Archive for the ‘software-hardware’ Category

Interview with SounDemon, the Sound Chip Hacker

July 6, 2009

In November last year I wrote a post about playing music with the graphic (VIC) chip of the C64, aswell as combining 4 channel Amiga MODs with 3 SID-channels. I e-mailed some questions to one of the programmers behind it and I was happy to get a reply from him the other day. : ) SounDemon is what I would call a sound chip hacker, since many of his works are based on exploring undocumented features of the SID-chip and exploit them. These things do not rely on CPU-power to create new sounds, that most music software does. In my opinion this is one of the most hardcore ways of making chip music that is somewhere inbetween hardware and software. For hardcore hardware chip music, I would recommend you to go to Brisbane, Australia right now for EPROM-music. But anyway:

CHIPFLIP > So first just a bit about what you are doing and what you have done in general. Education and stuff.

SOUNDEMON > I’m studying computer science at Abo Akademi in Turku/Finland. At the moment I seem to spend all my time running to choir practices and doing math exercises for school.

CHIPFLIP > How did you get into the SID chip?

SOUNDEMON > I think the first music routine I wrote was for the Dekadence 4kb demo Perkele. BriteLite asked if I could do a tune that was very small, in order to leave room for as many demo effects as possible. The obvious solution to this was to code a custom player. So, I got into writing music routines and by experimenting I somehow managed to invent a few new sound routines.

I must add that I have always liked the idea of programming music. This is the only way to gain full control over the sound. I was inspired by old C64 composers (Galway for example) who had to work this way, before fancy editors were available.

CHIPFLIP > Tell us a bit about your different projects. How did you come up with ‘the new waveform’ in Pico? How did you do it and what does it actually do? Will there be new experiments with the waveform editor?

SOUNDEMON > As with Perkele, we needed a very small tune for Pico (which is also a 4kb demo). I decided to include some metallic drum sounds by using the “testbit trick”. While trying different parameters for the sounds I got some weird pitched sounds. Only after releasing the demo I spent some time analyzing the behaviour of the SID chip to find out how and why the trick works.

The routine works by directing a steady stream of angry bits towards the noise generator of the SID. The result is a confused SID chip playing sounds it’s not supposed to play. For a more technical description see:

It might be possible to create more sophisticated sounds with this method than has been done so far… (hint hint)

CHIPFLIP > I once heard something about a 2 tone filter (“new waveform”) for the Atari Pokey, but can’t seem to find the information back right now. But have you heard about this?

SOUNDEMON > I’m not sure what this is. I believe most 8 bit sound chips (including the SID) use a shift register based approach for generating noise. This explains why it might be possible to get the same kind of sounds on other machines as well.

CHIPFLIP > Could u tell us a bit about your sample shocks from x2008? How is it possible to play 4 channels of 8-bit samples? And ofcourse, how about the Vic audio?

SOUNDEMON > I must first clarify one thing. In our x2008 demo there’s two new
major routines: A “MOD” player capable of mixing four digi channels AND the 8 bit sample playback routine. These are NOT the same routine, but they can of course be combined as we did.

The MOD player was written by The Human Code Machine. MOD players have been written for C64 before. The one by THCM is special because it actually sounds good and allows the screen to be turned on. (How fun is it to have MODs playing if you can’t display anything on the screen?) It’s based on straightforward code that uses cleverly precalculated tables to do the hard work. Somehow THCM managed to fit these tables and a MOD into 64kb of memory. I still suspect he cheated by hiding a memory expansion
unit inside my C64! (I haven’t found it yet)

The 8 bit sample player was written by me. 8 (and even 12 bit) sample playback has been done on C64 before, but this is the first routine that sounds clear and doesn’t use all raster time.

The VIC audio is just a fun trick. It’s absolutely nothing special codewise. It’s a bit like the 9 sprites on the same raster line trick by xbow where the idea is the achievement, not the actual code. That is why I gave the credit for inventing this technique to AMJ. He came up with the idea and after that the code was done in about 10 minutes.

CHIPFLIP > Do you always use your own software when you make C64 music?

SOUNDEMON > I don’t even have my own software. When using my own routines I just use Turbo Assembler to edit the player source and music data. I seldom reuse a player because they are typically coded for a specific tune. This is of course very time consuming so I do it only when it’s necessary. Usually because of tight size or raster time constraints.

CHIPFLIP > Are there other soundchip hackers that you know of?

SOUNDEMON > What exactly is a sound chip hacker? I like the sound of it, though…

CHIPFLIP > What will be your next shock? :)

SOUNDEMON > I will continue coding on network routines for C64… Something
interesting might result.

CHIPFLIP > Could you give us a few examples of 8-bit code, music, and graphics that you think are special?

SOUNDEMON > I liked Royal Arte by Booze Design a lot. I always liked the flow in
Extremes and Follow The Sign 3 by Byterapers. The 6 sprites over FLI routine by Ninja must be the most insane piece of code ever written.

CHIPFLIP > Do you have anything else you would like to add?

SOUNDEMON > I find it funny how this 8 bit sample routine became such a success.
I have always considered samples on C64 quite boring! Writing a sample player didn’t seem so interesting… But once I got an idea on how to implement the routine I wanted to try it. I guess the result was a bit more exciting than I would have expected…

Finally I must add that the 8 bit samples in Vicious SID wouldn’t have been the same without Mixer. He did an excellent job utilizing the routine! He also spent LOTS of time experimenting with the routine.

Handmade Electronic Music: Bending vs Building

April 29, 2009

In 2006, Nicolas Collins‘ released his book Handmade Electronic Music (Google’s pirate copy here). Yesterday he presented the second edition at STEIM in Amsterdam, an institute which has been active in this area for 40 years already. Collins has similar authority on the subject, being a professor of Sound and a very experienced low-level sound art performer.

I haven’t read the book, but if little-scale lists it as an inspiration it is probably a very good read. I went to the presentation expecting to get an insight into this practice, since I think it is interestingly placed inbetween chip music and circuit bending. It is more than just circuit bending because it doesn’t rely on readymade systems (just components). It is like chip music in the sense that all the audio/music is handmade; it doesn’t use large chunks of sampled audio or algorithmic compositional elements (like most other electronic music).

Nic (btw, not Nick) started with two performances: one with a group of people poking a circuit board to make sounds, the other one with a lit candle performing similar sounds. Fire-driven music is nice stuff and with Nic blowing wind on the candle, the sounds would change. So now, in a broad sense, there is chip music made with fire, wind and water. Hope to see more elements!

“Last time I was here I talked so much, so this time I will show examples instead”. Assuming that everybody was there the last time, Nic instead ran the DVD included in the book. It was like a very long Youtube session with 1 minute clips of handmade electronic music. Definitely very interesting, for a while, but it was not what I was hoping for. The clips were more like tech-demos and noise than performances with musical instruments. That statement is of course leaning towards musical conservatism, but sometimes we need that too, eh? : ) I can continue along those lines by saying that most of the devices made very similar sounds. You know, those scratchy and pitchy pulse wave sounds that the Cracklebox at STEIM made already in the 1970s. If you’re not in the mood, it gets pretty tiresome after a while..

But I also think that chipmusic and demoscene practitioners could learn a lot from the conceptual and noisy ways of sound art and circuit bending/”building”. It is funny how circuit bending, chip music, and the demoscene is sometimes presented as related to eachother, eventhough they are so different. Chip music is (too often) about 4/4 happy bleep pop and using default samples of LSDJ. Demoscene music is (too often) about perfectionism and competition. Circuit bending is (too often) about tech-concepts and predictable noise.

What they do share, is a fascination with the possibilities (aka limitations) of hardware that is old or open (enough). In the demoscene, hardware is losing some of its priority to make room for emulators and design/concepts instead of coding brilliance. Chip music seems to get more tech-fundamental at the moment, and as for circuit bending I guess that hardware will keep on playing a vital role (eventhough “software bending” such as glitchNES has appeared). What it ultimately amounts to, is a discussion on what a technological system is and also if/how a computer composer can operate independent from capitalism and culture. (any suggestions? hehe)

I think that “handmade” goes just as much for software as hardware. You often forgot the extent to which some chip music is handmade. At the end of the day, that might be more relevant than the mantra of “commodity subversion” and if so, maybe chipmusic is more similar to circuit “building” than circuit bending. Well. Just continuing the ramblings about how to contextualize and explain chip music so we don’t have to be blamed for being DJs/gamers on stage. We can play as much or as little live as other electronic musicians, damnit. Ciao gringos.

Sequencing Computer Peripherals

April 23, 2009

I just found a version of Bohemian Rhapsody performed by an Atari800XL, 8″ floppy drive, TI 99/4a, 3.5″ floppy drive and four HP ScanJets. It’s apparently the hottest youtube-clip in Canada right now, yip yip! The same author also has Funkytown performed by C64/modem/printer and TI99/4a. Mentioned as his inspiration is James Houston’s Big Ideas (Don’t Get Any) which had a slow start of its Internet career, but has received lots of internet attention by now. It’s James’ final project for design school, so the visual aspect is also well worked through. A very special clip. It’s a ZX Spectrum with scanners, harddrives, and printers that performs a Radiohead-cover. James “placed them in a situation where they’re trying their best to do something that they’re not exactly designed to do, and not quite getting there”.

While many chipmusicians claim to re-purpose technology, sequencing computer peripherals like this doesn’t even involve a sound chip! The first time I came across it was on the Commodore 64, where software would play music with the drive header. There is a youtube example of the 1541 drive playing Bicycle Ride For Two (originally from the first “chipmusic” record Music For Mathematics, 1962). There is also atleast one application to do this: 1541-music (1987), but don’t test it if your diskdrive is dear to you.

Back in the days, computers did not have a DAC (digital-to-analogue converters) which turn bytes into vibrations for loudspeakers. There is a peculiar story from 1966, when Tanzanian visitors to Sweden were treated with a printer playing their national anthem! Supposedly, this was the easiest way to make computer music for these engineers, although there was squarewave music elsewhere in Sweden at the time (where some pretty hardcore arpeggios were eventually made).

At the time, keyboards and screens were not common place either. Even in 1975 the Altair 8800 was just a box with switches and lights. The American hobbyist Erik Klein bought this computer as a kit and “30 hours later it was running with only one bug in the memory!” He happened to notice that the Altair was interfering with the nearby AM-radio, and he figured out how to control the tones and play his own music – “with nary a glitch“. Possibly this is the first piece of computer music made outside academia/art/videogames. But, the sounds are not digital and an AM-radio is not really a computer peripheral anyway.

On another (ir)relevant note, peripherals have been re-purposed in the C64 demoscene. If you run out of memory or CPU-power on your Commodore 64, you can use the 2 KB RAM and 6502 CPU inside the 1541. One example is the demo Deus Ex Machina (C64 2000) by Crest. Jeff’s song “Crossbow” apparently plays from the diskdrive.

So, the lesson to learn is that computer peripherals can be great tourist attractions and can probably be used for even more bizarre things. I’ll finish off this post with some more examples of music with peripherals.


  • Paul Slocum and his dot matrix synth, used for exhibitions and the excellent music project Tree Wave.
  • Sue Harding’s Dot Matrix music. youtube. Does not involve any programming, but rather trial and error style by printing images and see how they sound. Notice the Amigas!
  • Little-scale has a number of printer projects and an arduino tutorial aswell.
  • Half Arsed Printar Shreddage at youtube. Feeding samples into a dot matrix printer head.
  • Gijs Gieskes’ Image Scanning Sequencer
  • Amiga-drive performing El Condor Pasa (stepmotor) youtube
  • Amiga-drive performing a melody (“spinmotor”) youtube
  • Amiga-drive playing a sample. youtube


  • Tape Composer (C64 2009) Compose music for the Datasette (the “tape deck”). It plays back either through the motor, or through audio tape decks (the music you make is saved as data that sounds like your music, uh when you play it as audio) more info here. When I tried it I didn’t get much sound out of my datasette.
  • Tap Music Composer (ZX Spectrum 2007) I forgot how this works, but the results sound like data-cassettes in the right tones.

All Possible Digital Music By Definition

February 13, 2009

Discovered in 1927 by David Champernowne and Alan Turing, the CC is known to contain all possible digital data and is obviously prior art to all possible digital sounds. Which means, no new digital sound was ever invented nor recorded nor ever will be. […] I think most people would agree that this demonstration does not play any good music in this video, but nonetheless it is a matter of fact that the process in progress does generate all possible digital music by definition.” (video)

I am guessing there is a typo somewhere up there, but I hope not. That statement is great. So, VironCybernet just uploaded loads of new videos with interesting generative works, technesthetics, fractal music, digital hardware audio stuff (aka software), speechsynthesis, and so on. I don’t know much about what technology he uses (much of it seems to involve the Propeller microcontroller), but the looks and sound go straight into my telnet heart!

Z80, forgive me

November 25, 2008

Ok, the new thing is 1-bit music made with Z80! Just like with the AppleII-post before, this is me being an astonished newbie. It sounds so nice and data, I can’t believe it’s not a sound chip!

Z80 is the processor found inside shit loads of 1980s machines, for example Gameboy, Sinclair ZX 80/81/Spectrum, MSX-computers, Commodore 128, TRS-80, Amstrad CPC, and more funny named ones including Galaksija, Tatung Einstein, Coleco Adam, Data Science XOR, Grundy NewBrain, MicroBee, and Tiki 100.

So yesterday I found this and this, filled with mp3s of Z80-music made with the ZX Spectrum (or clones). This is how it works, according to Yerzmey: “Z80 chip produces all sounds and sends them into BEEPER and AUDIO-OUT connector of ZX SPECTRUM (jack) through ULA chip”. Normally you can play 1 channel square waves, but with the 3.5 MHz of this Z80 you can play samples and get up to 8 channels of sounds! So this is another example of chip music that does not only play sounds immanent inside sound chips, but uses the CPU to create a sort of softsynth.

There is a bunch of different software to make Z80-music on the Spectrum, but curiously enough there is only one tracker in the list that Mister Beep shows. They all seem quite interesting. Apparently there is no editor for Tim Follin’s 3 channel sample playing routine. But this month TDM used it anyway, in a coop-track with Mister Beep, where he composed his bits in assembler: Insane organist. The most bizarre Spectrum software might be this one, which lets you compose true data music: you save the song data on to cassette, and when you play it in a normal cassette deck you can hear it again! I haven’t tried the software, so I am not sure how it works, but it sounds like this.

There are lots of people developing things for the ZX Spectrum still, like ZX Spectrum Orchestra. Demosceners hang out at and there’s daily action. A few days ago Yerzmey announced that people are playing Beeper music on Atari XL/XE! It does not use the Pokey, but rather the GTIA (which apparently generates a click sound when typing). Mister Beep also released a new ZX Spectrum demo this month: check it. And get these ones:

Alberto Gonzalez – The Light Corridor (slow and foxy)
Andy Mills – AnoGaia (funky and mini-squeeky)
Ben Daglish – Dark Fusion (rockfunk)
David Whittaker – Brave Starr (micro epic)
Fuxoft – Starfox (rockfunk)
Jason Brook – Rastan Saga (adventure tonality)
Tim Follin – Agent X part 2
Tim Follin – Future Games


Drum Machine (1984), 'photo' by Mister Beep

Italo Disco Noise Digi Screen-music

November 12, 2008

That headline almost attracts anybody, right? :) A few weeks ago the Dutch C64-party X’2008 took place and saw some very technically impressive releases. The winning demo by Booze Design is the new favourite demo of the C64-demoscene, because it is a 15 minute masturbation in coder brilliance. (part1@youtube) Also, the music possibilities of the C64 has taken a big leap forward. Fanta released a song with 4 channels of 8-bit samples and 2 channels of ‘traditional’ synthetic channels. (mp3) Normally the C64 only plays 3 channels of synthetic sounds. The code magicians behind it is Soundemon and The Human Code Machine (nice name!), who also released a demo called Vicious Sid. This one also plays Amiga MOD-music, but there is a part where the music is made by the graphic chip, VIC! The screen shows lines that produce a sound of Soundemon singing with his choir. (mp3) The funny thing is that all these revolutionary music techniques are used to play italo disco-ish music. I personally like it a lot, but there is something quite funny about coding something ultracomplex to play italo disco. :) I will return to these new techniques in a future post.

For now, I would like to contrast these coder porn with some noise porn aswell. Yesterday there was a very refreshing C64-demo released by the Australian coder/musician/graphician A Life in Hell: Fuck The Scene. In many ways it breaks with the flows of the demoscene since it has chunky graphics and dirty glitches. It does include some pretty complex code that appealed to many people, but I like trash style! So noisy and great. While we’re at it, here are 3 other tips for demoscene stuff that is noisy rather than flowing. (any tips is much appreciated)

PWP – Robotic Liberation (Vic20 2006) youtube
Booze Design – Industrial Breakdown (C64 2003) exe
Satori – Trashtank (PC 2002?) exe/mpg

…and just to have something easily clickable in all this nerdery, I embed PWP’s Vic20-demo Robotic Warrior from 2003.

Interactive Chip Music Embroidery

August 4, 2008

Just a little piece of poetic data floss for you to enjoy in the summer nights: Rebecca Stern’s LilyPad Embroidery (2008). It’s a charming mix of chips and floss, which looks pretty damn good to me. It produces lights and sounds, which are affected the amount of light around it. Doesn’t sound or look all that amazing, but with this chip embroidery style it doesn’t even have to. I like it anyway. > via visualcomplexity

video @ blip

Video of the Day

June 22, 2008

Via ablogforyou I just learned about this video, which is a remix of Radiohead – Nude, played by a machine orchestra consisting of ZX Spectrum the computer, Epson LX-81 the matrix printer, HP Scanjet 3c the scanner, and a bunch of hard drives acting as loudspeakers! Amazing work by James Houston. Hmmm, couldn’t get this embedded for some reason, but here’s the link: Big Ideas (don’t get any).

From Pac-Man to Pop Music

June 19, 2008

Karen Collins, one of the few scholars writing on 8-bit music, has just published a book called From Pac-Man to Pop Music. Apparently, it is the first book to present a wide range of texts about music for computer games and game technology, written by both academics, composers and programmers. It includes several interesting texts on adaptive and interactive music composing, and games audio and marketing. And, since I am mentioning it here, also texts about chip music. Actually, yours truly wrote a chapter about the history of chip music. By now the text is not mega fresh, but I got good feedback from prominent chip folks during the writing of it. The main purpose of the text is explaining the history of chipmusic from a non-commercial perspective, ie demoscene rather than videogames. I briefly explain the foundations of the demoscene on one hand, and computer music on the other. My main point is to differentiate chipmusic as medium from chipmusic as form. Chipmusic as medium is any music made with a specific medium (a range of soundchips from the 80s, typically) and chipmusic as form is a music genre made with any kind of technology. YMCK is mostly referred to as chipmusic although they use Mac, while DJ Scotch Egg is typically referred to as breakcore although he uses Gameboy.

What I argue in the text, is that the term chip music was first used in the Amiga demoscene around 1989 by artists such as 4-mat, Duz, and Turtle. Prior to that, it is hard to find references to the term chip music, as it was probably just called computer or digital music. When the term chip music was first used then, it did not refer to music with waveforms coming straight off a chip. The waveforms were sampled and manipulated by the composer using sample-based trackers such as Noisetracker or Protracker. Furthermore, chip music seems to have been predominantly happy 4/4 music often flirting with C64-aesthetics. This might be annoying to people arguing that chip music is 1) not a music genre and 2) based on sounds generated in realtime by a chip. But although it might break a common historiography of chip music, it is because of using a sociocultural perspective rather than (techno)logical empiricist one.

As I wrote in the previous post, computer generated music has been around for almost 60 years now. By tracing the birth of chip music to 1989, we can (atleast theoretically) differentiate between chipmusic as we know it today, and the pre-1990 chipmusic which was mostly made in the names of science, art, conceptual music, and videogames. Technically speaking it might be 57 years old, but culturally I argue that it is 19 years old. This is a theoretical distinction that might seem unnecessarily absurd to some, but I find it useful. Atleast at the moment. I believe that the production, dissemination, and approaches within today’s chip music shares more with the (old) demoscene rather than videogames or art. Fun rather than monetary, playing with limitations rather than concepts, static rather than interactive.

Chipmusic – hardware or software?

April 14, 2008

After writing about Linus Åkesson’s work, I have been thinking more about the definition of chipmusic. Again. So, his hardware chiptune project is really hardcore – programming a chip from scratch to generate music. But, it can also be seen as the complete opposite to chipmusic.

If we use a technical definition of chipmusic – any kind of sounds synthesized by a soundchip – Åkesson’s project is not chipmusic. The sounds that his program generates could might aswell have been something completely different, that wouldn’t sound like chipmusic at all. But, when you use a machine with a specific soundchip inside it (which computers and consoles had until the mid 1990s), the chip has a framework that you need to adapt to. A Gameboy or a NES can play four sounds simultaneously, a Commodore 64 can only play three.

However, most music software goes beyond the absolute limitations of the soundchips. With LSDJ for Gameboy, one of the channels can play two 4-bit samples at the same time, by mixing the sounds. On this channel you can also use a software-synthesizer, which produces sounds not inherent to the chip. On the Commodore 64, clever programmers quickly figured out how to play samples on a “fourth” channel of the soundchip already in the early 1980’s.

Back in the days, it was important that music could be used in games or demos. Composers could not make music that used a lot of processor power to generate sounds, since programmers needed the power to make cool games or demos. Today music seems highly prioritized in the development of games and demos, but back then music wasn’t supposed to use much of RAM or CPU. (Collins) Still today, a lot of software used to make chipmusic doesn’t use a lot of CPU-power, although it is usually not made for games or demos. LSDJ is one example of a tracker that uses more software synthesis, although it was actually used in the Gameboy demo Demotronic (2002) by 1.000.000 BOYS. (and although I am biased here, I think it is one of the best demos ever)

I still argue that the birth of chipmusic as we know it today was around 1990. Although computer music had been made for almost 40 years already, the term chipmusic was not used. Also, it seems it was around this time that chipmusic as a genre was formed. Before the home computer revolution in OECD-countries in the 1980’s, computer music was mainly made in the name of science and fine art. It seems it was about technological progress and pioneering, or conceptual art based in cybernetics, sci-fi, and social critique. (Chandler & Neumark 2005) Basically, it was a lot more abstract than what we call chipmusic today.

I have not seen the term chipmusic in use before the Amiga demoscene started mentioning it around 1990. The term referred to music that flirted with the game music of the C64 and was essentially used in intros and cracktros. But, the sounds were not synthesized by a chip – it was sampled waveforms. It is a fact often forgotten in contemporary history writing of chipmusic. In its infancy, chipmusic was not about realtime synthesized sounds but rather a musical genre – and I would argue this is still the case. But personally, I like the technical aspect of the term chipmusic as I can easily label my music without saying what music style I make. It’s like saying I make “guitar music”. Very convenient + confusing.

Apparently I’ve missed out on good stuff, but thanks to the c64music blog I found out about little-scale who makes very nice chip-related things. Right now, I want to mention his automaton ep (2008) which uses the same technique as the early Amiga chipmusic composers: extremely short samples. All the songs on this EP are based on a single 11.6 millisecond sample, and it effectively proves that this technique can result in pretty complex soundscapes. I made something similar in a tune called Jonkvrouwe (download). This tune is some kind of quick n’ dirty bleep reggae, but it illustrates this technique in an interesting way. There is just one sample of a a screaming girl and I change the loop-positions of the samples while playing the song. When the loop is short enough it sounds like chipmusic! I wonder if it is?