added memory space! After all, the worst case is that the application crashes if it didn’t follow the LAA coding rules.
taking over the role as
Some remarks It’s important to stress the need for backing up the original boot.ini file. You may need to restore the original file to be able to apply updates for the product. There is another way of breaking the 2GB barrier. If you somehow can use more than one program at the same time by using stand-alone synths or maybe a second host on the side, you will be able to use more memory even without the 3GB switch enabled. Don’t expect to ever get access to the full theoretical 2GB or 3GB. One cause for this is that memory just like hard disks gets fragmented. When the fragmentation gets really bad, XP can’t find a large enough block to give to the application when it requests more memory, and the request is denied. Don’t enable the 3GB switch on a machine on which you’re using TASCAM GigaStudio. As the sampler engine in GigaStudio is a kernel-mode application, it will suffer heavily from the 3-to-1 memory split. The upcoming Giga Virtual Instrument, GVI, will be a standard user-mode plug-in, however, and therefore it will to gain from LAA applications.
our main enemy.
The future There is a 64-bit/32-bit hybrid mode in the 64-bit Windows XP version that can raise the memory limit for 32-bit applications even farther. 32-bit LAA applications running on 64-bit Windows can access up to 4GB rather than 3GB, since Windows itself is not forced to use this particular space. The real solution to our memory needs is a 64-bit world where the current limits are totally blown away and maximum values are counted in terabytes rather than a couple of gigabytes. Unfortunately there are still not many alternatives if you decide to go with a full 64-bit solution right now, in terms of sequencers and hardware drivers. Cakewalk has already been there since last year with their native 64-bit edition of Sonar, and some of the major sound card manufacturers have released drivers for their cards, but that’s about it. The support is coming step-by-step and when the time is right the market will probably explode. Let’s hope that time will come sooner rather than later. VI
www.magesy.ru
VIRTUAL
INSTRUMENTS
29
i n t e r v i e w VI
Jerry Gerber A very different kind of poster boy for the V.I. medium
Interview by Nick Batzdorf
Jerry Gerber is a highly skilled “serious” composer who works with sample libraries and software synthsesizers. Rather than making “mock-ups” with these tools, he treats this as a serious medium of its own, with its own capabilities and unique challenges. While he has done a lot of network television scoring in the past, for the last decade Jerry has been living in San Francisco, producing extended “concert” works on CD. His music combines sample libraries, synthesizers, and live musicians (especially singers) very skillfully. Please go here and listen: http://www.jerrygerber.com/virtualinstrumentsmagazine.htm
30
VIRTUAL
INSTRUMENTS
You’ve commented that there’s too much emphasis on acoustic realism in our medium. But you’re working with the Vienna Symphonic Library. Here are my thoughts. I’ve been doing this for 20 years, and I’m still like a beginner. Every day I learn new things, and this is very complicated—I would need 200 years to master this. I know that. But I think what happens is a lot of people compare the virtual orchestra to a live symphony performance, which is absolutely absurd to me! It’s just nuts! There’s just nothing that’s comparable to a live symphony in a hall; that’s a unique experience, and no digital orchestra is ever going to match that. Now, when I way that it doesn’t mean the digital orchestra can’t be incredibly powerful or musical or expressive in its own right. But it’s going to do that with its own attributes and its own qualities. If you compare let’s say a high-quality digital recording of the New York Philharmonic to a really high-quality recording of a virtual orchestra, now you have a little closer comparison. At least now you’re dealing with the same medium: you’re listening to the music through two speakers. People talk a lot about panning and how important it is. I just don’t agree with that. Even if you’re sitting in the best seat in the house, right in the center, listening to music over two speakers is a very different experience to listening to music live in a hall. So to me the panning template doesn’t make any sense. That’s what I mean when I say How much of this do we owe to the symphonic medium, how much do we inherit, and how much do we just have to let go and say This is another medium that requires a different exploration? There are a lot of similarities, especially when you’re dealing with orchestration, because you’ve got the problems of transparency and orchestral weight and blend, you’ve got the orchestral problem of how to get every instrument to sound—and that’s really important, and that’s what we have in common with symphonic writing.
www.magesy.ru
and turn them upside down—all kinds of techniques that composers use, everything from contrary motion to retrograde to augmentation, diminution, change of interval, change of meter, change of accent…all that stuff.
Fig. 1: The staff view in Cakewalk Sonar of Jerry Gerber’s Symphonic Movement 1, a work in progress. He composes in this screen, entering the notes manually.
But other than that it’s an open field. I think that the new tools are so new that people don’t know how to approach it, so we naturally refer to the symphonic orchestra. And we should, because it’s an incredible musical achievement—I mean listening to 100 good musicians playing together in a nice hall is a profound experience. So if the focus were more on good music, good structure, good musical development, I think these problems of realism would take care of themselves. There are always going to be people who reject the medium because they think it’s supposed to be a symphony orchestra, and there are always going to be people who love the medium whether it’s a symphony orchestra or not. That letter I wrote (“Trends,” 4-5/06) was out of frustration. I got this letter from some supposed musician who was basically telling me to stop doing what I do! It was just a bizarre letter. I think to myself, if you’re against this medium or you don’t like it, just don’t listen to it! I imagine photographers got the same flack when photography first hit the scene and people were saying “Whoa, it’s trying to be a painting!” But people just have to do what they love, and do the best they can. Whatever you put out there gets criticized anyway. That’s just the way it is. Always easier to criticize than to create, huh? I find it easier. :) Since most of our readers are interested in songwriting or writing for media, we deal with short forms. Your works are extended. Can you say anything about the structure? I basically start out with two or three primary
ideas. In the movement I’m working on right now, there are three. One is a rhythmic motive: an eighth and two 16ths. It’s repeated. Another one is a 4-note melodic motive that goes [he plays]: G, Ab, G, Gb. Just those four notes. And then the other one is [plays a series of notes that start with a minor/major7 arpeggiated sound but continue]…it’s that theme. How I work the form is, those are the primary ideas and I develop them, make variations on them, repeat them, take them apart
How much of that technique is conscious while you’re doing it? Half of it. There’s always an intuitive part. For me, no matter how complicated the texture is melodically and contrapuntally, it has to work harmonically. I’ve heard a lot music that’s very complex and very intricate from a contrapuntal point of view, but I’m not crazy about the harmonies. My favorite harmony sometimes is Irish folk music. I’m a tonal composer, even when I work with 12-tone rows I end up working harmonies in a tonal way. But for me the big challenge compositionally is combining tonality with chromaticism. So much of pop music, for example, uses seven notes. You’ll hear modulations to other keys, but basically composers are thinking in terms of seven notes. But there are 12 notes. I think chromaticism creates some of the most interesting music, for example the Mahler symphonic movements. There are some sections where you use arpeggiated ostinatos, and people associate that with minimalism. Is that how you’re thinking? I think that rather than write minimalist music or write serial music or write this kind of music, I think I’m a pretty eclectic composer. I try to make styles my own and integrate them, but there are a lot of times and passages in my music when I’m not interested in chord progressions, I’m interested in building up a texture and keeping a very pronounced
Fig. 2: An event list in Gerber’s Symphonic Movement 1. Note how many controller moves and patch changes go with the uncluttered score shown in Fig. 1.
www.magesy.ru VIRTUAL
INSTRUMENTS
31
www.magesy.ru
www.magesy.ru
tonal center, and that will create a minimalist effect—sometimes by overlaying four or five different contrapuntal parts but not necessarily changing chords. To me, tonality is like gravity: you can play with it, you can suspend it, you can work with it, you can hide it…but you can’t get rid of it. Sort of like flying from here to New York—as long as those engines are on and they’re flying the airplane, the plane is not going to crash. But you turn those engines off, and immediately gravity comes into play. Music is like that in the sense that there’s a natural tendency to resolve tones. I don’t know how much of it is a result of the physics of sound and the overtone series, how much of it is cultural bias, and how much of it is just a result of how our brains have evolved. But I find that music that never resolves tonally doesn’t reach me deeply. In some ways, minimalism is a reaction against serialism. Whereas serialism was an attempt to completely eliminate tonal gravity, minimalism is an attempt to affirm the primacy of tonal gravity. I mix in both directions. It’s fun to play with tonality, though, and suspend it and hide it and not resolve things, and thwart listeners’ expectations sometimes. I guess the trick is to do that and still have it make sense musically. 34
VIRTUAL
You’ve commented that some of your less experienced students come to you with a concept of orchestration—they might bring some nice textures—but they don’t have a concept of the form. I think form is the last part of composer’s technique to mature, and it’s difficult to develop form if all you write are short cues for visual media. Most film music, the cues are anywhere from 20 to 40 seconds long, sometimes three minutes, sometimes even four or five, but that’s about it. So it’s hard to develop long form. Like anything else, you just have to practice over and over again. I don’t know how much of it is learning and how much is intuitive talent; there’s obviously a combination with everybody. I think in some ways the technology is very intoxicating and very seductive. Just like having a fancy word processor’s not going to make you into Shakespeare, having the most elaborate equipment’s not going to add to your musical chops unless you’ve studied music too. Where would you have a student start “practicing” form? I would go to the master composers like Mozart, Mahler, or Beethoven symphonies, or even Bach fugues—which are short but
INSTRUMENTS
incredibly complete forms—and I would ask the student to just sit there with a score and just listen to the piece 20 times over a week, and just absorb the textures and the way music is developed. That’s if they want to do that! Most students who have come to me want to do film music. I can understand that—it’s practical, it’s one of the few ways a composer can make a living. Very, very few, maybe a handful of composers in the United States make a living writing serious classical music that isn’t for film or television or computer games. If you’re writing for television, you don’t have to worry about form too much. The form’s dictated. But there’s also speed. Some people crank out ten minutes of music in a day. Well, it may be ten minutes of music, but what kind of music? When I was doing the Gumby television series and I had 750 cues over a year and a half, I wrote very, very fast—I had to do six episodes a month. But it wasn’t my best music. My best music takes time, it takes time for music to age. I mean, it took Brahms 25 years to write his first symphony! It’s not uncommon for composers to spend years—five, six, nine years on a work. (CONTINUED ON PAGE 62)
www.magesy.ru
Are you in the
Mungo Giveaway?
AmpliTube 2 offers everything needed to craft your own signature guitar tones with 5 separate modules including Tuner, Stomp, Amp Head, Miked Cabinet, and Rack Effects, on 2 series/parallel guitar rigs. 14 Preamps and EQs, 7 Power Amps, 16 Cabinets, 6 Microphones, 21 Stomp Effects, and 11 Rack Effects make AmpliTube 2 an incredibly comprehensive guitar Amp and Effects Modeling plugin. Ultra accurate modeling of the most sought after guitar gear, with unique DSM™ techology.?Over 80 guitar gear emulations, including popular amp models based on Fender®, Vox®, Marshall®, Mesa Boogie®, THD®, Supro®, and more. Stomp Box effects modeled after rare vintage gear including models based on Arbiter® FuzzFace®, Ibanez® Tube Screamer®, MXR® Dynacomp™, MXR® Phase100™, Electro Harmonix® Memory Man™ and more, with AmpliTube’s unique DSM (Dynamic Saturation Modeling) technology. Each individual guitar amp component is modeled separately, allowing you to create 20,000 different combinations. But it’s easy to use—no parameter is ever more than two clicks away—and neatly organized with over 400 different programmed guitar Amp and Effects preset combinations covering virtually any famous signature guitar tone you may dream of.
Enter to win an IK Multimedia AmpliTube 2 Guitar Amp and Effects Modeling Powerhouse
And while you’re there, we invite you to take a few moments out of
your schedule and subscribe to our favorite magazine before the upcoming price increase. Why would anybody want to miss the best writers in the industry, helping you get the absolute most out of your softsynths and samplers? You’ll pay only $12.95 a year (newsstand price: $59.40) or $20 for two years. You’ll be helping us print more issues and more pages, which will only benefit you (and us). Most importantly, you’ll congratulate yourself for having been so intelligent every time Virtual Instruments Magazine shows up in your mailbox.
Just go to www.VirtualInstrumentsMag.com/giveaway to enter for a Very Good Chance to win.
www.magesy.ru
r e v i e w VI
IK Multimedia Classic Studio Reverb Real-time control meets the character and flexibility from the glory days of outboard
by Lee Sherman
IK Multimedia Classic Studio Reverb, $399.
IK Multimedia US, LLC 1153 Sawgrass Corporate Pkwy., Sunrise, FL 33323. 954/846-9101. http://www.philharmonik.com
Formats: PC—RTAS, VST; Mac— RTAS, VST, Audio Units.
License: requires included Syncrosoft dongle.
36
VIRTUAL
s the name would imply, Classik Studio Reverb (CSR) is a suite of four reverbs modeled on the sound of popular outboard classics. While they may or may not be modeled after specific units, but they definitely strive for the quality, character, and flexibility that was a hallmark of the glory days of outboard. All the classic types are included: the bright metallic character of 1950s plate reverbs, hall and room reverbs for simulating the sound of acoustic spaces, and 1980s reverse reverbs that can really liven up a rock snare drum. The software is available for PC in VST and RTAS formats, and for Mac in VST, RTAS, and Audio Units; a TDM version for Pro Tools HD systems is under development. CSR comes with a Syncrosoft USB dongle, as has become the norm with high-end audio software. If you already have one Syncrosoft dongle, you can consolidate authorizations onto one dongle, saving USB ports.
A
Happy tails The four plug-ins—CSR Plate, CSR Room, CSR Hall, and CSR Inverse—share a straightforward common user interface that looks like a generic effects rack. You can tell which one
INSTRUMENTS
you are in by its color, and also by the labels on the controls, which change to reflect the type of reverb in use. The software has two modes. Easy mode has six knobs that control the most basic parameters, such as Mix, Diffusion, Reverb Time, Low Frequency Gain, High Frequency Cutoff, and High Frequency Damping. And Advanced mode lets you get under the hood to access over 100 individual parameters. These parameters include the size of the acoustic space, speed at which the reverb builds up, and modulation. That all of this can be done in real time is unprecedented in software. Advanced mode is also where you’ll find the modulation matrix. Four different modulation sources can be routed to up to eight destinations, allowing modulation of nearly any parameter from two LFOs or two envelope followers. The A/B comparison switch comes in handy when you’re trying to determine what affect a parameter has on the sound. Built-in presets ensure that even novices can get a great sound from the outset. For added movement, it is possible to create Macros that assign a single slider to con-
www.magesy.ru
r e v i e w VI
trol up to eight parameters in real time. Macros allow you to morph between various settings with the flick of a slider. You can limit the slider’s travel so that only specified minimum and maximum values are allowed. Real-time control over reverb has been around since the ‘80s, but few musicians took advantage of it since it was tricky to set up. IK has made it very easy to set up, so perhaps that will change.
In use Like the perfect party guest, Classik Studio Reverb fits into a mix without standing out too much. Where other reverbs are brash and obvious, Classik Studio Reverb is lush and pure. There’s little to no coloration of the original source material. Traditional reverb processing can’t match the realism of convolution processing, but it puts far less strain on your computer’s CPU
and allows real-time control. A single instance of the plug-in in Ableton Live on a dual 2GHz PowerMac G5 caused my CPU usage to rise about 10%, a perfectly reasonable increase considering sound. I tested the reverbs on a variety of source material, including drums, sampled grand piano, and vocals. As expected, CSR Plate and CSR Hall were outstanding on vocals, while CSR Room excelled at adding ambience to solo piano tracks. The real surprise in the suite is CSR Inverse. By applying an envelope to early reflections, it can turn standard drum loops into complicated polyrhythms. A good traditional reverb processor may not be the flashiest tool in the shed, but for precise control over the bread and butter sounds most musicians need daily, it is by far the handiest. VI
www.magesy.ru
r e v i e w VI
Linplug Octopus A boatload of distinctive and highly programmable electronic tones
Review by Jim Aikin t was the late ’80s, and Yamaha was sticking with FM synthesis (in the DX7II) while the rest of the world went crazy for sample playback workstations. The way I heard the story, a Yamaha manager insisted that their corporate vision was correct because “FM can make any sound!” He was wrong, of course. FM is pretty bad at realistic emulations of most acoustic instruments, though it does excellent mallet percussion, electric piano, and harp. One is bound to suspect that either this manager couldn’t hear the difference between the DX7II’s “strings” and a sampled orchestra, or he heard the difference but didn’t think it mattered musically. If we narrow the scope of the claim to nonrealistic, synthesizer-type sounds, it starts to make a lot more sense. A good FM synth is an extraordinary tool for producing a huge variety of tones and textures. Throw in a cou-
I
LinPlug Octopus, $149
LinPlug Virtual Instruments GmbH, Eibischstr. 6A, 12357 Berlin, Germany, www.linplug.com
Formats: Windows VST, Mac AU & VST
License: emailed authorization code
38
VIRTUAL
INSTRUMENTS
Fig. 1. In Octopus, FM modulation routings are in the matrix on the upper left, while envelope and LFO amounts are programmed in the matrix on the right. Waveform editing is handled at upper right, and individual envelopes are edited at center left. The effects and step sequencer are along the bottom.
ple of analog-style filters, additive waveform editing, some looping envelopes for producing rhythms, a simple step sequencer, and a few basic effects, and you’ve got a powerhouse on your hands. LinPlug Octopus, for instance. While it’s not the first software-based FM synth to hit the market, it puts a lot of desirable features in an affordable package. Its sound is warm, icy, fat, delicate, whatever you want. And calling it an FM synth is selling it a bit short. Basic sample playback is also included.
www.magesy.ru
Wrap your tentacles around it… Octopus continues the proud DX7 tradition in another way as well: it’s not exactly easy to program. LinPlug departed from the familiar FM design, in which each oscillator is tied to its own amplitude envelope generator, the pair being referred to together as an operator. In Octopus, the eight FM oscillators are in one part of the panel, while the envelope generators (up to 32 of them) are in a different area, as seen in Fig. 1. Each envelope can control any combination of oscillators, and any oscillator can be used as an FM modulator for any other. If this synth used patch cords, you wouldn’t be able to see the panel for the dangling spaghetti. The two filters have no frills, but they sound good. Each offers lowpass, bandpass, and highpass mode (no choice of rolloff slope). Three controls are included for cutoff, keyboard tracking, and resonance. The signal routing matrix allows the filters to be positioned in series, in parallel, or in a mix of the two modes. You can route some oscillators through the filter(s) while sending others direct to the output, and the filter output can also be used as an FM modulator. In another departure from tradition, Octopus has no LFOs. Instead, since all of the envelopes loop, any envelope can be used as an LFO with an editable waveshape.
I’m not completely sold on this design. First, editing the speed of an envelope-based LFO without altering its waveform requires moving at least two envelope points. Second, there’s no LFO onset delay (unless you choose to modulate LFO depth externally using a MIDI controller), because the attack portions of an envelope are skipped when it’s in LFO mode. Third, the random waveforms found on many other synths’ LFOs are not possible. The effects section includes chorus, delay, reverb, 4-band fully parametric EQ, and a distortion knob. Pushing the envelopes The manual doesn’t say how many segments the envelopes can have—more than you’re likely to need, it’s safe to say. Editing envelopes is easy, and they have most of the features we’ve come to expect of multi-segment envelopes, including tempo sync with snap-to-grid editing, adjustable curvature for each segment, and selectable loop start and end points. Each envelope is associated with its own multi-segment keyboard and velocity scaling curves, which affect the output level. Velocity and the keyboard can also modulate envelope speed, but you get only one parameter for each. This affects the speed of the envelope as a whole.
Envelope depth can be controlled via MIDI using mod wheel, pitchbend, aftertouch (mono or poly), breath controller, foot controller, expression controller, or CC 16, 17, 18, or 19. There’s no scaling or offset for these control inputs—they’re all or nothing. In practice, this means you’ll sometimes want to use two separate envelopes to control the same signal, one modulated by MIDI, the other not. This will entail a little extra editing, but with 32 envelopes available, you’re not likely to run short. Envelopes can be unidirectional or bidirectional, and amplitude envelopes can be additive or multiplicative. This is an unusual design, and very useful. Octopus maxes out at twelve notes of polyphony. Twelve should be plenty for most musical parts, but while using the synth in a song, I spotted a voice allocation bug that cut the actual polyphony in half (or worse) when the main amplitude envelope had a long release time. In a passage with repeated 6-note chords, some of the chords would have several dropped notes. Since I alerted LinPlug to the problem, they’re working on a fix. I also found that when synced and looped envelopes keep looping for long periods of time, they drift slowly out of sync with one another. If you strike a new note every eight
www.magesy.ru
VIRTUAL
INSTRUMENTS
39
r e v i e w VI bars or so, however, you’ll never notice the problem. Davey Tones’ locker Each of Octopus’s FM oscillators can produce the first 32 harmonics of the overtone series in any mix, with separate phase control for each harmonic. (A trivial bug causes the overtone display to be incorrect and inactive until you select an oscillator.) The usual pitch ratio and frequency parameters are provided. Keyboard level scaling of oscillator output level is not available directly, but you can set it up using an envelope. By programming sawtooth or square waves and then running them through the filters, you can easily turn Octopus into two parallel 4-oscillator analog subtractive synths—very nice. Situated alongside the FM synth is a parallel 1-oscillator sample-playback synth. This uses the same pair of filters, which means you can use a sample as an FM modulator (but not as a carrier) by routing it through a filter. Eight slots can be loaded with the samples of your choice, and each can be given a key
zone, but the zones can’t be layered atop one another. Looping can be switched on or off for each sample, as can keyboard tracking, and their output levels are adjustable. Octopus crashed two host sequencers (Cubase and Live) when I tried to load certain samples. (WAV file headers—don’t ask.) I’ve alerted LinPlug to the problem, and they’re working on a fix. But most of the samples that I tried loaded and played correctly. Steppin’ out When either of Octopus’s two step sequencers is switched on, the synth becomes monophonic, but when both are switched on, two voices sound, playing the two sequences independently (in sync, of course). Steps are always a 16th note long, but rests and sustain steps are allowed. The sequence editing features, while simple, include “shift left” and “shift right” commands, which are vital in this type of device. These sequences always start when you play a key on the keyboard and stop when you lift the key. Some sort of quasi-intelligent
keyboard-reading mechanism allows you to choose minor, major, or diminished patterns from the same sequence by playing several keys at once, but it’s undocumented, so I’m not entirely sure what’s going on. If you want to go even farther out, try grabbing a microtonal tuning from among the hundreds installed with Octopus. Tuning files of your own can be created using the freeware Scala software (Google “scala microtuning”). Sunken Treasure I’m a sucker for FM synthesis, so I like Octopus a lot. It’s versatile, and it sounds great. While I’d like to see a few enhancements, starting with more modes for the filters, Octopus is priced very reasonably, so if you’re curious about FM or just need a good all-around synth, dive right in. VI Jim Aikin writes regularly for VI and other music technology magazines. To learn about his latest novel, visit www.musicwords.net.
Waking up Steinberg V-Stack
random
tip
teinberg V-Stack is a popular program for hosting V.I.s and plug-ins on a remote computer. It’s efficient, it works reliably, it’s flexible and easy to use, and it’s only a $50 download. (There’s a Mac version too, but the following only applies to the Windows version.) But V-Stack has one bug that persists even after several versions: for some reason it stops responding to MIDI input after about ten hours, after which you have to launch it again and re-load all your programs. This happens whether you’re using MIDI interfaces or sending it MIDI over ethernet using MIDI Over LAN (Musiclab.com) as we are in this screen dump. Given that musicians who still have their computers fired up after ten hours are usually on a deadline, this seemingly minor inconvenience can be quite frustrating. Did you know that you can often wake up V-Stack’s MIDI input by hitting Play on the transport (lower right)? It doesn’t always work, but it usually does. While you’re slapping yourself on the forehead, we should mention in passing that if you’re trying to assign a channel strip to an output on your audio interface but only see the stereo bus, hit F4. That will bring up the VST Outputs window, where you can turn all your audio interface’s outputs on (see the light blue IO symbols to the left of each “lighted” bus at the upper right). Now they’ll show up.
S
40
VIRTUAL
INSTRUMENTS
www.magesy.ru
www.magesy.ru
r e v i e w VI
PSPaudioware Neon HR Linear Phase Precision Mastering Equalizer Review by Jason Scott Alexander
PSPaudioware Neon $149; PSP Neon HR $299
Contact: PSPaudioware; +48 60 196 31 73; e-mail:
[email protected]; Web: www.pspaudioware.com
Minimum System Requirements: PC – P4/Celeron 1 GHz or faster, 256 MB RAM, Windows XP; Mac – G4 1.25 GHz or faster, 512 MB RAM, OS X 10.3.8; Formats: DX, RTAS, VST, AU
License: PACE iLok key required for HR version (not included), Standard version uses code numbers.
42
VIRTUAL
SP Neon is a full-featured linear phase mastering equalizer. Linear phase means that the filters delay all the frequencies by the same number of samples; standard IIR (infinite impulse response) filters shift some frequencies in time more than others. Through a nasty little phenomenon known as transient smearing, that phase distortion is usually perceived as a loss of space and depth in recordings—and that includes samples, which of course are also recordings. There are many IIR equalizers that sound good in spite of that, but for applications in which you want to maintain a pure signal, phase linearity is a good thing.
P
Overview PSP Neon opens to a clean and uncluttered interface; despite having an extensive array of features, using it is completely intuitive. Like all of PSP’s plug-ins, Neon’s graphics are designed to look like a piece of hardware. This is enhanced with touches like a Power
INSTRUMENTS
button that bypasses the unit, and a zoom wheel for the waveform display. Neon is available in two versions. The standard version operates at 44.1 kHz and 48 kHz sample rates, or for roughly double the price you can get PSP Neon HR (high resolution), which adds support for sample rates up to 192 kHz plus a special ‘FAT’ (Frequency Authentication Technique) smoothing algorithm we’ll discuss later. There are two other differences with the HR version: it can equalize each channel in a stereo program separately, and it can work with mid-side stereo sources. MS is a popular, mono-compatible stereo recording technique that uses a cardioid (directional) mic pointing forward and a figure-eight mic to pick up the sides. PSP Neon offers eight bands of equalization, each of which can be assigned one of seven filter types. Those are: low cut 12dB per octave; low cut 24dB per octave; low shelf; peaking; high shelf; high cut 12dB per
www.magesy.ru
r e v i e w VI octave; and high cut 24dB per octave. You can engage/disengage bands individually, and each band is given a unique color so you can differentiate easily between the active bands. A large soft blue “LCD” shows you the curves for active band in its own color, and the overall equalization curve is shown as a thick black line. If you’re using a stereo instance of the plug-in with only one channel active, a thin black line shows the other channel’s EQ curve. Instead of using the frequency, gain, and Q controls to adjust EQ curves, you can simply grab a filter’s adjustment node. These are shown as small colored dots along the curve (see Figure 1). The EQ graph also includes an info bar along its bottom, offering pertinent information on the current filter settings. Systems The higher-end PSP Neon HR includes the standard version, so you don’t need to download it separately. System requirements include a 1GHz Celeron/Pentium PC or Mac G4 processor or better, running Windows XP or Mac OS X 10.3.8 or newer. The standard Neon version uses code entry authorization, while the deluxe Neon HR version uses a PACE iLok hardware dongle that you must purchase separately (if you don’t already have one). I have long since removed the PACE iLok drivers from my PC after noticing that they cause performance dips, and therefore I only tested the standard Neon version on that machine (a 3.2GHz Pentium 4
I found the Reset tabs to the left of each parameter row quite handy for clearing all eight bands in one click. In order to do its magic, the plug-in handles audio in fairly large chunks called frames. The optimal size of these frames in samples depends on the length of the filter, or the
Fig. 1: Instead of using the standard knobs to adjust EQ curves, you can simply grab a filter's adjustment nodes (the small colored dots).
plug-in’s operating resolution. While the standard version’s filter size is fixed at 4096 samples, the HR has two additional sizes: High mode doubles it to 8192, and Max quadruples it to 16384. Kicking in FAT mode doubles the filter length again. This oversampling mode curbs the warping caused by digital filtering, as I easily detected on a variety of source recordings through A/Bing using the FAT toggle. With it engaged, music sounded much smoother with more
I swear by each of the exquisite EQs I compared Neon to, yet none comes close to delivering the transparent high frequency detail I heard from it in Linear Phase mode. running Windows XP SP2 and using Steinberg Nuendo as a host). However, the iLok is fine on Macs, so I installed both versions on a dual 2GHz G5 with 4 GB RAM, Mac OS X 10.4.3, and Pro Tools|HD Accel hardware with a 192 I/O interface. Performance The peak or cutoff frequency for each of the eight bands can be set from 20Hz to 20kHz in very precise 1Hz steps, with Q ranges from 0.05 to 20.0 for peak filters and 0.35 to 1.41 for shelving and 24 dB/octave cut filters. The Q of an equalizer is the frequency divided by the bandwidth; the higher the Q, the narrower the range of frequencies that gets affected. Gain knobs yield ±24 dB maximum cut or boost in 0.04 dB steps, and double-clicking on any control allows for manual value entry.
For this reason PSP Neon has a minimum phase error Infinite Impulse Response (IIR) mode, in which the latency is extremely low and is reported back to the host correctly. To be honest, though, you’ll have a hard time going back after listening to Neon in Linear Phase mode.
analog-like high frequency response. This requires a lot of power, though, so the instructions advise you to raise the host program’s buffer very high. Larger buffers lower the CPU load (up to a point), but not all hosts or audio interface drivers offer a large enough setting, which makes Neon use a lot of computer power. Furthermore, because the linear phase filters are long and their length varies according to the resolution setting, the latency isn’t constant. And it’s significant—delays can be well in excess of one to three seconds! So as to not overtax your host application, PSP Neon doesn’t report its latency correctly, so the latency can’t be compensated for in linear phase mode. This is typically not a problem in a mastering setting with everything running through the equalizer at the same latency, but it is an issue in other applications.
[As we went to press, PSP informed us that version 1.1.0 will include the option to report latency in LP as well as IIR modes, a fixed latency mode, and an internal overload for Pro Tools TDM users.] Bright lights I swear by each of the exquisite-sounding nonlinear EQs I compared Neon to: Sony Oxford OXF-R3 (7-band), Massenburg DesignWorks Hi-Res Parametric EQ 2.0 (5band), and Univeral Audio’s Cambridge EQ (5-band). Yet none comes close to delivering the transparent high-frequency detail I heard from PSP Neon HR in Linear Phase mode. Regardless of the sample frequency, absolutely no coloration (i.e. phase distortion) ever appeared, and the material I was mastering benefited from Neon’s open sound. Even with fairly substantial curves thrown at them, hi-def recordings seem to pass right on through with transients left unscathed. Typically problematic high-dynamic GigaStudio string ensembles gained incredible depth and clarity across the stereo field, quite unlike any digital EQ has afforded me in the past. With FAT mode switched on, PSP Neon HR sounded just that much silkier on the highs with not even a hint of aliasing, and the double-precision extension of the filters allowed me to punch up low frequencies with extremely high Q boost settings. The 15 - 20% CPU hit to the beefy test machines with both Linear Phase and FAT engaged in Max resolution modes shouldn’t come as much surprise, and it shouldn’t sway you at all from considering this plug-in for your mastering stage. Used as channel insert EQ in its Standard IIR mode settings, Neon sounded better than some processors that cost up to three time the price. As totally transparent and highly flexible sculpting tools, PSP Neon and PSP Neon HR truly shine. VI
www.magesy.ru
VIRTUAL
INSTRUMENTS
43
f e a t u r e VI
more
online
www.virtualinstrumentsmag.com
Download the scores, MIDI files, and MP3 that go with this article at www.VirtualInstrumentsMag.com
From Sketch to Score Translating sequenced music to a score for live players presents some unique challenges. The highly skilled and talented young composer Thomas J—who’s also unparalleled at orchestral MIDI programming—goes over his process for doing that, offering pointers along the way.
By Thomas J. Bergersen
In their traditional role, orchestrators take composers’ sketches and flesh them out for full orchestra. They often get a piano sketch or a couple of staves’ worth of music from which they’re expected to make colorful instrumentations. In this day and age, however, more and more media composers are working in sequencer-based composition environments, doing the bulk of the actual instrumentation as they go along. Sometimes they don’t even have the training to write for real orchestra. For these reasons, the work associated with the “orchestrator” position in the film music industry is perhaps in need of a slight redefinition.
44
VIRTUAL
INSTRUMENTS
www.magesy.ru
f e a t u r e VI In this article I’ll offer a view into the role of modern orchestration and go through an example of the process. People work in different ways when transforming a MIDI sequence into an orchestral score, but what they all have in common is that there are no shortcuts. This article describes my way of working, from the point of view that you’re the composer and you’re working with an orchestrator. However, all this applies if the roles are reversed—and most applies if you’re filling both roles. As an example, we’ll use the first 12 bars of the included piece “Mojo Madness,” composed for this article, and follow the process as the MIDI file is transformed into a full orchestral score in the popular notation program Sibelius (see sidebar). You can download PDF versions of the score and also an MP3 of the audio file at www.VirtualInstrumentsMag.com, then click on More Online. Samples that don’t translate One of the first problems you’re likely to run into when you work out your sketch in a sequencer is that samples don’t always translate directly to a score. Examples of this include special effects not covered by samples, player-specific instructions, and irrelative time signatures. You might use a measured
The parts have been transferred into the Sibelius notation program and the work begins.
Figs. 1 & 2: The clarinet part is cleaned up in Cubase by first selecting the notes, quantizing the attacks to 16th notes, and finally quantizing the note-offs to 16ths as well.
tremolo sample that works in 4 but not in 3, for example; in the mock-up it just appears as a long note. When I work with orchestrators I make a habit of noting these occurrences as they appear in the work using MIDI markers. This makes it easy for the orchestrator to transfer the idea to paper, even if the mock-up is lacking. More on this later. Similarly, if you’re using sampled snippets
of an orchestra playing effects or tutti hits, etc., you should try to discern exactly what is going on in the orchestra, and either detail this in your instructions for your orchestrator/arranger or try to mock it up if time permits. You may also provide a set of isolated sound bites as an aid for dense parts that may have drowned as a result of less than ideal instrumentation, too much reverb, or simply overly complex writing.
www.magesy.ru
VIRTUAL
INSTRUMENTS
45
f e a t u r e VI dynamic marking for the passage in question. The more markers you offer the orchestrator, the more accurately he will be able to reproduce your musical intentions without it turning into guesswork. Sorted details To make the job yet easier for the orchestrators, it’s a good idea to keep your tracks in the order of the traditional orchestral score layout (picc, fl, ob, cl, bsn, hn, trp, trb, tba, vln 1, vln 2, vla, vc, cb). In the “Mojo Madness” MIDI file, the unused tracks in my 630-track template have been deleted and only the active parts are left in. Due to several practical considerations related to the set-up of my template, the MIDI file doesn’t follow the standard orchestral arrangement to a T. However, maintaining a certain structure avoids confusion down the line, particularly in large-scale projects. If your MIDI tracks are all over the place, you’ll constantly be searching up and down for that particular track. Chances are you’ll even miss crucial parts of the instrumentation. Quantizing I know some composers who like to quantize their tracks before they export the MIDI file. The idea is to make the notation software represent the music correctly, so you won’t have to input all the parts again. Since the basic rule of a realistic digital orchestra simulation is to avoid any quantizing (in my opinion), your sequence will invariably contain offset notes. For example, if you’re playing a passage with a violin patch that has a slight attack ramp you may be subconsciously compensating for that delay by playing a bit early on the beat. These notes will translate into nonsensical notation like 32nd rests, or 64th slurred notes at the beginning of a bar/beat. Quantizing the music before you export the sequence will lock the notes properly in musical time. Note that no sequencer as of yet has the ability to automatically detect and discern between triplets, dotted 8ths, or complex rhythmical patterns. Therefore you will have to go through each instrument part and apply the correct quantization setting to the applicable section. In Figs. 1 and 2 you can see how the clarinet part is cleaned up in Cubase by first selecting the notes, then quantizing the attacks to 16th notes, and finally quantizing the ends to 16ths as well. This can be a long and tedious process.
The score has been completed and is ready for part extraction.
Mark my MIDI If you have a good idea what you want, it’s a good idea to get in the habit of sequencing as much of the music as your samples will allow. This will avoid surprises when you hear it played by the live musicians. (The exception is if you’re working with orchestrators whom you want to give more creative input.) Make notes! Dynamic markings will save you and/or your orchestrator lots of time at the notation stage. Please import the included 46
VIRTUAL
MIDI file “Mojo Madness – Uncleaned Midifile.mid” into your sequencer, and take note of the markers. Markers are critical in this work situation. The first marker, “8 clix free,” indicates that the piece is to be performed to a click track and that the musicians should be given eight quarter note metronome clicks before the first bar. This is a count-off, in other words. Then the second marker, “dyn: f – aggressive” denotes the composer’s intended
INSTRUMENTS
Export/import Export the sequence to MIDI and import it into your notation software. When you import the sequence your notation software will probably give you a set of options. It can take some experimentation to get the most accurate representation of the MIDI file, so spend some time with this until you’re happy with the way things look. Note that unless you quantized properly,
www.magesy.ru
f e a t u r e VI the notation software will never represent your MIDI file accurately. In Sibelius you are given a standard set of import options (see Fig. 3). Finding the right combination of minimum note values and triplet formatting can be crucial in order to get the import looking as neat and clean as possible. Create a new orchestral score (I have created several manuscript papers/templates in Sibelius for various musical purposes), and import the MIDI file to a separate window. Unless you have a very large monitor, for this process it’s worth investing in a dualThe imported MIDI performance hasn't been quantized, and it needs some work to become readable. monitor set-up—one for the MIDI sketch view and one for the ochestral score. Some orchestrators prefer to print out the Madness score is transposed score with an all the parts. We could fix the notation and MIDI sketch first and transcribe from there, in open key signature. add small notes—“trill begins here,” “flutter which case a dual-monitor setup obviously ends here,” etc.—as we go along, but I usualisn’t necessary. When you create the score, be Adaptation ly reserve this for the next (and final) stage. sure to use a template that matches the size Now it’s time to begin adapting the score. If we didn’t quantize earlier, we’ll usually of your work and represents your instrumenStarting with the first instrument group, we have a pretty messy reference (see the score tation. move or copy all the parts from the MIDI excerpt above and the more complete PDF Set all the time signatures and tempo sequence one by one into the new orchestral version on the VI site: “midi import – changes up front. In general, people have score. messy”). Since we did quantize, we probably grown accustomed to the use of C-scores Let’s say the flute is on top of the list. have a more readable reference to serve as (meaning concert scores that are not transThere’ll probably be various flute tracks, such our base (see “midi import - cleaned” here posed) and open key signatures (accidentals as sustains, staccatos, runs, flutter etc., so we and on our website). are written in as you go along). The Mojo proceed systematically until we’ve transferred If we encounter parts in the MIDI sequence
The Sibelius notation program in a nutshell S
ibelius (www.Sibelius.com) is a professional notation program for Mac and Windows. Whether you need a full-blown composition tool or a just a means to transfer sequenced MIDI files to a proper score format, Sibelius offers everything necessary to create profession scores with a high quality look quickly. The program is simple to work with and stable as a rock. Its score layout is as customizable as you’ll ever need it to be, and the trademarked “House Style” concept provides full control over the look of the score, with details ranging from preferred positions for articulations, distances between notes, staff justification, and so on. For further flexibility, the program is fully object based, meaning that every score detail (note heads, beams, slurs, braces, etc.) can be resized, reshaped, and moved around freely on the virtual score sheets. Sibelius comes with an abundance of predefined templates (called “manuscript papers”) for different purposes: Big band, Brass choirs, various choir configurations, Army band, Guitar, Jazz quartet, Classical, Modern & film orchestra, String orchestra, Woodwind quartets, and so on. If you can’t find what you need, you can create your own quite easily. Inputting notes is also easy as pie, with several options ranging from real-time MIDI recording to step entry with the computer keyboard or mouse. The software is designed around key commands, and I strongly recommend a mouse with two buttons for the Mac version, just to make life easier. A handy floating window called the
“Keypad” gives you easy access to note values, markings, beams and accidentals, and the floating “Properties” window lets you control the layout aspects of the score, including staff type, font type/size, playback behavior and much more. Sibelius supports plug-ins, created with a built-in programming language called “ManuScript.” It offers user unparalleled control over the software since you can program plug-ins to do everything from proofreading to aiding in the composition process (retrograding, inverting, and bass figuration are just a few of the plug-ins the software ships with). Another plug-in that creates harp pedaling, one even checks your score for proper accidentals and/or respells them. Still another one checks for parallel fifths and octaves! The software has full MIDI support, should you want to use professional sample libraries with it. Configurable musical interpretation rules make sure that your score is played back with true musical expression, and in accordance with the dynamic markings and other annotations. You can also purchase a special Sibelius version of Garritan Personal Orchestra that automatically reads and interprets performance indications. Sibelius supports video playback for scoring to picture. If you need an easy and intuitive program to create beautiful looking scores, look no further. – Thomas J. Bergersen
www.magesy.ru
VIRTUAL
INSTRUMENTS
47
f e a t u r e VI where we’re unsure what is going on, we refer to an audio version of the sequence. Having an audio reference is important throughout the process. Check it regularly, and don’t forget to listen for dynamic changes. These may be subtle but nevertheless play crucial roles in the overall feel of the composition. The orchestra is extremely sensitive to dynamic changes, so I recommend taking advantage of that and making your music more expressive by giving them plenty of dynamic markings to chew on. The actual orchestration After clean-up, what first appeared to be fairly involved syncopated rhythms turn out to be simple eigth notes. work begins once the parts have all be transferred (see the PDF file on the VI site: “Mojo Madness – Stage 1”; in this stage the large modifications to the orchestration at this orchestrator every now and then, and as such parts have been transferred from the MIDI file point if you haven’t developed a good inner the scores are sometimes (very rarely, and to the orchestral score). ear yet. That is, of course, unless you’re using usually only with very dense orchestrations) First, we go through the piece and correct a sample set (such as East West Quantum sent to proofreaders just to make sure they’re all the notation errors, adding necessary Leap Symphonic Orchestra Silver) that gives ready for the score preparation process (the markings such as dynamics, playing techyou a fairly accurate representation of the process of extracting parts and creating the niques, and tempo indications. Then we changes you make within the notation softsheet music for the orchestra and conductor). VI adapt all the special notes (markers) from the ware. original MIDI sequence, look for mistakes in Thomas J. Bergersen (25) lives in Norway. voice-leading and harmony, and make sure Be an orchestrating Guy He is a composer working in the media music no orchestration holes that were present in Adapting MIDI music into full orchestral industry. His credits include hundreds of producthe original MIDI file have made their way scores is a long and cumbersome process that tions in film, TV, trailers, radio, and video into the final score. often leaves a lot of creative decisions to be games, as a composer, orchestrator, and At this stage we also make sure the note made. It usually takes about a day to transfer arranger. distribution among instruments in harmonic one to three minutes of fairly complex orchescontent is satisfactory. Among many other tral MIDI music things, the seasoned orchestrator will make to paper or sure woodwinds and brass have enough notation softspace to breathe. ware, which is Note the alternating 1st and 2nd clarinet why most film passage during the first couple of bars in composers Mojo Madness, for example. In that register don’t have the clarinet is at its most hollow, and so distime to do it tinctive in sound that even a single clarinet is themselves. enough to color the carrying string section In the PDF with a more moody tone. This is also the “Mojo stage in which we ensure that the bowings Madness – make sense, and that players with alternating Stage 2,” the instruments (and/or mutes) are given time to score has been make the switch. completed and Oftentimes a composer without a properly is ready for balanced orchestral template will produce part extraction. unrealistic volume relations in their Because such sequences. This may be due to harmonic an immense material played with large ensemble samples amount of (without compensating for the volume drop a work is carried divisi would implicate in a live orchestral setout in a short ting), or it may be as simple as a woodwind period of time section overpowering the brass. in the film It is then the orchestrator’s job to solve music industry, these problems and achieve the composer’s certain misFig. 3: The notation program Sibelius (see sidebar) greets you with a straightforward set desired relative levels among the instruments takes are of import options. in the orchestra. If you’re doing the orchestrabound to slip tions yourself, don’t be tempted to make past the 48
VIRTUAL
INSTRUMENTS
www.magesy.ru
www.magesy.ru
r e v i e w VI
Uberschall Urban Jointz A collection of useful “modern hip-hop and RnB flavas” seful.” That’s the word that kept coming to mind while auditioning Ueberschall’s Urban Jointz. From the musical arrangement of the samples to the layout of the disc, it became clear that a lot of thought went into making this library as useful—and easy to use—as possible. The core of the library is a set of 22 song Construction Kits, ranging from 78 to 114 bpm (most in the mid to upper 90 range), typically eight bars in length, with the keys listed. Each kit comes with a loop of the full mixdown, one of the entire percussion section, one of the entire melodic section, and then breakdowns of these sections (such as just kick drum, just rhythm guitar, and so forth). Even if the patterns are repetitive, all loops in a kit are the same length to make it clear how to layer them against each other in Acid, Soundtrack, Live, or similar programs. This makes it quick and easy to create breakdowns of the arrangements. Each kit comes with a companion Construction Singles folder that contains 5-18 individual hits (such as kick, hat, clap, synth sting, etc.) from the kits to help you augment the already-performed loops. The clean, notoverwrought compositions leave room for your own additions. In addition to the kits is an Additional Content folder. It contains 50 samples of MC vocals of varying length—all 99 bpm, all in the same key (which can then be warped by your application)—that includes solo and harmonized exclamations and raps, many of them rhythmic in delivery so as to fit well against drum tracks. The same goes for the
“U
Review By Chris Meyer
Urban Jointz $99.95
Ueberschall (www.ueberschall.com; distributed in the US by East West: www.soundsonline.com)
Formats: cross-platform DVD with 44.1kHz 16-bit Acid-friendly WAV and GarageBand/Soundtrackfriendly AIFF files; Rex2 format and Elastik soundbank also provided
License: Free to use as part of a musical composition including multimedia and library music, as long as any sample, loop, or sound is combined with another sample, loop, or sound.
50
VIRTUAL
INSTRUMENTS
140 - 99 bpm half- to 4-bar DJ scratches of voices, drums, orchestral hits, and effects— most are rhythmic so you can easily fit them into a song, in comparison to the harder-touse freestyle vocals and scratching that come with some other libraries. There is also a collection of 80+ 98 bpm 1to 4-bar clean and wah rhythm and simple lead guitar loops (all in C minor), which are nicely underplayed and as a result again fit in well with other parts from other libraries. This also applies to the 35 - 97 bpm 4-bar synth and electric piano bits, which also resolve around C minor. All these samples are duplicated in Acid (WAV) and Apple Loops (AIFF) folders. Most of the samples, minus the single hits and MC vocals, are also available as Rex2 files. Ueberschall has also created a soundbank to load into their Elastik VI plug-in (which we will be looking at in a future review). The cover declares “modern hip-hop and RnB flavas.” Indeed, the styles range from urgent rap through smooth lounge and R&B styles, with an emphasis on the melodic. Even Ueberschall’s license agreement is to be commended for its reasonable, plainEnglish approach. For example, “You may not use construction-kit-mix tracks, demo tracks, or demo songs, or any sample, loop, or sound that is a full orchestrated, ready-to-use mix for your musical production without combining them with another sample, loop, or sound. These rules result out of the requirement that at least a minimum of creativity has to be incorporated into a musical production before you gain the right to commercially exploit it.” Sounds fair to me—especially when a library has been designed this well so it could be commercially exploited. VI
www.magesy.ru
r e v i e w VI
cycles vol. 04: momentary incursions The latest volume in a unique series
Review By Chris Meyer
momentary incursions, $99
Cycling ’74 (www.cycling74.com)
Platform: cross-platform DVDROM of 44.1 and 48 kHz 24-bit WAV samples, plus audio CD
License: Free to use as long as combined with other sounds or not resold as another library.
e reviewed the first three volumes of the avant-garde cycles libraries back in the premiere issue last summer; this latest addition contains nearly a thousand “events, hits, FX, and transitions,” ranging from under a second to over a minute in length. Most of the sounds are oneshots without a listed key, although Cycling74 continues their tradition of including excellent documentation, including a PDF and Excel spreadsheet. (Just an aside: more and more vendors are going away from having a printed or printable index, which I sorely miss because I like to make written notes as to favorites when I audition a library.) One of the chief characters that sets the cycles libraries apart from other industrial, spacey, or avant garde collections is that a
W
surprising number of the sounds have an organic quality and a sense of…well, tunefulness. Yes, there are also a number of harsh synthetic blasts and random data streams, but other sources include horns (including the percussive quality of the keys and pads on a saxophone), guitars, percussion instruments, and other metals, which are then often heavily processed. Particularly interesting is the “events” category, which contains a large number of sounds that would be very effective stings or sound design elements for logos and other visual work. Although many of the samples are squarely aimed as percussive hits, bridges or interstitials in music, I personally think it is better to view momentary incursions as a collection of very cool, surprisingly harmonious sound design elements. VI
www.magesy.ru VIRTUAL
INSTRUMENTS
51
r e v i e w VI
more
online
www.virtualinstrumentsmag.com
Audeon UFO synth Something old, something new....
Review By Jim Aikin
Audeon UFO: $139. Light version: $22.
www.audeon.fr
Format: Windows VST (Mac version under development)
52
VIRTUAL
s software synthesizers proliferate, it has to be getting tougher for instrument designers to come up with fresh ideas for new programs. Oscillators, filters, envelopes—we’ve seen it all before. If anything, the trend seems to be to buckle under the pressure and code up a software version of an instrument that first achieved fame when Jimmy Carter was president. Bucking the trend, a small French company called Audeon has come up with a genuinely new spin on the concept of the oscillator. Their synth, UFO, simply doesn’t sound like anything else. It’s well named; unless coaxed to produce “normal” synth tones, it tends to sound vaguely like something that oozed out of the soundtrack of a 1950s sci-fi movie. This impression is reinforced by the deliberately retro panel graphics. Add an unusual 2dimensional modulation routing set-up to the mix, and UFO is anything but a me-too synth.
A
Plug Me In, Mr. Wizard! The basics of UFO (the letters stand for Unique Filters and Oscillators, by the way) are not exactly esoteric. It has two oscillators, two LFOs, three AHDSR envelope generators, two effect processors, and two formant filters in
INSTRUMENTS
Fig. 1: The two-dimensional control surfaces on UFO’s panel look very much like low-res CRT screens, and the signal flow is represented by jagged blue bolts of “electricity” straight out of a Tesla coil. The CRT parameters can be modulated vector-fashion, as shown by the colored arrows.
series. It also has a proprietary filter with a response that “morphs” from lowpass through peaking to highpass. I wouldn’t mind seeing another oscillator and another LFO added, and a little step sequencer for modulation would be sweet too, but there’s nothing essential missing from the voice line-up. UFO is 4-part multitimbral, so it can play separate parts on four MIDI channels at once, but some of the more interesting possibilities come from layering two or three patches on the same MIDI channel. (For an audio example, visit virtualinstrumentsmag.com. The phrase in the downloadable mp3 uses three of UFO’s four parts.) There’s no way to save a multi set-up, but since UFO is strictly a VST instrument this is not a big deal—the set-up will be saved with your sequencer file. Velocity, key number, aftertouch, and mod wheel are available as modulation sources
www.magesy.ru
(more on modulation below). In addition, UFO has a standard MIDI Learn set-up, and its parameters can be automated in the host sequencer. Oscillators UFO’s oscillators have two modes of operation, called K-Osc and V-Osc. K-Osc is less traditional, so let’s start there. The main parameters of a K-Osc are arrayed on two of UFO’s CRT displays. A display on the left has the words “Chaotic,” “Bright,” and “Mellow” arranged in a triangle, while the right display has the terms “Odd harmonics,” “Noise,” “Even harmonics,” and “Periodic” arrayed along the edges. Each display has a little white dot that you can drag around with the mouse, thus creating a waveform that is, perhaps, mellow and a bit chaotic, with strong odd harmonics and weak even harmonics but no noise. These settings are not ways of modulating the oscillator after a basic underlying wave has been produced—they control the manner in which the wave is generated in the first place. There’s also a slider called “Transient duration” below the left display. Low values cause a quick raspy burst of tuned noise during a note’s attack, while long values can be combined with high chaos settings to turn the
tone into a sustained gargle. By modulating a few of these parameters from aftertouch or a mod wheel, it’s pathetically easy to come up with a tone that sounds like a mad scientist’s unstable circuit. The V-Osc uses only one CRT display, with the words “Saw,” “Square,” “Pulse1,” and “Pulse2” at the four corners. No points for guessing how this works. There’s also a ring modulator, and either oscillator’s output can be blended with the ring mod out. The output of each oscillator can then be fed to either or both filters using a crossfade slider. Filters The resonant filters are switchable between 12dB and 24dB per octave. They can also smoothly crossfade from lowpass through peaking to highpass operation. In addition, each filter has a CRT for controlling cutoff and resonance in a 2-dimensional manner. The formant filters are standard. Each has a bandwidth slider and a CRT for adjusting cut/boost amount and center frequency. Another feature I’d like to see added would be a gain-compensated distortion stage on the output of the resonant filters. UFO is not too good at conventional “fat” synth sounds, and distortion would help. There’s no distortion in the effects section either, just the usual
chorus, flange, delay, and reverb. I had lots of fun running UFO through iZotope Trash, a hugely versatile distortion plug-in. Modern modulations When you click on the color-coded button to the left of any of the modulation sources, you’ll see little colored arrows on the CRTs and sliders that show where this source is being applied. The arrows can point in any direction, so an LFO can modulate the filter cutoff and frequency at the same time, for example. While the mouse work needed to edit the modulations is a bit finicky, UFO shows you what’s going on in a clear, intuitively obvious way. A single CRT dot or slider can be modulated by all eight standard mod sources at once, in varying amounts and directions, and that’s before MIDI Control Change messages are added. Using a touchpad controller like the high-end JazzMutant Lemur to control the CRTs could lead to some serious expressive possibilities. Individual envelope segments can be modulated from external MIDI sources. The LFOs offer sine, triangle, square, sawtooth, stepped random, and smooth random waves. Each has modulatable sliders for depth, rate, onset (CONTINUED ON PAGE 63)
www.magesy.ru
VIRTUAL
INSTRUMENTS
53
v e r y VI
d e e p
c l i n i c
Apple Logic Pro’s Ultrabeat Part 2 of a 2-part tutorial on the built-in groove machine. Last issue we explored the signal flow, assignment and drum mixer, and modulation capabilities of Apple Logic Pro’s Ultrabeat drum synthesizer. Now we’re going to dive into the synthesizer section itself, specifically its sound generators and filter. There’s not enough space to cover absolutely everything in Ultrabeat, but the end of this article you will be able use Ultrabeat to build your own drum Fig. 1 Ultrabeat offers four sound generators: two oscillators, a noise generator, and a ring modulator.
sounds.
tarting from the left of the synthesizer interface, Ultrabeat offers a selection of three main sound generators: two oscillators, a noise generator, with a fourth option being the ring modulator to the right of the filter section, as shown in Fig. 1. Logically, you might wonder why the ring modulator was placed after the filter, since as a sound generator it feeds into the filter. Because the ring modulator requires both oscillator signals (as we’ll describe later), it does logically come after the oscillators. These generators allow a mind boggling array of options for sonic construction, all
S
By Orren Merton
54
VIRTUAL
INSTRUMENTS
based on classic methods of sound generation—but served up the Ultrabeat way. More than just a phase Both oscillators can generate the classic waveform shapes we know and love from subtractive synthesizers. But rather than giving you the same old “analog-style” selections of sine wave, square wave, triangle wave, sawtooth, and so on, Ultrabeat goes farther by offering you phase oscillators. Instead of restricting you to selecting a specific waveform shape, the phase oscillators offer you the controls to shape the waveform man-
www.magesy.ru
v e r y ually. If that sounds complicated, don’t worry—it’s not. When you first click on the Phase Osc button on either oscillator, you are presented with a standard issue sine wave that will sound at the pitch set by the pitch slider. By clicking on and dragging the Slope knob, the Saturation slider, and/or the Asymmetry slider, Ultrabeat lets you shape that sine wave into other waveforms, standard or completely original. Turning up the Slope knob steepens the waveform, making the wave sharper and the tone more nasal. Turning up the Saturation slider flattens the waveform peak, ultimately changing the sign wave into a square wave. Turning up the Asymmetry slider tilts the waveform, changing the wave into more of a sawtooth waveform and making your tone more edgy. As you can see, by adjusting these three controls you can easily manipulate the waveshape into anything you want. For example, if you want a sine wave, leave the three parameters at their minimum value. To transform it into a square wave, increase the saturation slider to 1.0 (maximum). If you want a triangle wave, raise the slope knob to about -.17, and the asymmetry slider to 1.0 (maximum). But that’s kid’s stuff. Ultrabeat lets you shape the wave into far more than these traditional waveforms. Let’s say we want an edgy wave not quite sawtooth-edgy but close, and with just the peak of the waveform saturated. First click on the Slope knob and drag it to
d e e p
Fig. 3 In FM mode you can use oscillator 2 to modulate oscillator 1, and then you can modulate the FM Amount knob to modulate how much modulation the modulator is doing. Got all that?
volume of the oscillator. Please refer to part 1 for the clinic on modulation. What’s the frequency (modulation) Kenneth? If you click on the FM button next to the Phase Osc button, you will place oscillator 1
c l i n i c VI
tor 2’s waveform is being modulated by performance velocity, which affect’s how it modulates oscillator 1. Now let’s turn up the FM Amount knob to about .40. Finally, let’s modulate the fm knob using LFO2 (see Fig. 3). Now the amount of effect that oscillator 2 has on oscillator 1 is being modulated by both performance velocity and LFO2. And let’s not forget pitch and volume modulation that we can do as well!
Sample size Oscillator 2 offers three different modes. The first is the phase oscillator mode described above; clicking the Sample button next to the Phase Osc button engages Sample mode. You can load one of the included Ultrabeat multi-samples, or bring in your own samples in AIFF, WAV, or SDII mono
By adjusting these three controls you can easily manipulate the waveshape into anything you want.
into FM (frequency modulation) mode. In freor stereo interleaved formats. quency modulation synthesis, the frequency To load a sample, click the tiny downward of one waveform is modulated by another arrow at the top left corner of the waveform waveform. In FM mode, oscillator window to access the sample menu (see Fig. 1 generates a standard-issue sine 4). You can even audition sample files from wave that will sound at the pitch this window. If you click the play button in set by the pitch slider, and the the dialog window, you will hear the frequency of oscillator 1 will be unprocessed selected sample. If you click the modulated by oscillator 2. Preview Sample In Ultrabeat Voice checkbox, As you have guessed by now, if then you will hear the selected sample Ultrabeat excels at anything, it’s through the currently open drum voice, modulation. First off, you determine how much oscillator 2 will modulate the frequency of oscillaFig. 2 tor 1 using the FM By adding a tiny bit of saturation you can clip just the very peak of Amount knob. The comthe waveform. plexity of oscillator 2’s waveform determines about -.44. Then click on the Asymmetry slidhow complex (or how many partials) er and drag it almost all the way up, let’s say its modulation of oscillator 1 will be. to about .83. Finally, click on the Saturation And you can modulate the FM slider and drag it up slightly to a value of only Amount knob using any of the LFOs around -.04 or so—just enough to saturate and envelopes, or using parameters. Fig. 4 the waveform lightly (see Fig. 2). You can also modulate oscillator 2 Now you’ve created a nasal, edgy sound itself, which affects how it will modu- Click the arrow at the top left of the sample waveform window to load, unload, or show a sample in the Finder. with just enough saturation for that extra late oscillator 1. In other words, you punch. This waveform could be the basis for a can modulate the modulator while nice fat kick drum with a bit of grit, or a modulating the amount the modulator will including all its sound generators, modulasharp digital pulse, depending on the pitch modulate oscillator 1. tions, EQ, and so on—just as if it were already you assign the wave. The oscillator 1 waveform window gives us loaded. Pretty cool! If you want to give this waveform’s edgia view of how oscillator 1 is modulating oscilRemember, Ultrabeat is designed to be a ness a sense of depth and motion, you can lator 2. For example, let’s start with oscillator sample playback instrument; it’s a rhythm modulate the Asymmetry parameter of oscil2 in phase oscillator mode, generating a comsynthesizer that allows you to select a sample lator 1, and the Saturation parameter of oscilplex clipped sawtooth-ish wave with its satuin one of its sound generators. So don’t lator 2. You can also modulate the pitch and ration being modulated by velocity. So oscillaexpect any sort of crossfading or advanced
www.magesy.ru
VIRTUAL
INSTRUMENTS
55
VId e e p
v e r y
c l i n i c
sample keymaping or any of that. In fact, you might have become excited when you saw the “layer” slider, thinking that you could set layer crossfades. Unfortunately, only the factory samples can have velocity layers. You can, however, use the little red Play button to the right of the waveform to determine if you want your sample to play in reverse mode or forward. Still, Ultrabeat does let you adjust how oscillator 2 will use the sample you’ve loaded. Under the sample waveform display are two sliders: Min and Max. These two sliders let you select where the sample will start triggering at minimum velocity, and where the sample will start triggering at maximum velocity. You have completely free reign in adjusting these sliders. For example, let’s say you want your clap sample to play completely through at full volume, but to completely miss its loudest attack at low volume. Click on the min slider and drag it about halfway across the sample waveform. Leave the max slider at the beginning of the sample (see Fig. 5). Now the harder you play, the more of the sample will trigger; the softer you play, the closer to the middle of the sample oscillator 2 triggers the sample. Building models Oscillator 2 isn’t done giving up its secrets just yet. Not only can it use an analog-style or sample waveform as a sound generator, it can even model a basic vibrating string as a sound generator if you click the Model button to the right of the sample button (see Fig. 6). Again, Ultrabeat is not trying to include all of Sculpture in its model section—this is a basic percussive model useful in creating drum sounds. Your model string can be “tuned up,” so to speak, by dragging the dot inside the X-Y controller known as the material pad. The material pad lets you set the relative stiffness and inner loss of your string. Inner loss (the Y axis) corresponds to the high frequency dampening of the string. A higher inner loss value results in a smoother, more muffled string. Lower inner loss values will allow your string to vibrate longer during the decay phase. Stiffness (the X axis) controls the rigidity of the string. High stiffness values result in
a very rigid, inharmonic string; low stiffness values will result a more vibrant, harmonic string. Both parameters can be adjusted from 0 to 1. Your modeled string is triggered, or in Ultrabeat terms “excited,” by one of two exciters to the left of the material pad. Click the button to select either type 1 (a tighter, more “plucked” sound) or type 2 (a looser, more “bowed” Fig. 6 sound) to trigger your string. When oscillator 2 is in Model mode, Ultrabeat models the properFinally, you have the Resolution ties of a virtual string and the virtual “exciter” that vibrates the slider. Sonically, raising or lowering string. the value of this slider will increase or decrease the number and complexity of the overtones. What the Resolution slider is actually doing is increasing frequency and tone of your white noise. You and decreasing the calculation resolution of can set it to either high pass, low pass, band the modeling algorithm; as such, watch your pass, or off (bypass). Then you use the Cut CPU meter! The resolution slider has a miniknob to dial in the cutoff frequency, and of mum value of 3 and a maximum value of 30. course you can modulate it using it’s Mod The modeling oscillator is particularly useful and Via menus. for creating industrial-sounding percussion The filter also has a Res knob to boost the and organic-sounding “found drums.” For resonant frequencies immediately surroundexample, drag the dot in the material pad to ing the cutoff frequency. Like most analogthe middle of the inner loss (Y) axis, which style filters, lower values offer little resonance, corresponds to a value of .5, and leave the while higher values make the filter sound like stiffness at a low value, around .25. Now click it’s self-oscillating (feeding back into itself). on the type 1 exciter, and raise the resolution In addition to the filter, the Noise to about 20 (see Fig. 7). When you trigger Generator includes a Dirt knob. This knob lets your drum, it will sound like you are plucking you “roughen up” the white noise and make a harmonic, loose, percussive string, with it sound more grainy (or dirty). Low values moderate decay. will keep your white noise “pure,” while high Does it sound like any actual percussion values will give you a very grainy, almost disinstrument? Certainly not any stringed instrutorted noise. ment I’ve played! But it definitely sounds rich, And as if that’s not enough, you can even percussive, and ready for the rest of Ultrabeat. In fact, try using a unique, rich programmed physical model in oscillator 2 as an FM modulator for oscillator 1 as discussed above. You can create sounds using Ultrabeat you can’t create any other way. Remember, of course, that regardless of the type of sound generator you choose for either of the oscillators, you can modulate the pitch and Fig. 7 volume just the same.
Fig. 5 Use the Min and Max sliders to use your playing velocity to affect how the sample is triggered.
56
VIRTUAL
INSTRUMENTS
Drag the dot in the material pad to create your own virtual string; the results might or might not resemble any actual The other generastring, but its response will be modeled on real world materials. tors There isn’t enough room to go into every single aspect of Ultrabeat in a 2-part modulate the dirt parameter! The Noise clinic. But before we explore the Generator is especially useful for cymbal filters, modulators, and step sounds or buzzing noisy rhythmic sounds. sequencer, the other two sound The Ring Modulator to the right of the main generation engines in Ultrabeat filter is pretty simple—a single volume slider deserve at least a quick mention: and two modulation menus. But sonically its the Noise Generator and the Ring sound is very complex; it modulates the output Modulator. of both oscillators to generate its own sound, The Noise Generator creates— which means both oscillators must be on for you guessed it—noise, and offers the ring modulator to function. nearly as many parameters as the In other words, even if you don’t want one oscillators (see Fig. 8). It has its or both of the oscillators themselves to generown filter, so you can tailor the the ate sound for your drum voice, they still have
www.magesy.ru
v e r y
Fig. 8 Use the Noise Generator to both generate and shape white noise to use as part of your drum voice.
to be on to trigger the ring modulator. If you don’t want to hear one or both oscillators along with the ring modulation, leave them activated, but turn the volume of the oscillator to zero. Cold filtered The big knob in the center of Ultrabeat’s synthesizer section is the multimode filter, which also contains its own distortion unit (see Fig. 9). This filter is global in that it processes all the active sound generators that are sending signal into it. Of course, not all active sound generators need to send their signal through the filter, as discussed in Part 1. The large red arrow in the center of the filter determines if the sound generators will first pass through the distortion unit and then the multimode filter, or the multimode filter first. The multimode filter offers four filter types: low pass, high pass, band pass, and band rejection (band rejection means the bands inside the band are cut, not the bands outside like a band pass filter). You can select either a 12dB or a 24dB slope for the filter. Like the filter in the Noise Generator above, the Cut knob determines the cutoff frequency at which the filter begins to affect the signals, and it can be modulated by its Mod and Via menus. Unlike the Noise Generator filter, in the main filter its Res knob has its own Mod and Via menus, allowing the resonance to be modulated. The filter’s distortion unit can function as either a “bit crusher” effect (reducing the bit rate of the signal) or as a classic distortion effect (which doesn’t reduce the bit rate, but rather clips the waveform). If neither the Crush nor Distort type is selected, the distortion unit is not used. Regardless of distortion unit mode, you will have the same three controls. The Drive
knob determines the amount of distortion; lower drive settings produce less distortion (or bit crushing), higher drive settings increase the distortion (or bit crushing) of the filtered signal. Then the Color knob lets you adjust the tone of the distortion. Lower values result in a darker, warmer tone, higher values result in a brighter, harsher tone. Finally, the clip knob has a slightly different function depending on the mode of the distortion unit. In distortion mode the clip knob sets the output level of the distortion unit. But in bit-crusher mode the Clip knob determines the level the signal needs to be before bit crushing begins. For example, if you have the distortion unit in bit-crusher mode and set the clip knob to -8.000 dB, the level of the signal into the distortion unit would need to be -8.000 dB or higher before the bit crusher would start crushing bits. Grüv As you can see, there’s still more to Ultrabeat that what we’ve in these two issues. But by now you should have a firm grounding in how Ultrabeat is organized, how Ultrabeat handles modulation, and how Ultrabeat generates sounds. The included patches in Ultrabeat are a great starting point, but as with all truly advanced instruments, to
d e e p
c l i n i c VI
Fig. 9 All the sound generators in Ultrabeat can feed into the global multimode filter and distortion unit.
really unleash the ultra in Ultrabeat, take what you’ve learned here and get grooving! VI Orren Merton is the author of a number of books on audio software for Course Technology, including three books on Logic.
www.magesy.ru
VIRTUAL
INSTRUMENTS
57
f e a t u r e VI more
online
www.virtualinstrumentsmag.com
download the free Mind Meld algorithmic rhythm generator program.
Buick Sessions:
Sound design in music production Part 2 of our look at inspiration and production: a free algorithmic rhythm generator program. By Peter Buick
The Mind Meld algorithmic lyric and music program (discussed in the 12/05 – 1/06 issue) is still available for free download www.VirtualInstrumentsMag.com. While this issue is being printed, the new rhythm utility will be made available there too. (Hence any screenshots used in this article are subject to change in the final version—but we’re sure you’ll cope. The manual is included as a PDF in with the archive. )
58
VIRTUAL
INSTRUMENTS
Inspiration 101 Last time we looked at pitch and idea seeds for melodic lines coming from words and phrases fed in to the Mind Meld program, written by yours truly. We’ll be going back to looking at pitch later in the series, but this time let’s look at rhythm. Rhythm is an interesting part of our musical arsenal, as it supplies a wide gamut of choices we can apply to our (more limited) range and sequence of pitches. That is, as long as we can break out of the rhythmic mould. There seems to be a current (Western) tendency to follow a club dance mentality of four-on-the floor, superimposed with quarteror eighth-note on-beat guides/posts. There are of course, examples of off-beat rhythms, perhaps borrowed from reggae and even ’70s disco, and some intertwining of 16th note blips. But largely everything fits on the onbeat 16th note “grid.”
www.magesy.ru
f e a t u r e VI Similarly, arpeggiators in most synths are evenly spaced in time, and they start on the beat and stay on the on-beat. Only more original synths like the Wusik Wusikstation, Steinberg’s X-Phraze, and the Access Virus T1, to name just a few, dare to break that embargo. But even then they cross a fuzzy line called phrase sequencing rather than just an arpeggio. There are some good reasons for all this rigidity. The danceability of a structured rhythm and the rhythmical resolution (predictability) of a rhythm loop causes chemical reactions in the brain. So when we use “clever” rhythms in our productions, we need to consider how they supplement a provided framework. If we don’t provide the grid reference, the intricacies of the clever beat could leave our dancers with one leg in the air. We’re only using the term “clever” as a term to describe the concept of extraneous rhythms that add to the basic grid building blocks. We’re not suggesting intelligence.
make it skip. Or the off-beat tambourine between the kick drum pattern, which makes it jerk. These grace rhythms are totally valid, but there should be other ways… A straight kick drum beat. Solid but boring and lacking any groove.
See the website for the audio example SDMP02_03.mp3
Here is the clever percussion beat without the reference present at all. See the website for the audio example SDMP02_01.mp3
By adding an interleaved sound (in this case tambourine) we can add interest without destroying the dependable beat. See the website for the audio example SDMP02_04.mp3 Here we have a straight snare beat. No Drummer of the Year prizes
“.…the intricacies of the clever beat, could leave our dancers with one leg Once we add the grid framework, the percussion rhythm makes (more) sense. See the website for the audio example SDMP02_02.mp3 The importance of the reference should be quite apparent. But we need to add interest and contrast to this with the clever fill beats in order to maintain interest and originality. A common way to do this is with grace rhythm notes: the quieter fill notes on the snare beat that
in the air.”
www.magesy.ru
VIRTUAL
INSTRUMENTS
59
for this one. See the website for the audio example SDMP02_05.mp3
However, without a reference, the snare and grace notes are left hanging. So we can integrate the kick drum with or without its grace notes to help define it. See the website for the audio example SDMP02_07.mp3 One of the other ways of adding clever beats is to keep the original number of basic rhythm beats, but to change the voicings of selected notes to form a sub pattern that presents the ancillary rhythms. We can move the beats or copy them, depending on what story we want the original to tell still. Remember of course that silence (rests) can also be used as one of the alternate voicing sounds, leaving the imagination free to fill in the gap. And like nudity, often the suggestion is more appealing than the bare truth. Dynamics and effects can work in similar ways. See the production section below for more details.
By adding grace notes (usually quieter), either using the same sound or a contrasting one (in this case rim shot), we can transform this predictable cliché and make the feel shift. See the website for the audio example SDMP02_06.mp3
“….if you fill in all the gaps, you just get a hiss, and no longer get to hear the message”
60
VIRTUAL
INSTRUMENTS
We start with our single (beat-rich) tone pattern, such as hi-hat, shaker, or tambourine. See the website for the audio example SDMP02_08.mp3
www.magesy.ru
VI
Then we use two other contrasting voices to express some sub patterns by either moving or copying each note. See the website for the audio example SDMP02_09.mp3 One common error is adding too much! Often people create a fantastic groove—but then they fill in all the holes with extra notes, either using the same sounds or alternative ones. The point is, they remove the space and contrast that had originally made the beat groove. Most unfortunately, they tend to do this to the original, make no copy, and forget what they had just done before that. Production 202 In the last article, we looked at the dissection technique to take parts, or aspects of some clone copies, to produce (simultaneous) performance variations. Now that we have a rhythm in place, we’re going to explore using textures to arrange it. Textures From the rhythm discussion above, you can see that if we added all the rhythm part beats together and played them all with the same note and sound, we would end up with pretty much all the holes filled; it would not be what we call a “groove” at all—more of a machine gun. So we have naturally created a textural rhythm, out of our musical rhythms, by assigning different sounds per beat. Of course the truth is that this is a pretty straight beat and that it isn’t a groove, even with the textural icing. Most real grooves tend to have the important fifth element player: silence. So from that production viewpoint, once we have defined the song’s groove, we should be very careful about filling in any of those gaps. Take the analogy of Morse code played with white noise. If you fill in all the gaps you just get a hiss, and no longer get to hear the message. It’s not a great analogy, but hopefully you get the point! But we can enhance the role that textures have in making a groove.
Firstly we need to consider the importance of contrast and to review how we implement it constantly. Blanket and constant effects processing may set the mood, but they may also mask the subtle changes in the story. Equally, we don’t just want to keep changing things for its own sake. Repetition and predictability (musical and sonic patterns) do play a role in maintaining a listener’s comfort zone. But that has to be balanced against boredom and fulfilment. With synth sound programming there are some common examples of how the sound can be related to performance and production. The way the note is played could change the attack or release envelope of the sound. For instance, short notes could appear to have a longer decay (or reverb), while long sounds may end abruptly. A filter could open or close depending on the playing style. Extra oscillators or layers, could switch or fade in, depending on how many notes are being played, providing natural thinning on large chords and sound doubling on more simple parts. These programming techniques may not first appear to be part of our rhythmic and textural journey. But from the previous examples, one can see that using the sound on a per-beat basis can provide subrhythms. Meanwhile, we need to consider when we should and shouldn’t fill any gaps (with sonic changes or background hub-ub). Most importantly, we need to consider whether we are supporting or distracting from the piece each time we introduce a sonic change. Getting off the beat Whilst we tend to use the audio part of our host sequencers as mere tape machines for “live” stuff (like vocal and guitar), the audio facility opens up a plethora of creative production options. By “audio” recording the outputs of our (virtual) synths as audio parts, we can change the rhythm and textures in ways far beyond what the manufacturer had envisaged. Whilst there are some plug-ins and synths that cater to this beat-bybeat processing, they are unlikely to match the originality of the manual method. We can apply this to any sound in our sequence, not just the enabled few, and we can use or entire arsenal of plug-in processing of our duly enabled synths. This includes vocoders and processing audio through synthesizers. So going back to our earlier example of our on-the-beat and evenly spaced synth arpeggiators, we can record them as audio for processing. Then we can time-shift them, insert silences, rearrange the order, substitute beats with other sounds, add single note-effects processing, and generally manipulate the phrase. Because the arpeggio phrase is even, if we keep audio snap on, we should get glitch-free seams. If not, most sequencers provide crossfades on overlapping joints, or for gaps you can draw in a false volume fade (or automation) on each beat. Until the next time, we hope you have fun considering rhythm and texture and the role they can play in your music making. VI Peter Buick is a free-lance journalist, lecturer, and composer (specializing in interactive music for media) in London. BBC-trained with a strong history in studio engineering, these days he finds sound design and writing custom software a seminal part of music composition and production.
“….like nudity, often the suggestion is more appealing than the bare truth”
www.magesy.ru VIRTUAL
INSTRUMENTS
61
VI VSL (CONTINUED FROM PAGE 24)
their curve. It’s also possible to reverse their effect, for example if you want one parameter to go down as another goes up. So let’s say you have a sustained marcato in the middle slot. If you put progressively shorter accented articulations along the horizontal axis, you can use the Speed control—which once again senses how fast you’re playing— to switch automatically to the articulation that’s the appropriate length. Or it could switch from slow legato or portmento transitions to faster ones. The delay incurred to calculate the speed is unnoticeable. An even more astonishing use of the Speed control is when you go from performance legato to performance trill programs along the horizontal axis. At slower speeds you play legato notes in one bow or breath, and as you keep playing faster, the player switches seamlessly to a trill program. Sometimes it’s hard to believe that you’re actually playing what you hear—the transitions are that smooth. Nothing else can do that. The cover text of this issue asks whether the Vienna Instruments player can read your mind. That feature may be coming in Version 2, but this one does the next best thing. Performance The Vienna Instruments player is solid and reliable, but it is still an early release. While I still give the program the highest recommendation for Mac users, in general the Windows
release feels a little farther along. For one, I was able to run the Windows version with a 64 sample buffer vs. 256 on the Mac. Now, that’s a somewhat unfair comparison, since the Mac is running Apple Logic Pro while the Windows machine is just running Steinberg V-Stack (a basic instrument and plug-in host, not a full-featured DAW),
You could theoretically call up every articulation you need and play it live; the only practical limit is memory—the computer’s for loading programs, and your own for remembering which controllers or keyswitches bring up what programs. and the audio hardware is different. Nonetheless, the Windows version just seems friskier in general, and it’s less prone to clicking and popping when you start to strain the host computer. This may well just be a temporary Logic Pro/Vienna Instruments issue, though, and that doesn’t mean the performance in Logic or on Macs is sub-par. We’ll report any significant performance differences in MOTU Digital Performer 5, which is on its way here. As to memory access, I was able to load 1.745 GB of samples into the 2GB installed on the Windows machine, which is absolutely phenomenal—over half a gig more than anything else has ever loaded on that computer.
There it is: 5.13GB of samples accessed on a single machine! The VSL-Server shows 3.05GB, Logic Pro 2.08GB. The DAW isn’t straining either—you really can work with that kind of a load if you install 8GB in a G5. Note that the 2.08GB includes Logic itself and other sample libraries, in this case running in Native Instruments Kontakt 2 sampler; the 3GB is about the most VSL you can run on a single machine.
62
VIRTUAL
More importantly, it’s loading 29,784 samples; for perspective, a typical single articulation Patch might be 25MB and 400 samples. There’s a wide range, but that translates to maybe 75 patches. The only Windows issue is that if you attempt to load more than the program can handle, it brings down the house, requiring a
INSTRUMENTS
restart. It would be nice if it just cancelled whatever it was loading and threw up a dialog. The Mac Audio Units version works a little differently from other instrument plug-ins. It actually runs outside the DAW’s memory space, and it behaves halfway between a stand-alone program and a plug-in. With 5GB of RAM, the test Mac G5 was able to access 3.08GB of memory and 44,225 samples inside Logic (which includes the memory Logic itself uses). This is not the exact breaking point—that’s impossible to determine until you’ve crossed the line—but it’s very close to the precipice. When I filled the G5 all the way with 8GB of RAM, the VSL figure stayed the same, but then it was able to load an additional 1.9GB into Native Instruments Kontakt 2 inside Logic. It all ran fine! That’s 5GB of samples loaded inside a single machine, which is simply unheard of. But that was the breaking point—attempting to load even one more plug-in caused Logic to crash. The way OS X works is that each program is able to access a theoretical maximum of 4GB, which in the real world ends up being about 3GB after some overhead. Since Vienna Instruments runs outside the host program, it has its own additional memory space. It’s almost greedy to speculate why things break down at 5GB rather than 6 (3GB for each program), but my guess is that the total number of files open on the machine may have something to do with it. Having the Vienna Instruments be a separate program does have one drawback, though. When you go to open its window in Logic Pro, you get a dialog box to click on to open the Vienna Instruments window; this takes over the computer, since it’s running outside the DAW. If you save a screenset with the Vienna Instruments window open and then switch to a different screenset, the window is gone when you come back. While it’s a pretty minor inconvenience, this unique behavior is still suboptimal. With so much cell-switching to keep track of, it’s often necessary to work with the Vienna
www.magesy.ru
VI Instruments window open. Perhaps a ReWire version of the player would be a solution. In any case, 1.745 GB on a 2GB Windows machine and 3+2GB on an 8GB G5 is absolutely spectacular performance—especially when you take into account how efficient the player is with buffers.
memory access performance. That, combined with the incredible breadth of the library itself, allows you do to some things nothing else can do. VI
Conclusion All the major software samplers have improved by leaps and bounds in many important areas, but they haven’t necessarily focused their efforts on making the wonderful new libraries they spawned easy to manage. The Vienna Instruments player sets new standards in that category, and also in its
delay, and fade-in time. A voice sync switch locks together the LFO waves in all sounding voices. Neither the LFOs nor the delay times sync to the host clock.
GERBER
percussive and rhythmic tracks are performed. I will sequence it and produce variations by putting in lots of program changes and lots of envelope changes, and create variety. Sometimes I’ll play stuff in, but it’s not any faster I find.
(CONTINUED FROM PAGE 34)
I might write 30 seconds of music a day now. It’s 30 seconds that ten years from now I’m going to be proud of! But if you work six or eight hours a day and you’re a slow composer, you’re still going to produce a lot of music. What slows me down is not just composing, but sequencing. That basically adds about twice the time to everything you do.
AUDEON (CONTINUED FROM PAGE 53)
Look—up in the sky! One of the big advantages to software synthesizers is that they can all share the same
Because you have to switch articulations. Exactly. I find switching articulations slows me down so much that it takes away from the performance anyway, so I just sequence.
What is your working process—do you scribble and then play, do you sequence and then scribble… As of the last ten years it’s changed, because I used to be a pencil and paper composer—I’d just sit at the piano and compose. I’m a pianist, I still improvise every night at the piano just to get raw ideas and inspiration. But when I actually start composing, I just go to the sequencer in staff view and start inputting notes. The discipline involved there is to not jump ahead and start thinking about the orchestration…well, the orchestration always happens with the composition. But you have to first get a melodic idea down that you like before you start sculpting every note. When you’re sequencing you’re not just composing, you’re actually sound-sculpting every note, because every attack, every velocity, every note length has to be right, and you have to go in and make all those changes. If you start backwards and start thinking about the mix before you have the composition done…well, I’m not saying that’s wrong, just that it’ll produce different results.
Is the movement you’re working on now the first piece you’ve done using VSL? Yes, I just got VSL in February. It’s a better library than what I was using before. One of the things that makes that “organ” sound is the woodwinds. But the VSL woodwinds are the easiest woodwinds to write for, to get them to sound good—just the best I’ve heard. It’s not only the way they’re recorded, the crescendos, the decrescendos, but the other thing about that library, which I’ve never heard before, is the tuning. You play a chord and it doesn’t sound slightly out of tune.
So you live in the notation editor. I use the event list a lot and I use controllers a lot, but all my composing is done in the staff view. All my copying/pasting…
What do you do to shape phrases? If every note in a phrase has the same attack and velocity, it’s going to sound horrible. Well, it depends on the texture and the musical style, because sometimes you want repeated notes to have a hypnotic effect. But if you have repeated notes that are part of a primary melody, man you gotta do a lot to get them to sound good.
How much of what you do is performed and how much is programmed into the sequencer? Almost all of it is programmed. A lot of the
I hear an attitude in all their notes. I think if the players pour their souls into every note they’re sampling—which sounds kind of oxymoronic in the sense that how can you pour your soul into one note?—but you can! And there’s a big difference between playing a note and feeling it so strongly that it comes out with expression. If you catch that in the sample, you just have better samples.
hardware. There’s no need to limit your music by buying one instrument that purports to do it all. UFO is peculiar enough sonically—visceral, colorful, varied, and unpredictable, but seldom crisp or punchy—that I would never want to rely on it as my only synth. But it does things no other instrument will do, and that makes it a welcome addition to my evergrowing virtual rack. VI
Jim Aikin is the author of Power Tools for Synthesizer Programming and Chords and Harmony (Backbeat Books). His online avatar is www.musicwords.net.
With VSL you don’t have to use controller 11 and 7 to shape notes as much, because you’ve got so many samples with built-in crescendos and decrescendos. Do you use the Repetition Tool in VSL a lot? I’m running VSL on two computers. They are both running GigaStudio 3, they’re both at absolutely full capacity—I have no more room for any samples. But I have enough technique that I can accomplish the same thing a different way. And one of the ways I do that is with velocity envelope, note lengths, and location of the note on the beat. You can really get a smooth legato line, and you can get a repeating note line that doesn’t sound like a machine gun effect by sequencing it correctly without using those tools. I couldn’t load the Performance samples without taking out my single-note samples. It’s a compromise. If you can’t produce good music with four instruments, what makes you think you can produce good music with 25 or 30 or 100 instruments? When you have all this gear, the temptation is to make these thickly-orchestrated works that are actually over-orchestrated! A good composer can make a really interesting piece out of just one instrument. If you over-orchestrate music that’s not that complex, it sounds really creepy, really dated! A master who did this would be Stravinsky. If you look at scores of his music, they’re supercomplex. But the music is complex. One of the temptations for younger, inexperienced musicians is to over-orchestrate with this stuff just because they can. I’ve heard MIDI composers say you’re always limited by the medium and you have to write for the samples, and you change the phrase to fit the sample. To me the answer is that you’re always dealing with the limitations of the medium. If they’re not technical or artistic, they’re financial or social—life is fraught with limitations. You have to look at the positive and see all the options you have. VI
www.magesy.ru
VIRTUAL
INSTRUMENTS
63
VI t r e n d s Scripting: the craze over customizing Native Instruments Kontakt 2
ampler developers are allowing their instruments to respond to MIDI commands in all kinds of interesting ways. TASCAM GigaStudio 3 has iMIDI rules, which are stored with each program. It also has the Dynamic Expression Filter and Portmento Reshaping Filters its programmers discussed in our December/January issue. The latest craze is over the scripting feature in Native Instruments Kontakt 2. Scripts are programs you load up to instruct the sampler how to respond to MIDI. This applies to all the parameters of Kontakt 2’s extensive digital signal processing, including its convolution processor. While you and I can just load them up and use them, scripts are written in a programming language, and they can be very sophisticated. You’d expect the scripting feature to be one of those nerdy things nobody other than professional developers would pay any attention to. Not so—a vibrant community of people who write Kontakt scripts has popped up online. (VI writer Frederick Russ’ VIControl.net forum has a whole section dedicated to Kontakt scripting.) Most of the scripts around are free downloads, and they can add some rather amazing features to existing sample libraries.
S
To insert a script, simply load a program into Kontakt 2, then click on the wrench next to the program name to go into Instrument Edit mode. Click on the Script Editor tab.
64
VIRTUAL
INSTRUMENTS
The script that really fueled the craze is SIPS—the Solo Instrument Performance Suite (available here: http://www.theokrueger.com/kontakt2.htm, along with many other useful scripts). It was written by a talented multi-instrumentalist named Robert Villock (“Big Bob”), who unfortunately is having some health problems right now, and who we wish a speedy recovery. Bob analyzed the transitions between notes in various instruments, how the pitch shifts slightly, how they cross over, and in general what happens when you play legato. This unbelievably ambitious set of scripts, which actually comes with a 31-page manual, is his contribution. The SIPS suite has presets for many different instruments, since the transitions are always different. As a result, some samples that by today’s standards are pretty much unusable have been given new life; the difference can be astounding. Meanwhile, a fellow named Nils Lindberg in Sweden has written a series of scripts inspired by the Vienna Instruments player reviewed in this issue. See http://nilsliberg.se/ksp/scripts/scripts.html. One of the scripts is a Speed control that automatically switches between programs depending on how fast you play. Another script is by a fellow named Olivier Frappier (http://jazzphoton.free.fr/) who wrote a sympathetic resonance script for sampled pianos to recreate what happens when you play the sustain pedal. This makes use of Kontakt 2’s convolution processor. There are quite a few other scripts floating around the internet, and you should be able to find them from the links on this page. This is turning out to be an exciting advance in sampling technology, and what’s interesting is that it’s a grass roots phenomenon fueled by enthusiasts. VI
www.magesy.ru
www.magesy.ru
www.magesy.ru
