Procrastination Polka
written and performed by Dragan Espenschied
using two Handspring Treo 180s
Replay instructions
1. Get two Treo 180 or 180g from eBay. The telephone unit does not need to work, it is enough if the Palm OS is responsive and producing click sounds.[1]
2. Get the Palm Software M.Play and install it on both devices.
3. Convert the provided Midi files PPolka-lead.mid and PPolka-beat.mid to a M.Play library file using the converter that ships with M.Play. Install the resulting Trial Library on both devices.
4. Run M.Play on both devices. Tap the icon that looks like a folder or whatever and select the Trial Library. The play list will now show the names of the two Midi files.
5. Select a different Midi file on each device and tap bot play icons simultaneously.
6. DANCE!!
Composing
Listen to PPolka-phone.wav for the original sketch, recorded on an answering machine.
A Piezo Beeper as present in the required Palm devices can only play one sound at a time. In order not to require four devices to play four notes together for a harmony, I use a classic sound chip music technique called Arpeggio: All notes of a chord are played one after another very fast. Classic sound chip compositions are usually synced to the Vertical Blank of the monitor, which is 50Hz or 60Hz.[2] So the notes change 50 or 60 times a second, creating the foamy sound we all love!
The Palm OS can only play Midi files, so to make the Arpeggiation I had to insert loads of note events into the Midi data stream. After some testing it became clear that
● I had to use chords broken into 4 notes. If i would use for example 3 notes, the Arpeggiation would last too short for a 1/32 note and then the Palm OS would fuck up the timing totally;
● 96 ticks per 1/4 note are the optimal Midi time resolution because it makes a 1/32 note 12 ticks long. An arpeggiated 1/128 note then takes 3 ticks;
● a song replay speed of 185 BPM is optimal, the Palm OS can not really replay anything faster. So in one second i would have (185 BPM * 96 ticks) / 60 = 296 ticks per second. One 1/128 note takes 3 ticks, so 296 / 3 = 98,666... note changes per second. I could use almost 100 Hz Arpeggio![3]
Do you know any Midi sequencer that allows you set up and work with all these details? There is only one: Cakewalk. If you hate Midi like I do, you will love Cakewalk.
It allows you to edit every single Midi event and has the most control-freaky copy/paste functions I have ever seen. Don’t you just stare in awe at this beautiful interface? No simulated knobs, no wood textures, no fake blinking lamps, no idiotic pictures of rack synthesizers ... but lots of numbers for all the timing stuff that is really important. There is almost nothing in between the user and raw Midi data! Too bad Cakewalk was discontinued ages ago. Honestly i have no idea where from one can get a copy nowadays.[4]
Anyway, I proceeded laying out the notes in two tracks, one for each Treo. Of course the “rhythm”-Track with its arpeggios had much more Midi events than the “melody”-Track that only contains a few clear notes. When I uploaded the two Midi files to the Treos, the rhythm track would always play slower and the timing of the two devices playing simultaneously would run out of sync after only a few bars.
I figured out that, given the extremely slow processor speed of the Treos, the sheer amount of Midi events is stressing the CPU too much and delaying the replay. The solution was to put into both the rhythm and the melody track the same amount of Midi events, occuring at the same time in both tracks as well. I converted the rhythm track note events to KeyAfterTouch events that do not produce any sound on the Treo. Then I mixed these events with the actual note playing events of the melody track; and the other way round, I converted all melody events to inaudible events and mixed them with the rhythm track. Now both Midi files have exactly the same byte-length and my two Treos can play for hours perfectly in sync, i just need to press play on both at the same millisecond.
Again: only Cakewalk made it possible!! Can you do that shit with Ableton, Cubase or the “Loop Tool” of the month? You can’t bro! What you can do there is only rolling your cursor around some graphics that look like professional knobs or faders, stuff that emulates instruments and concepts from decades ago for the “phat analogue sound”.[5]
Recording
As the Treos beeped to my satisfaction I thought that recording would be the easiest part. After all these are just two primitive Piezos! However none of my microphones was able to pick up the sound with acceptable results. After some days I gave up refining my recording setup and asked sound engineer Robert Kaiser for help. He just took the best Microphone available to him, did a Firewire recording, fooled around with the EQs ... and now the recording sounds as if you would stand next to the Treos. 1024 Thanks Robert!!
It was difficult though, because the Piezo beepers only seem to be very loud. But this is mostly a psycho-acoustic effect. High notes are perceived as louder by the human ear than low notes, and the pulse wave with its radical changes in its amplitude also sounds very loud, but in fact the signal is very weak. And there was no special microphone to pick up Piezo sounds, so Robert had to use high amplification and deal with the resulting noise.
Some pixels above I was making fun of simulated Interfaces present in contemporary “professional” music software. Interestingly, Robert’s sound engineering software tools (mainly EQs, Limiters and Compressors in Apple’s Logic 9) sported relatively abstract interfaces. It is understandable that as a sound engineer you need tools that look professional and neutral. You might be mastering a ZZ Top cover band one day and some Euro Trance the next day. I know that there are “mastering plugins” that seek to simulate the sound of tube amplifiers and make an according appearance on the computer screen, but serious sound engineers would hardly use them. In general, mainly composers and producers of music seem to long for something that reminds them of a tangible world.[6]
Choice of Interface
I still refuse to touch software that emulates and/or mystifies sythesizers. Maybe I could pump out 10 tracks a week if I would use the latest music software? I could even try to find a virtual synth that sounds like a Piezo and looks like a Palm. But then I wouldn’t know why I would have chosen this virtual synth. Because I could have used one that sounds like the London Symphony Orchestra conducted by Jerry Goldsmith with all the same effort.[7] More or less, any sound that I would want I could ultimately create. But what sound do I want and what is the process that in the end makes me want something?
Every interesting creation tells the story of how it came into existence simply through the fact that it exists. Of course the precondition for the narrative to work is that people are able to read this story, that there is some general knowledge, or rather interest, about for example how music is usually created. I am convinced that the literacy about creation processes of media products is already quite present and will only continue to grow. As more people at younger ages gain access to all kinds of software, I am sure almost everybody will find five minutes in their lives for playing around with some kind of music software and more or less will be able to grasp the concepts. Just like almost everybody used to sort of understand how a Rock’n’Roll band is supposed to work, the notion of how digital music is made is spreading broadly.[8]
I totally appreciate this development. My opinion is that as many people as possible should be making music. And I am deeply convinced that an artificial mystery has no place in such an environment. Bands and musicians used to cultivate their auras by explaining under what conditions they recorded their latest album or what exotic instruments they used: Maybe it happened on a boat , maybe they put a drumkit imported from Goa into a staircase, maybe there were riots outside while the recording happened, or inspiration struck next to the graveyard where someone famous is buried, whatever ... Today nobody could care less where the laptop running Ableton Live is plugged into the wall, all digital instruments are available for free, hardware is cheaper than ever. And the most awesome sounds ever heard, who really created them? Maybe it was a new VST-Plugin that happened to be released last week, or a lucky stumble upon an anonymous MP3 file on the Internet? This stuff happens to everybody every day, because, let’s face it, most people use more or less the same software in more or less the same way on actually the same network. This might sound boring, but in fact it is a great chance for narration: The amount of signs available for communication is prospering and sprouting. A cultural standard, like a file format standard, affords exchange.
In the case at hand, low-tech digital music, aka chiptunes, hooks into the naive narrative of the computers’ quest for more “fidelity”. The idea is quite old, reaching back to research projects of the 1960’s. At first, as this folklore tells us, computers where only beeping and showing zeroes and ones; now, thanks to natural progress, they show so-called photorealistic scenes and play music that sounds like the aforementioned London Symphony Orchestra. If this trajectory should continue, the computer would become more and more sophisticated in making people forget about itself. -- But, apparently, people already seem to love computers and their strange, inhuman ways. They believe it’s more practical to click an icon than to say supposedly natural voice commands. They like Super Mario exactly because he is a very computerish figure embedded in a very digital-abstract scenario. WTF, some even like fractals, and an unbelievably large number is proclaiming Microsoft Paint® their favorite art tool!
From learning how to formulate good Google queries to fooling with the underlying systems of social networking sites, everyday computer users experience and understand in increasing rates how computers work. And I think they will want to entertain their minds with art that tells them something about the current times.
It seems impossible to impress this audience, their members approaching infinity, with some arcane preconditions about how art “happened”, or with uniqueness as an artist, or that a a knob was turned the other way round than it was written in the manual. Because they have, collectively, seen it all. You can only narrate and have to chose a quality level in the collective computer usage experience for doing so. I think the most interesting story to tell today is that everybody can tell one.
[1]You can actually use any Palm OS 3.x device with a built-in Piezo Beeper. However the Treo 180s have the loudest Beepers of all Palms I ever owned. So, for best results!
[2]50 Hz for PAL, 60 Hz for NTSC, depending on where you live. Today these broadcasting standards are slowly replaced by a plethora of crappy digital protocols that barely work.
[3]A lower number would have been bad for the composition. Most chord strikes should be only 1/16 notes, running through all four notes of a chord two times is required so it sounds remotely like a harmony.
[4]How could Cakewalk become so awesome? I suspect because the Microsoft Windows® operating system it runs on is so badly designed for audio that there are always timing problems to be encountered and Cakewalk needed ways to fix them.
[5]Protip: Apply Note Velocity to change the pulse width on the Treo! No extra event needed!
[6]Probably they dream of standing on a stage playing instruments in front of an enthused audience that is envying their gear! A sound engineer is usually someone in the back, making everything sound well -- a very crafty task.
[7]aka Piratebay.org
[8]And, as you probably know, it works much easier than how a Rock’n’Roll band makes music.