Turning music and sound into three-dimensional worlds often yields something that fields like a trip through space. But this feels like a real trip. Through pulsing, glowing starfields, “Versum”‘s audiovisual movements are brain-bendingly transformative. Artist Tarik Barri has created an integrated world of sound and image that makes the interface and the compositional realms seamless. It seems as though this really is a musical universe, through whose harmonies of the spheres you can fly like. Boldly going, indeed.

Ingredients: Max/MSP/Jitter, Processing, Java, SuperCollider, GLSL [the 3D shading language], and … some serious skill and time, I imagine.

The work has been in development for some years (not surprisingly, given the results). But it surfaced again as we brought up the 3Dconnexion SpaceNavigator hardware as a practical controller for 3D. See Create Digital Motion:
Look at Me, I’m Flying: SpaceNavigator Hardware + Blender

Tarik’s work resurfaced after a presentation in the UK. Reader janklug writes:

I’m just back from the M4_u Max/MSP/Jitter conference in Leicester (was great, btw), where Tarik Barri presented his project ‘Versum’, both as an installation and as a performance.
The user (and in case of the performance, Tarik) navigates through this incredible 3D-space-sequencer-universum with the help of a SpaceNavigator; glowing objects floating in this space produce sound, and as you approach them, they even give this nice doppler effect…
It was totally amazing to be able to float between pulsing rhythm-planet-objects and shiny drone-beams; navigation was easy and natural. Tarik uses a combination of Processing and Max/MSP; don’t know which one the SpaceNavigator is connected to.
Having tried this, I immediately ordered one; I think it also could be a great interface for M4L…

More information:
http://tarikbarri.nl/projects/versum
PDF documentation [2009]

Significantly, it’s really the act of flying that controls the music. That remains interactive, but it’s the movement through the three-dimensional space that determines what you hear. As the artist explains:

This virtual world is seen and heard from the viewpoint of a moving virtual camera with virtual microphones attached. This camera, controlled in realtime by means of a joystick (or any other kind of controller) moves through space, similar to how first person shooter games work. Within this space, I place objects that can be both seen and heard, and like in reality, the closer the camera is to them, the louder you hear them. So when the camera moves past several visual objects, you simultaneously hear several sounds fading in and out. Consequently, the way the camera travels past them actually causes melodies and compositional structures to be seen and heard.

The visual position of each object coincides with the panning of its sound: objects to the right of the camera will also be heard on the right, and those behind the camera will be heard from behind in case a surround speaker setup is used. This principle also applies to the Z-axis, meaning that sounds can be heard coming from above and below if the speaker setup supports it.

That’s the essential question, to me, when looking at 3D environments for music. What about the dimensionality will interact with the music? Is it something spatial, or will there be other sorts of interactions? (New Zealander-turned-Berliner Julian Oliver worked extensively with game engines, for instance. One solution for him was modifying the “gun” in those games to be an implement for doing things in the space, turning swords into plowshares after a fact by making the gun produce music rather than kill virtual entities.)

So, now you’ve seen some of the technical demonstration. But Tarik uses his work as an environment in which to make audiovisual performances. Here’s what some actual live playing looks like, in a beautiful, meditative piece called “Eleven”:

In fact, the biggest challenge to me of a piece this awesome is that you want an immersive environment, not just the small, rectangular screens that are often all festivals and venues can afford.

Holodeck, anyone?

More:

There’s something about symmetrical rhythms, it seems: through the power of math, these rhythms sound really good. We’ve looked a couple of times before at the spread of the Euclidean Algorithm for producing rhythms; see below. Wouter Hisschemöller has updated what began as an in-browser Flash tool to build an free and open source, Java-based MIDI utility. You dial in the rhythms you want, and now, with the addition of MIDI output, you can play those rhythms in any software of your choice. (Ableton Live plays the part of the MIDI recipient in the video above.)

Yes, you can actually make music with these nifty geometric interfaces:
Euclidean Patterns Demo 1 by Wouter Hisschemöller

Euclidean Patterns Demo 2 by Wouter Hisschemöller

Lots of detail and documentation on how to use the tool on Wouter’s updated blog post from earlier this week:
Euclidean MIDI Patterns

Previously, on Euclid Music Television:
Euclidean Rhythms in Ableton MIDI Clips for Polyrhythmic Good Times

Circles and Euclidean Rhythms: Off the Grid, a Few Music Makers That Go Round and Round

The beauty of modeling an instrument is that it involves ideas – taking a design from one context and translating it to another. With software, we’re able to put sound-making things everywhere, from obscure game consoles to a tab in your web browser that can distract you with music instead of Facebook updates. In the process of moving those ideas from place to place, we discover things.

Just ask Shannon Smith. He’s been on a great tear emulating favorite synthesizers in free toys for the browser. Through the power of the Internet, the New Zealand-born, California-based developer heard from Japan-based Monotron designer, who shared tips like these:

- filter doesn’t sound nearly as aggressive at maximum peak.
- LFO is slower at minimum (about 15s period) and faster at maximum (>1kHz).
- monotron resets the LFO at the moment the ribbon is touched, so it
works like a simple cycling EG at slow LFO rates.
- monotron has fixed intensity keytrack. cutoff tracks ribbon position
by factor of two. only tracks ribbon not pitch knob.

(We get to enjoy a much cooler industry that keeps friendly, and leaves the competition more often to the engineering departments than to the legal departments.)

And so there’s something wonderful about getting to fiddle with squelchy sounds in the tab of Chrome or Firefox. A few examples:
“Webotron” (Korg Monotron)
“Webotribe” (not-even-out-yet Monotribe)
Yamaha CS01
A 4-op FM synth
A (mini) Minimoog

There are useful tools, too, like a Java patch editor for the microKORG XL. Amazingly, it can actually transmit MIDI to the keyboard:
microKORG XL

Shannon tells us a bit more about the development process…

How they were developed:

All Java (interface and sound). It’s not really possible to use flash to generate sound real-time with low latency. Java also has built-in MIDI support that works in your browser which is pretty handy.

I was curious if things like this would be useful on tablets (particularly if someone got them working with HTML5 in place of Java – it’s a bit sobering that our “futuristic” Web tech represents a step backward in some respects):

I have considered writing apps for tablets and have been meaning to look into it but can never find the time. Also the market seems pretty saturated with much better products than I could produce in my spare time.

Shannon’s background:

I studied Electronic Engineering at a university in New Zealand and now work full time as a developer for a GPS company in California.

The most important lessons learned by doing these:

Hmm, that’s a tough one. I guess an appreciation for just how hard it is to digitally generate good sounds from scratch. I assumed before I started writing synthesizers that doing it digitally would be trivially easy compared to the analog days. In the digital realm you can do things with a few keystrokes that would have taken dozens of components and hours to wire up physically. Unfortunately even though it’s easy to get something working quickly there are some rather nasty artifacts that creep in when you do things digitally that means you have to be very careful anything you do doesn’t generate frequencies outside of the limited range dictated by your sampling rate. Anything you generate that falls outside this range folds back down into the audio spectrum and makes it sound crap.

I’m actually supposed to be writing games in my spare time. I only started writing synthesizers because I thought it would be a way to have decent sounding music in my games and keep the download size very small. I started out trying to do a Nord Lead emulation but utterly failed and realized just how complicated it was to get a good sound. Even though it was a failure it was a lot of fun to try and I continued writing them and lately I’ve been writing many more synths than games, also I tend to finish (mostly) the synths which is something I can’t seem to do with games.

Lots more goodies to explore:
http://www.angryoctopus.co.nz/

I think the first time I really understood handheld music making was when I first tried Nanoloop on Game Boy. While the more-popular LSDJ tracker is powerful, Nanoloop’s interface was unlike anything I’d seen before: aggressively minimal, it embodies in its interface design the feeling of a blank sheet of paper. Adding an idea feels like composition, like genuinely exploring open-ended possibilities and discovering what melodies may result. Now, Nanoloop – already on iOS – is available for Android, too.

It remains simple stuff, the sense of what a music maker looks like when designed for your hand rather than translated to mobile from desktop, studio-style workstations. There are six fixed channels, each assignable to a synth (FM, noise, or filtered wave) or a sampler. Then, each channel takes eight patterns. There’s a step sequencer, the ability to resample, and song editor with loops. The sampling capabilities are especially nice on Android, as you can now sample from the mic or load samples right onto the SD card. (The mic you can use on iOS, but not the SD card, of course.) You can export OGG files to your library, turning your Android device into a DJ-set-ready pocket music library, or send and receive projects via email. Via iTunes, you can even exchange files with the iPhone version, in case you have an Android phone and an iPod touch or some similar arrangement.

This isn’t a port of the Game Boy nanoloop, developer Oliver Wittchow is quick to note. It’s not for chip music, or emulating game consoles. It’s for … well, Android music. But make no mistake: while the tool feels fresh and native to the new platform, it also carries the spirit of the Game Boy version. And that spirit is handheld music making, not just the aesthetics of the chip, but the feeling of using a minimal device scaled to your hands, something you can use on the go.

Oliver tells CDM that he has used a lot of native code (via the NDK), as he did non-object-oriented C (not Objective-C) in the iOS version. He says it’s about 1:1 C and Java: “I could use the iOS code almost unmodified and get the sound engine and touch input to work immediately. I had an almost fully working nanoloop within few days – without any graphics though.
Now I ‘just’ had to write Java code for GUI, file access, recording functions and the different menu structure.”

The results, he says, are a little strange if you’re looking at the code (lots of getters and setters, C mixed with Java), but it works well.

Android owners, I’d love to hear how this works on your device. Let us know – just be sure to fill out a bug report. As the Market page notes:

Nanoloop works on a variety of devices, including HTC Desire, Motorola Droid and also midrange phones such as HTC Legend and XPERIA X10 mini. However, this is the first release and of course it has not been tested on all Android devices yet. If you experience crashes or other problems, please report them via e-mail or the anonymous bug report form at www.nanoloop.com. Thank you.

Now if we can just have some luck with Bluetooth MIDI at our hacklab tomorrow, this could be another candidate.

Oh, and most of all, I’d love to hear the music you make.

nanoloop on the Android Market (a stunning EUR1 … that’s a no-brainer. Better than an espresso shot!)

nanoloop.de [iOS, Android]

See our previous Android music round-up

For the historical record, here’s nanoloop for Game Boy on CDM – from 2004. It’s actually come quite a ways since that release, but the spirit is the same.
http://createdigitalmusic.com/2004/11/updated-nanoloop-20-game-boy-instrument/

In your pocket, or perhaps orphaned in your closet, is a sophisticated piece of electronics going to waste. So, whether you’re suffering from iPhone envy or simply want to put toxic used electronics to useful musical applications, the cell phone noisemaking project at GetLoFi could help make a happier, noisier world.

The ingredients:

• Software running on the Java virtual machine (good, old-fashioned J2ME)
• A hack for proper input and output (from the various headset connectors)
• Sequenced MIDI files

The result: otherwise silent, ordinary cellphones become living musical creatures. Install files, plus information on how to make the circuit for I/O, went up on GetLoFi early in October, but thanks to Michael Una for pointing this my way over dinner recently.

Everything you need:
MIDI Loop Sequencer aka Cell Phone Noisemaker

What if you could make any device or any software a re-programmable musical instrument, effect, or soundmaker? Your phone could be a touch-controlled effect, your tablet a sketchpad for interactive drum sequencers. Patches assembled on your desk on a computer could be taken with you in your pocket. And what if you could do all of this for free, using a time-tested environment?

libpd, authored by Peter Brinkmann, takes on that vision. It’s a way of making Pure Data (Pd), the visual development tool for interactive music and media, more accessible across a range of applications and gadgets. It lets you embed Pd pretty much anywhere. It’s not a new version of Pd. Instead, it makes use of the standard, “vanilla” distribution of the free and open source software. What’s different is that it separates the sound processing part of Pd from the part that talks to audio hardware, allowing Pd to run on a greater variety of mobile devices and inside other applications.

libpd:

• turns Pd into an audio synthesis and processing library
• liberates Pd from GUI and drivers
• allows for easy communication between Pd and the code into which it is embedded (so you can send and receive messages with your Pd patch)

Today, a team of developers and testers (including myself) is releasing the first version of libpd. It’s free to use on any device you wish, and free to modify. Because of its licensing, you can even build commercial applications with it. (That is, yes, it’s open source – but yes, it can also be useful if you’re a commercial developer. You don’t have to choose.)

http://gitorious.org/pdlib | community discussion

We’re just pleased to have a tool that makes experimenting with sound and music quicker, easier, and more flexible and compatible. It’s more fun that way.

Supported Platforms, What You Can Make

Right now, today, you can use libpd with:

Android: Thanks to Google’s NDK (Native Development Kit), you can use libpd with any Android device running OS 1.6 or later. Note that devices without the Google Market are often non-standard in other respects, so your mileage may vary, but we’ve found a wide variety of devices work quite well, including the Motorola Droid and Droid X, HTC Legend, and Google NexusOne.

iOS: iPhone and iPad models with the latest, armv7 processors work (3GS, iPad); we’re working to extend compatibility across more devices. Working with Peter Brinkmann, the RjDj development team contributed (and continues to contribute) free code that’s making iOS support compatible and high-performance. But the Objective-C classes mirror the Android and Java classes, meaning the two will stay in sync, and once you’ve learned one, the other will be a piece of cake. (Or coffee. Or cocoa. Or whatever.)

In each case, you just need libpd, Pd for making your patches (graphically), and a copy of the SDK for each mobile platform you want to use.

Additionally, you will soon be able make user interfaces for libpd using cross-platform HTML5, via Chris McCormick’s project WebKitPd. (It’s not quite ready for consumption yet, but will also be free and open source.) Android was the impetus and initial test platform for libpd, so right now it’s the most mature. But we hope to improve iOS compatibility and testing next.

Some Sample Apps to Try for Android

libpd is really aimed at developers who want to embed Pure Data into mobile devices, games, and so on, and soon also people working with Processing, Open Frameworks, and the like.

But if you’re eager to try this out as an end user, there are a number of packages you can try. They don’t show off everything libpd and Pd can do, but they do allow you to load up something on your device and make some noise.

Download the test packages from the libpd site:
http://gitorious.org/pdlib/pages/Packages

It includes a scene player for RjDj (see below).

Among the code included in the repository is one complete app, Peter Brinkmann’s own Circle of Fifths. He tells us:

Circle of Fifths: I wanted a circle of fifths tool for the subway, with exactly this kind of GUI. It also nicely illustrates the newly possible separation of concerns — Pd only does DSP, and an elaborate GUI is built somewhere else. It’s a demo and not optimized for universal consumption. In particular, it’s a bit CPU hungry because it’s actually simulating six Karplus-Strong strings in real time.

Of course, if string simulation is what you want to do, this also illustrates that you can – even on a phone.

Chris McCormick has created two libpd-based apps, one of which I feature in the video above. Can of Beats also makes use of WebKit as its UI rendering engine. Chris describes what he’s made:

Can of Beats: This is a procedural hiphop beat generator. The beats are generated using simple hand-crafted probability weightings for each type of sound at each position in the beat. In the Android app, you can also input simple melodies or basslines to go along with the beats.
http://mccormick.cx/projects/CanOfBeats/

Garage Acid Lab: This is an algorithmic, 303-style acid bassline generator. The app will make you an infinite number of different acid bass lines and garage style beats. You can also have some fun with the cutoff filter and delay unit settings with a kaos-pad style input. I want to work on this app a bit more to provide an ‘advanced’ mode which will let you write custom basslines, beats, and have more control over the effects. http://mccormick.cx/projects/GarageAcidLab/

RjDj, RjDj team. If you don’t know it by now, RjDj is a fantastic application built on Pd that makes interactive musical and sonic experiences deliverable in the same way as a digital album, not only to musicians, but anyone who wants to experience music and sound in new ways. libpd makes use of code contributed by RjDj. Future development on RjDj will use libpd. (More on those libpd-based versions, and the evolution of RjDj and RjDj Voyager, soon.)
http://www.rjdj.me/

Where to Get It, Where to Get Involved

1. Get the library. To get started, download libpd from its Gitorious source repository:
http://gitorious.org/pdlib

You’ll need Pd, too, if you don’t have it; vanilla Pd builds are available from the official Pd download page.

2. Join the community. You can discuss patching for libpd, developing using Pd, and making instruments and effects and other sonic creations for gadgets everywhere on our new community group:
http://noisepages.com/groups/pd-everywhere/

That will be a location specifically dedicated to the unique challenges of working with mobile gadgets; of course, see also the other great community resources for Pd.

Who made this: libpd was conceived by a team including Peter Brinkmann, Hans-Christoph Steiner, and myself, with input from the RjDj team (particulary Martin Roth). It was primarily developed by Peter Brinkmann, who applied his talents and the work he has done in JJACK, a Java API for JACK, with additional contributions and testing by our team and by Chris McCormick. Major thanks to Martin Roth and the folks at RjDj, to Miller Puckette (creator of Pd), and the generous attendees of our first hackday at the NYC Patching Circle, along with others who are testing now.

Tutorial Next Week; Your Feedback Wanted

I’ve been preparing a tutorial for working with libpd Android (initially), to be followed eventually for developing on iOS devices once we have a better handle on making that go smoothly. We’ll have a complete tutorial for you by next week. Processing is then my next priority.

An FAQ will also be available by then. That means, first, ask some questions!

Got specific questions about what this is for? How to get started? What you’d like to see in the tutorial? Ask away.

And please do get the discussion going not only here in comments, but in the Pd Everywhere group. (Noisepages registration is now open; if you have any trouble, let me know and I’ll sort you out.)

Those crazy Amiga artists were ahead of their time. The lightweight real-time music engines and formats they began were uncommonly efficient, and allowed the exchange of elaborate electronic music using a minimum of resources – with some accompanying compositional and sound design ingenuity required, as well. As a result, getting a phone handset to reproduce their work today is a pretty manageable task, and some of the music available is concise and clever. Pop on some headphones, load up some tunes, and you may feel you’re starring in your very own Amiga point and click adventure the next time you hit the grocery market.

There are a number of trackers and mod players for mobile platforms from iPhone to Windows Mobile, but Android is now in on the game thanks to Xmp (Extended Module Player). Using Android’s JNI-based NDK interface for accessing native code from Java, the “experimental” queue up some files and play back on your SD card. My sense is that this hasn’t been widely tested, which is where you come in: got an Android phone? Ideally, got some obscure models of Android phone? Load this up and see if you’re getting the retro tracker music love. Let us know in comments how it goes.

Full downloads and code for Xmp, a command-line mod player for Mac, Windows, Linux and pretty much every OS every invented, along with the experimental Android port:
http://xmp.sourceforge.net/

Thanks to Dan Galpin, developer advocate at Google, for pointing this out to me.

Now, this brings me to my open question. Suffice to say, someone could build a pretty player interface for Xmp, with playlist support and the lot. But what about actually editing files on your Android device, as you can on iPhone, PSP, GamePark, PC, Mac, etc.? It’s possible that the Xmp code could be used as a template for porting the engine of something like LittleGPTracker. But looking through quickly, I wonder if Xmp itself might serve as a real-time engine? It’d also be interesting to design a tracker interface that took the UI patterns of platforms like Android to heart, rather than just reproducing interfaces designed for other platforms. If you’re interested in such a project or have some insight into what might be practical, let us know in comments.

Swarm Music Album Black Allegheny from Evan Merz on Vimeo.

We’ve heard albums made by singular compositional minds and by bands. What would an album sound like if composed by swarm intelligence, by computer evolutionary models of individual agents or bots? That’s the question asked by composer Evan Merz in his new, full-length album “Black Allegheny.” (At top: the composer explains in a video.)

Western musical and creative tradition is steeped in linearity, from the forward motion of the music staff to the mythos of Aristotle’s Poetics.

So, maybe it’s little wonder that generative music – music that may not have linearity, or a beginning, middle, and end – hasn’t exactly been a big hit with the kids. Pioneers like Brian Eno have helped spread the gospel of generative music, but apart from lots of interesting experiments, there hasn’t been a lot of actual musical content. If you were to make a stack of generative music albums, your listening list would be fairly short.

All of that could be about to change. Programming code, the essential medium in which such models can be developed, is more accessible than ever. It’s also more visual, thanks to the popularization of tools like Processing, which can help make the abstract rules of generative music easier to grok. Merz, for his part, has taken on the challenge with his own Java-based software.

Saying the bots “compose” the music may be a little misleading. Generative music needs rules to operate. Before Eno, there was John Cage, whose “chance” compositions were as much defined by choices of materials as by ranges of indeterminacy. Merz makes a nod to Cage’s notion of a “gamut,” a collection of raw musical elements used as the input in the chance system. Here, though, Merz is aided by something Cage didn’t have: a swarm of intelligent “agents” can navigate those materials via simple rules, giving the music form and substance. Because they aren’t aware of the big picture, the music evolves more naturally, rather than being subjected to an over-arching narrative.

Or, as Merz puts it, “the tiny ant on the ground knows only what it sees around it.”

So, that’s the theory — what does the music sound like? Far from “ennui,” as Merz puts it, to me the results are organic. The structure is emergent from its materials, sounding almost like a natural physical process, like watching ice melt. The content ranges based on the gamut; like a lot of generative music, some sounds a whole lot like Brian Eno’s work. Others borrow from minimalist composers (Reich’s music itself might be seen as partially generative), and others take on an edgy urgency. The models that determine the bots are based on a popular, simple mathematical predator/food model, one often used in these works. Sometimes, you might imagine that evolutionary struggle playing out in the music.

You can read more about the process of developing this tool and the compositional ideas behind it at Evan’s blog:
Black Allegheny, Swarm Generated Music [Computer Music Blog]

For more explorations of sound and composition, check out Noise for Airports, which recently featured the work:
http://noiseforairports.com/

And you can stream the album or buy it for yourself for the light price of US$5 — though I’d like to see a software release, since that would mean each playback could be different. (Eno released an album in software form in the 90s, though tracking down the software now is evidently impossible – anyone with tips?)
Black Allegheny @ Bandcamp [Stream / download purchase]

Swarm Controlled Sampler – Becoming Live from Evan Merz on Vimeo.

Quick: you need to produce a music score. It needs to look really great. The deadline is looming. You’ve lost your serial number for [insert program here]. What do you do? The answer might surprise you.

Lilypond is something of a cult secret in music notation circles. It’s free software for high-quality computer engraving, it runs on any platform (Mac, Windows, Linux), and it produces exceptionally good-looking output, often exceeding leading commercial programs in particularly tricky notational situations. But it could easily scare off beginners, because it isn’t necessarily graphical software. The tool generates its output from text files, a bit like the way in which a Web page is rendered from an HTML file.

What beginners don’t know is that text entry doesn’t have to be slower or more daunting – especially if you choose a tool that assists you in the entry process.

Lilypond’s language for basic music entry is actually reasonably simple. If you want a g flat, for instance, you just type “gf.” (Note: you will probably need to adjust Lilypond for your native language to get an abbreviation for “flat” that makes sense to you! Hint: “flat” is in English.) To change rhythmic durations, you use a number, so two eights followed by two quarter notes would look like “c8 d e4 f.” Because it’s text-based, you can be explicit about what you want, which avoids some of the pitfalls of graphical entry methods. If text is to be attached to a specific note, you specify which note in your text file. Most importantly, this means that entering and arranging notation doesn’t get any harder as the score becomes more involved. For complex measures with densely-packed material, or tricky notations from early music to modern composition, Lilypond continues to handle layout and rendering automatically, without intervention – just at the point many graphical programs will have you pulling out your hair.

Lilypond “Switch” How-to Crash Course

Entry itself can therefore move really fast, especially if you like to sketch out an idea on paper (or in a MIDI file) first. I recently completed my first score in Lilypond, and was surprised that – after the initial hour or two of entry – I started to really like it. Getting the first few bars in was a bit tricky as I got the hang of entry, but then, to my surprise, finishing the score went as fast or faster than it had in other programs.

Find the Right Tools

That isn’t to say you won’t want some help. Music notation is always involved, because of the sheer quantity of notational conventions used – even for fairly simple musical contexts. And while text entry makes copying fast, you’re likely to want some MIDI playback and entry assistance. In fact, I’d wager the quality of your experience with Lilypond will depend on choosing a front-end tool you like.

I experimented with various tools on my Ubuntu install, including some graphical programs that can read and write Lilypad files. If anyone is interested, there are a number of programs I can recommend you don’t use. In the end, I found that what I wanted was essentially a text editor – so I could take advantage of the speed of Lilypond’s text-based language – but with plenty of shortcuts so that I’d never get lost trying to look up how to input a symbol.

Frescobaldi was a real pleasure to use, if you have a Linux install. (That’s the case for now; efforts to port KDE and Python should mean Mac and Windows versions aren’t far off.) It’ll install a lot of dependencies on a stock Ubuntu install because it relies on KDE, but it’s a nice all-in-one tool. A PDF preview accompanies your text so you can see what you’re doing, and by clicking on a note, you jump to the correct place in the text. There’s instant access to online help and notational references. The nicest feature is perhaps the MIDI input using Rumor, which worked out of the box with an M-Audio USB MIDI keyboard I connected.

Frescobaldi

The other best-of-class tool I found is none other than omni-platform text editor jEdit, with the LilyPondTool add-on (Mac, Windows, Linux). (Thanks to a Twitter friend for the tip; thanks to Twitter’s terrible archiving, I’ve lost who you are, so say hi in comments?)

Grab jEdit, and perform two steps:

1. De-uglify jEdit. Yes, that default skin is hideous, and doesn’t look like any OS you’ve seen in the past ten years. Choose Utilities > Global Options > Appearance > Swing look & feel, and set it to something native for your OS. Reboot, and take a deep breath.

2. Install the plug-in. Incredibly, it’s a default option. Choose Plugins > Plugin Manager > Install > LilyPondTool.

jEdit’s LilyPondTool does a lot of what Frescobaldi does, with wizards for setting up scores and changing parameters and various clickable shortcuts. But it benefits from putting this functionality inside an extensible, standard text editor, which means you can do anything with LilyPondTool that you can with jEdit. And there are simply more options – there are more quick menu shortcuts for symbols, tweaks, and all the other little things you have to do in notation that you don’t realize you have to do until you get halfway through a score. That makes LilyPondTool a bit friendlier to beginners. It doesn’t have MIDI input as Frescobaldi does, but it does have MIDI playback. You even get nice tools for making templates and OpenOffice-based hyphenation of lyrics, plus a virtual on-screen keyboard to aid with entry.

http://www.jedit.org/
LilyPondTool

One part of the process I didn’t quite work out was the best MIDI import tool. There’s a simple Python script that ships with the Lilypond distribution, and it can be called from tools like jEdit+LilyPondTool. But converting MIDI to notation isn’t a simple task in any tool, so I’d have to research this further. Doing note entry in a proper MIDI sequencer, then adjusting the engraving in a Lilypond editor like jEdit or Frescobaldi seems a terrific workflow, though, if anyone has found a process that works for them.

What you might miss…

So, how does a free tool like Lilypond stack up against the newest version of, say, Sibelius?

Even if you’re a seasoned Sibelius user, I highly recommend doing at least one score in Lilypond, as it’ll give you a window into how Sibelius works, and the issues that arise in computer notation. I believe Lilypond may have even helped influence modern versions of Sibelius with its approach to engraving, though that’s only from memory – don’t quote me on that.

You will see some advantages of Sibelius. Sibelius’ graphical layout means there’s no separation between what you see and what you edit, and one big edge of the Sibelius engine from the beginning has been its ability to reflow even huge scores almost instantly. That visual process could become part of your compositional workflow, too, with on-screen clipboards for ideas and quick playback of ideas. Sibelius has also done a lot recently with DAW-style tools for MIDI, live tempo tapping, integration with ReWire, and so on. That makes Sibelius a powerful tool for creating high-quality playback right in the score.

Nice as Lilypond is, I certainly have an easier time imagining teaching students notation with Sibelius than with a text editor.

Advantages of the Lilypond approach

I commonly hear odd, defensive barbs about free software, especially in the music community. People will casually drop statements like, “but the open source community doesn’t innovate. They just rip off commercial software. And it’s just not as good.” As near as I can figure, this entire argument is often based on one or two bad experiences with OpenOffice a few years ago.

Now, some of this defensiveness comes from the fair perception that discussion of free software often centers more on philosophy than practicality. And there, I agree. Software is a tool. Philosophy matters, but you ought to be able to look at tools in more or less objective terms. You ought to be able to like using the tool.

Lilypond is a perfect counter-example. It is innovative software, period; now well over a decade old, it’s well-respected in the engraving community. I’ve been surprised to find out how many people use it, and they do so because it saves them time and headaches and they like the output.

It’s also just plain different from the commercial offerings. Its free nature means it can do things that commercial software doesn’t even try to do. (Can you imagine a major vendor unveiling a text-only notation app? Didn’t think so.)

As with any design, this means some trade-offs. They aren’t as simple as “Sibelius and Finale are for casual users; Lilypond is for hard-core geeks.” On the contrary, I found some real advantages.

Text input means backup, file exchange, and tracking revisions becomes a whole lot easier. Lilypond’s output is more like traditional engraved scores than anything I’ve seen from Sibelius or Finale, even when swapping fonts in those packages. Lilypond is uniquely equipped for doing early music notation; it makes a lot of alternative notations as easy as modern notation. Sure, that sounds like an “advanced” feature, but it’s an “entry level” feature if you happen to perform or research early music.

Also, despite improvements to things like Sibelius’ “magnetic layout” and other automated features, I find that even the newest versions of these apps still require a lot of tweaking after the fact. Lilypond still requires “subjective” tweaks – adding a page break where you want one, for instance – but I tried some tests with bars of music that broke my favorite commercial tools, and Lilypond was very hard to stump. There’s also the simple fact of the matter that with graphical tools, it’s easier to screw up the notation yourself, by attaching text to the wrong note or dragging something slightly out of place. Those changes are hidden in the graphical view, too, whereas they’re explicit in the textual score.

I don’t think one approach is necessarily better than the other. The point is, you need both. Something like Sibelius or Finale is just not going to evolve in free software. But something like Lilypond isn’t going to evolve in commercial software.

You also don’t have to choose one or the other. Thanks to MusicXML, an interchange format, you can exchange files between Sibelius or Finale and Lilypond easily. If you work a lot with scores, it’s worth a download – and the price is right.

And I don’t believe Lilypond is “for geeks only.” Give yourself a simple job, like a lead sheet, and pick a solid tool. Give it a real try, with a couple of evenings to get used to the language. I think whether you like the results will have more to do with personal preference. But I’m glad Lilypond exists, and I think you may find it’s something you want to add to your arsenal, even if that arsenal also includes one of these other tools.

For good measure, here’s a visualization of the open source contributions to the project.

code_swarm: LilyPond apr 23, 2010 v2 from Paco Vila on Vimeo.

Keep up with news from the project, and some good tips, at:
http://news.lilynet.net/

Share your experience

If you want to give this a try, I recommend both the jEdit and (for Linux) Frescobaldi routes. Each has links to Lilypond tutorials and documentation right in the program. I’ll try to work out a quick tutorial at some point, too; I’m planning a bigger set of scores and am going to give Lilypond the old college try.

Worked with Lilypond? Found a tutorial that helped you out? Got some tips? Trying it out and need help? Do share.

By the way, that score I worked on will be premiering as part of a party with operatic and musical theater types Monday in New York. Alongside digital music made by computers, it’s nice to get to work with humans, too, which is why I suspect notation will be with us for a long time to come.

New music party, NYC, Monday night 5/17

For computers, digital tech means the ability to turn anything into numbers. For humans, it means a chance to translate between gestures, ideas, sounds, and images. We can interface with musical, visual creations intuitively and collaboratively – now with ubiquitous, cheap touch and electronics. Two events take on that idea on the two coasts of the US; if you’re nearby, hopefully you can drop by, and if not, we’ll have plenty to share.

Multi-user Art, Networked OSC Workshops in NYC

Here in New York, mobile touch is put to the test in a gallery show in Brooklyn, with two workshops that can help you make your own work. Multi-User Art (image, top) uses the open platform mrmr and OSC protocol to allow visitors with mobile devices to manipulate installations. Step up, and a layout of controls is automatically pushed to your device, so you can push buttons, slide faders, draw, and otherwise control what you see — even with multiple users at a time. (For now, we’re stuck with the iPhone/iPod touch/iPad only, but I’m working on Android and browser-based ports for anyone interested.)

The artwork includes installations by myself, by mrmr creator Eric Redlinger, Superdraw artist Joshue Ott, and, using mirrors and light in place of projection, Chris Jordan. They range from three-dimensional, collaborative drawing to reflected light to moonscapes. The opening is free on Friday night:

But what if you want to learn to harness some of these same tools in your own work? We have two workshops Saturday, too.

Friday, 4/9: Opening, free – 7p; see exhibition information

Saturday, 4/10
11a-12p, free, Eric Redlinger presents an introduction to mrmr, a demonstration of how to use an iPod touch, iPhone, or iPad for control via OSC.

1p-4p, Joshue Ott and I will teach an in-depth workshop on using OSC for networked communication with free tools, focusing on Processing and Pure Data. We’ll talk about how Zeroconf (implemented by Apple as Bonjour) can create zero-configuration, automatic connections – no IP numbers to type. We’ll look at how you can use OSC to make software and hardware communicate across networks, for sound and visuals. And we’ll talk about how you can use tools like mrmr on mobile devices. $75. Class information, signup required!

All information:
Devotion Gallery, Brooklyn

Physical Controls with Arduino, Max, Pd in LA

Touch controls a bit too insubstantial for you? Prefer the tactile feel of a physical encoder in your hand? We’ve got you covered there, too.

Kevin Nelson writes us to share some new events he’s setting up at LA’s new CrashSpace hackerspace. (About time LA got a new, proper hack spot!)

There’s a free workshop Wednesday, using Max (though translating to other environments like Pd shouldn’t be hard), plus a more advanced intensive in May. Details:

First, I’m going to be giving a high-level talk this Wednesday, April 7 at 8pm on using a Monome to control things in the real world by integrating an Arduino with Max/MSP. The talk is free for members of the space, with a $10 suggested donation for non-members.

Flamethrowers! Arduinos! Monomes! This Wednesday, April 7th [CrashSpace]

Second, I’ll be teaching a course on building user interfaces with the Arduino and dataflow languages (I’m trying to make the emphasis on pd because it’s open source, but depending on the audience, Max and Max for Live might slip in there too). The curriculum and description haven’t been posted yet, but should be done and up by the end of the week.

The basic idea of the course is to target musicians who have dabbled in electronics and give them the tools necessary to empower themselves to build their own interfaces and instruments. It’s a two day intensive (8+ hrs/day) on May 15 and 16. We’ll be covering basic electronics & sensors, Arduino programming, serial communication between Arduino and pd/Max, and basic pd patching for midi routing or sound generation. The course is $150 for CrashSpace members, and $250 for non-members, and both prices include an Arduino and selection of sensors & misc components for the project.

http://csarduinodataflow.eventbrite.com/

What’s the Best Way to Document?

What do you prefer for documentation of these courses, for those of you not in NY or LA who want to follow along at home? (And hey, I can’t be accused of being too specialized geography-wise — I’m teaching a similar course in Portugal this spring.)

I’ve (ahem) sometimes promised more documentation than I’ve actually delivered, but in the meantime, I have been gradually refining some examples in Processing, Pd, and the like, so I’m feeling less shy about sharing them.

Suggestions? The more specific, the more likely I am to implement them. What do you want to see? In what format? Any sites you’ve found useful for this sort of sharing?

The more we can share this sort of specific knowledge, the more we as a community can help each other build our skills.