BY JIM AIKIN
There’s never been a better time to make music on a computer. All you need is a decently fast machine with a large hard drive, some non-cheesy audio I/O hardware, an Internet connection, and some patience. If you’ve got all that, you don’t need to spend a nickel on high-priced commercial programs — great music software is available for free.
Free software? Could it possibly be any good? Why would talented software developers just give their stuff away? Several reasons. Some of the most powerful music software in the world has been developed over the years by teams of composer/programmers working in universities. This type of software is maintained by volunteers and used by students, and nobody expects to make any money off of it. Other programs are “teasers” offered by commercial developers who hope you’ll like it so much you’ll buy the non-free programs.
Once in a while, a talented young developer will release a program as freeware in order to show off his programming chops and develop a buzz in the industry, a move that can lead to a full-time gig. Still other programs originate in the “open-source” movement. Computers that run the increasingly popular Linux operating system almost always use free, open-source software. As much as anything else, the open-source movement is a cultural jab in the eye at capitalism in general and monoliths like Microsoft in particular.
“Open source” means that if you’re a computer programmer, you can download the source code and modify it in whatever way you’d like. Most of us lack both the specialized skills and the need to do this — but even so, open-source software is a good thing, and not just because it’s free. The big advantage is that an open-source program isn’t tied to the fortunes of any single corporate entity. As users of Opcode Studio Vision (which wasn’t open source) learned the hard way, when a company closes its doors, years of creative work in the form of sequencer files can begin to quietly rot. Open-source software offers you some protection against that dire eventuality.
If your computer is a recent Macintosh, you already have a very nice piece of free music software: GarageBand, which is part of Apple’s iLife suite, is on your hard drive, ready to go. If you’re using Windows, you can download MuTools Mu.Lab Free (mutools.com), which is similar to GarageBand. I didn’t include this recorder in my Top Ten because the free version is limited to six tracks.
As I looked for cool stuff to include in this article, I sometimes had to choose between programs that are extremely powerful, but not easy to use and, on the other hand, programs whose friendly look and feel make them easy to use, but that are not very powerful, and perhaps not as interesting either. Three of my choices — Csound, Pd, and SuperCollider — line up solidly in the first column: amazingly sophisticated, but they’re brain-benders, and will require hours or days of study before you’ll even get your first sound out of them. The other items in this round-up are the kind that any computer-savvy musician should be able to start using quickly. You probably won’t have a use for all ten programs, but you’ll surely find something here that will ring your chimes.
With any complex piece of software, but especially with freeware, having a place to post questions will make it easier for you to get unstuck when you encounter the inevitable snag. Some programs have mailing lists. After subscribing to the list, you can post messages to it and get them answered (usually quickly) by experts. Other programs have user forums on their web sites.
The items listed in the following pages are not the end of the story by any means. I may have missed some extremely cool items. There’s a long list of freeware at making-music.blogspot.com/2007/10/all-about-free-music-making-software.html. Another important resource is KVR Audio (kvraudio.com). The KVR Developer Challenge (kvraudio.com/developer_challenge_2007.php) serves up a variety of intriguing free programs, most of them for Windows.
And now, in alphabetical order:
Ardour (For: Mac/Linux Site: ardour.org) A multitrack audio recorder, Ardour 2.7.1 doesn’t yet do MIDI tracks, but they’re slated for the 3.0 release. It requires an audio connection utility called Jack, which is also a free download. I haven’t used Ardour, but it looks very professional. The layout is reminiscent of Pro Tools. The owner’s manual can’t be downloaded, it’s online only; and as the development team freely admits, the manual is incomplete. But Ardour has been out on Linux for some years now, growing in power all the while, and now Mac owners can use it too. Definitely a contender for audio tracking.
Audacity (For: Win/Mac/Linux Site: audacity.sourceforge.net) Everybody needs to edit audio files, and that’s what Audacity is for. Not just cutting and pasting, either — this full-featured open-source program has all the EQ, normalizing, pitch shifting, noise removal, crossfade, and other utilities you’ll need. Plus, it isn’t just a stereo editor: You can assemble a multitrack montage, record stereo overdubs, and then save the mix as a stereo WAV, mp3, or Ogg Vorbis file. Unlike a full-featured multitrack recorder, Audacity doesn’t host plug-ins and has no mixer panel or automation. For that, you’ll need Ardour or a (non-free) Windows recorder.
Csound (For: Win/Mac/Linux Site: csounds.com)Csound is the graybeard of music software: Its roots go back to the 1970s. At that time, no computer could create or process audio in real time, so Csound is a text-based rendering language: You type a bunch of code and then hit the Render button to save your audio file to disk (or, these days, to listen to it immediately).
Why would anybody still use something so primitive? Easy: Csound is arguably the most powerful and best-sounding computer music system around. It does lots of types of synthesis and processing; it has a text-based scoring language that can run rings around a high-end MIDI sequencer; and it can do 64-bit floating-point audio, which is unimaginably clean. But it’s not easy to learn, and it doesn’t interface well with any other audio software in your computer.
Several “front ends” have been written to give Csound a boost into the modern era of windows and graphics. The best one I’ve found is called Blue. Written by Steven Yi and available for download from the Csound web site, Blue doesn’t eliminate the need to type code, but it has a multitrack timeline with mute and solo buttons and many other amenities. Csound and Blue are also compatible with general-purpose programming languages such as Python. If algorithmic composition is your thing, you could hardly ask for a better set of tools.
Gleetchlab (For: Macintosh Site: gleetchplug.com) A stand-alone program, deliberately eccentric, that’s designed for real-time experimental manipulation of samples. How eccentric is it? You can’t save or restore your settings because the author of the program wants you to relate to Gleetchlab somewhat the way you would have related to an old-time analog modular synthesizer, which had no patch memories. (Every time you fired it up, you had to plug in a bunch of patch cords.) Six long samples can be loaded into Gleetchlab at once, looped, and mangled using built-in effects with names like Cosmo (a spectral delay/reverb/granular processor), Neural, ThreshBass, and Rythmicon.
Illformed Glitch (For: Windows Site: illformed.org/plugins/glitch) If you never use drum loops, you may have no use for Glitch. But if you do use drum loops, but hate boring ones, you’ll love this program. Glitch is a VST 2.3 plug-in effect that can mangle loops in bizarre and unpredictable ways, turning a monotonous one-bar repeating riff into an endless kaleidoscope of killer weirdness. It has a bank of nine effects — TapeStop, Modulator, Retrigger, Shuffler, Reverser, Crusher, Gater, Delay, and Stretcher — each with four or five parameters plus a separate filter. When running, Glitch switches from one effect to another, either randomly or in an order that you specify. Switching can happen at any sixteenth-note. Sixteen separate patterns are a mouse-click away, and your knob moves can be automated.
I couldn’t get Glitch to work reliably in Steinberg Cubase 4.5. It operated well at first in Ableton Live 7, but then crashed Live the second time I tried to open its panel. My computer is not a squeaky-clean test machine, but yours probably isn’t either, so your results may vary, but I’d suggest testing Glitch before trying to use it at a gig. But even if it’s not completely stable, it’s too good to miss.
Mixxx Digital DJ (For: Win/Mac/Linux Site: mixxx.org)For DJs only, this open-source program can be controlled with a mouse or with any of a number of turntable control devices. The screen features a standard dual-deck layout, capable of loading and playing two songs at once and crossfading between them. You can send a separate mix to headphones, nudge a track forward or backward so its beat will line up with the other track, and so on. When installed, Mixxx asks where to find your song library, and then scans it, loading all the titles (plus length and bpm) into its own browser panel. A primitive flanger and equalizer are also provided. The manual is an online-only Wiki, so if you’re not wireless and get in a spot of technical trouble at a gig, you’re on your own.
Native Instruments Kore Player (For: Win/Mac Site: native-instruments.com) A teaser for the Kore multi-instrument and its expansion sound libraries, the free version of Kore Player comes with only 50 presets, but they sound terrific. A small amount of sound editing is provided via the eight macro knobs, which have been pre-assigned to useful parameters. Under the hood, Kore Player includes the sound engines for NI’s six main synthesizers (Reaktor, Absynth, FM8, and so on), so there are no sonic compromises. The download will take a while, as the installed sample library is over 300MB. Runs in VST/AU and stand-alone modes.
Pd (For: Win/Mac/Linux Site: puredata.info) A full-featured graphic programming language in which you connect modules by dragging “patch cords” with the mouse. Pd is mainly oriented toward real-time use, but as it’s a general-purpose program, you could create a multitrack recorder in it if you have the patience. Functionally, Pd is very similar to Cycling ’74 Max/MSP, which is not freeware. One important difference is that the graphic interface of Pd is primitive and ugly (as the screenshot shows), while Max is quite tasty to look at. Windows users who want to process MIDI data through Pd and send it on to their sequencer can do so using a free utility called MIDI Yoke (midiox.com). Pd isn’t just for MIDI, however: Its audio modules can be used to build your own synthesizers, samplers, and effects processors.
As with Max, the included tutorial files in Pd are actual running patches. You can experiment with them or block-copy them into your own patches, which makes the process of learning to use Pd considerably easier. But if you’re looking for instant gratification, Pd is probably not a good choice.
SuperCollider (For: Win/Mac/Linux Site: audiosynth.com) This high-end synthesis/processing system will intimidate the casual user, but SuperCollider has been around for a while, and it’s extremely powerful. It used to be Macintosh-specific, but it’s now fully cross-platform. (The Windows version is called PsyCollider.) Vaguely similar to Csound, in that you’ll be typing DSP code in the form of text, but the SuperCollider language is modern and object oriented. SuperCollider is reputed to be better than Csound at doing real-time processing, and its use of OSC (Open Sound Control) means you can run it on several computers over a network, so the potential processing power is greater. If there were more hours in a day, SuperCollider is a program I would definitely dig into.
u-he Triple Cheese (For: Win/Mac Site: u-he.com) Zebra2, the modular softsynth from u-he, is one of my favorite instruments, so I had high hopes as I downloaded Triple Cheese, in spite of its cheesy name. I wasn’t disappointed. Triple Cheese is clearly based on the same technology as Zebra2, but with a fixed set of modules. It uses an unusual type of synthesis: In place of dual oscillators and a lowpass or multimode filter, you get three comb filters, which seem to operate in series or parallel depending on what mode you choose for them. Two ADSR envelopes, two LFOs, and a single effect processor with various algorithms (phaser, reverb, and so on) round out the feature set.
The filters have knobs for tone and damping, and a choice of modes with names like Bowed, Blown, Damp, and Resonator. I was able to get quite a variety of thick, disturbing tones out of this setup. The time parameter of the delay effect seems always to be synced to the host’s quarter-note, though — not to the tempo, but to the actual quarter-note length. Now that’s cheesy. Or maybe it’s a just a bug. I forgot to mention that in the intro: Most free software has bugs. But then, come to think of it, most commercial software has bugs too. Whatever.