Search options (Page 1 of 4)

Wow, okay... Thanks all - sounds like I'd better get that dispute under way.

I wanted some of the major elements of Protracker/FastTracker behaviour available to me in modern DAWs, but without the fastidious attention to detail that frankly inhibited my creative process for years when I used to argue on forums about the minutiae of interpolation techniques, etc. I've coded my own MOD/XM replayers, I know what's involved in getting *the sound*, but I realised that life's too short and that what's just about good enough for my ears is more than good enough for casual listeners. It's great that people are still concerned about the details, but this is a very personal thing based on the sad fact that time is money and I need to actually finish tracks to deadlines rather than indulge in fiddling with bytes

SO... I made Chipslapper, a Reaktor ensemble. Yes, it requires a licenced (or pirated) copy of Reaktor, but I haven't the skillset required to make a C++/JUCE VST so it's the next best thing in terms of portability between Windows, OS X, Renoise, Ableton Live, Logic etc. You can see the details on the page there, but basically it does:

- all the waveforms joule listed plus freehand waveform drawing
- LFO'd PWM
- ADSR (but not freehand-drawn envelope)
- vibrato with a sweep (just like FT2)
- a nice big pitchbend value so you can draw pitch envelopes with DAW automation
- a choice of mono or poly and a basic delay that's designed to sound like multichannel pattern-delay
- a waveform slot for files imported via Reaktor's Sample Map system
- some control over sample start/loop points
- an extra waveform type which is basically all of the samples from ST-01
- a random panner with variable maximum width

It DOESN'T have an arpeggiator (I intended to use built-in DAW MIDI arps), pan envelopes (easily automated in the DAW, which felt like a better solution to me anyway) or randomised step modulation - but read on for more about that last one...

Stuff I wasn't bothered about included compression, filtering, or anything else that I'd probably be doing as part of my general mixdown. Again, fidelity (relative or otherwise) and historical accuracy weren't critical, but 'feel' was.

Then a few nights ago, I decided to redo the whole thing in Max/MSP, to use with MaxForLive in Ableton Live. Then I decided "fuck that" and remade the whole thing just using native Ableton Live devices instead - loads of Sampler instances filled with waveforms. It's called Chipslapper Rack and you can find it on the page linked above. And, although it's arguably more limited than Chipslapper in that it's only for use in Ableton Live, it's got a number of advantages (mostly outlined on my page). Pertinent to this conversation, though, are things like:

- it uses looped 64byte waveforms that I created myself (32bytes wouldn't loop in Sampler) rather than Reaktor oscillators
- Interpolation can be disabled or set to one of three quality levels in Sampler, so they're mostly off but you could change that if you wanted
- Step modulation for pulse waveforms is implemented and it's randomised With only changes on note retrig, as joule specified.
- I've built an Ableton Arpeggiator into the rack this time, preset to a close approximation of the kind of arps I often use - but configurable via Macros
- As well as a delay, which works a little better than in the Reaktor version, I've also included Redux for sample/bit reduction. Probably unnecessary, but once in a blue moon I'd use it for a cheesy automated drop/fadeout
- Input velocity can be adjusted on the panel Macros too, as well as the ADSR, the random pan width (0 == centre), portamento time (you can dig into the Sampler instances if you want to change between portamento and glide; sadly that option couldn't be Macroised)
- White, pink and brown noise, still beholden to the ADSR and most of the other features
- Best (worst) of all...I've now got the first three ST-XX sampledisks built in as the final three Waveform types; 'ST-XX Sample', needless to say, lets you scan through all of them. A few are missing from ST-03, as it's got more than 128, but who cares

Again, the big caveat - I wrote this for my own use. It does exactly what I need and it makes it really easy for me to get some of the sounds/behaviours I want in a fully-sequenced multichannel DAW track without having to render out of Milkytracker or whatever, as well as taking advantage of all the handy parameter recall stuff in Reaktor/Live. If anyone else finds either of them useful, that's great - I've put them both up for free, so go mad. I should say that I'm unlikely to implement any major changes or alter anything that somebody might think isn't 'authentic' enough, but since Chipslapper is an .ens and an .alp, the source is all there for people to mess with if they want.

Hope it helps!

Hi SketchMan3! I'm afraid there are various tedious reasons why Chipdisco isn't resizeable, the most stark and unilluminating of which is that Processing/Java applications don't resize well. It's possible to resize them, but because so many percentage and coordinate-based layout values rely on fixed dimensions and can't be specified as variables during the initial setup/loading, resizing causes everything to break. I spent a lot of time trying to work around this and did actually figure out how the problem could mostly be solved, but it required a few too many compromises and a lot more work than I could afford to put into it (especially after we moved house and had a baby).

In parallel with all this, I was getting annoyed by the fiddliness of the 800x600 interface Chipdisco originally used (I'd wanted to use it on my netbooks) and I just decided that, for the sake of being able to use Chipdisco comfortably the way I was trying to use it - in dark nightclubs after a few drinks - I would increase the size and make all those control elements huge and foolproof. No more pixel-sniping in the dark with a crappy netbook touchpad! The downside is that it forces me to take a 12" Macbook or a 14" VAIO out to gigs, but then my Eee 901 netbook had been struggling with playback occasionally and I'd been starting to doubt that this was a netbook-friendly app. Java's just a resource-hog anyway, especially with my inefficient code, but it's the only language I know

Readme/config files in zips: good question! Easy answer, too: the last time I fixed and rebuilt all the zips, I forgot to manually add them. When I get a spare moment, I'll add them into the build process so it's done automatically

Sorry about the resize thing. I know it's obnoxious of a program to refuse to resize gracefully and one day I'll crack the problem entirely, so it can be done freely (or proportionally). It's a problem for Processing in general, so I occasionally ask around in the Processing community to see if progress has been made. Some day...

Oh dear! Could you try running it from a shell window as described above to see if it puts out an error message? If it's starting up and then just going nuts, it should be printing a stack trace with details of the exception and that might hold some helpful clues. Thanks!

shiroshii - I'd like to support .it, but I'd have to get considerably better at coding in order to implement the filters. The NNAs, I *think* I could get my head around, but I'll have to research the format. I'd also love to support AHX, but that's a whole different kettle of fish...

Hi guys. Thanks, XyNo!

VCMG, Aeros - can you tell me a bit more, please, and I'll see if I can figure out why it's not working? Which OSs, first of all? It could be a really simple thing like the backwards (compared to Linux and OS X) directory slashes aren't being interpreted properly, thus breaking the paths. Or it could be something more sinister So the various methods of adding modules are as follows: Add button, which can add single MOD files (or XM or S3M). Dragging/dropping from an OS folder onto either playlist - this can take multiple files if you select a load before dragging. This method's not advisable during playback as it can cause a slight gap in the audio; the Add button shouldn't, though. Adding M3U playlist, which is just a list of fully-qualified file paths, one per line, saved as something.m3u. Remove any commented-out metadata lines (beginning '#'), as they can sometimes cause problems. Editing chipdisco_config.txt to autoload either a directory or an M3U playlist on startup.

I'm sure you've tried all of these, and though I've tested them extensively on OS X, I haven't had a chance to test on Windows - main reason I didn't bother, though, is that I didn't change any of the relevant code since the last version, which I wrote and tested on Windows XP... So yeah, the more you can tell me about the fault, the better - I'll try to fix it as quickly as possible!

Finally, the Java console will display a lot of feedback while Chipdisco is running: do a google search to find instructions for enabling the Java console in your OS, and it'll clearly state if it's had a problem finding, reading or drag-dropping the files (corrupted headers, incompatible formats, etc.), parsing M3U files, or reading the preset directory.

Cheers!