Composers Desktop Project, still relevant ??

VST, AU, AAX, CLAP, etc. Plugin Virtual Effects Discussion
Post Reply New Topic
RELATED
PRODUCTS

Post

Hey guys

I always heard good things about Composers Desktop Project over the years, stuff like it's really cool to mangle sounds, it's very powerful to transform any audio, etc.

I've downloaded it (it's now free) and to be honest, it feels like I'm on a DOS 80's program. Very clumsy, not fun nor intuitive at all and couldn't get anything out of it after more than hour trying & exploring it. Wasn't a very fun experience to tell you the truth. Am I missing something? :lol:

How relevant it is in 2015, with all the powerful plugins available, like all the granulators, lots of spectral plugins and tools, glitch plugins, Reaktor, Max For Live, Metasynth, etc ? I'm aware it was created long time ago when mangling tools were still rare and tough to run on computers. But to be honest, I'm not quite sure it still can be useful with all the modern mangling arsenal at our disposal nowadays.

What do you guys think? Are you a CDP user and only swear by it? If so, help me & tell me what I'm missing that I can't get with the other tools mentioned above :wink:

Post

I absolutely adore it. Paid something like £200 circa 2006 and it still stands as the best music investment I've ever made.

You're not a million miles away with how dated it feels though. No UI can really hide the fact that it's executing old-school command-line programs behind the scenes. I'd hoped that CDP being made free would inspire a new wave of interest leading to some improved graphical front-ends, but (other than the work Xenakios has been doing with Reaper that I'm yet to check out) this doesn't seem the case.

I'm a fan of your granular work, and FWIW I think you'll love it though once you've wrapped your head around it. Bear in mind that granular is called 'brassage' in CDP while 'granular' in CDP typically means processing with the sound 'split into moments' (via envelope tracking) as the base unit, e.g. the syllables of words or the beats in a rhythm. For 'pure' granular work though, I don't think Hourglass can be beaten in the offline processor world these days. I can barely even remember what CDP does brassage-wise as I rarely used those processes. I always preferred tweaking real-time while recording to disk or using Cecilia (then Hourglass when that arrived)

I certainly won't pretend it's pleasant to use, but there are things in CDP that still just can't be done elsewhere AFAIK. The waveset distortions are fantastic (waveset is CDP language for the portion of a wave spanning 3 zero-crossings - i.e. one wavecycle of a sine wave, or really odd divisions if a wave is more complex and contains more than one zero-crossing per wavecycle). I've had some brilliantly wacky results deliberately applying significant DC offset before messing with the wavesets.

Ultimately, it's all down to your workflow though. If you work mostly real-time with synthesis and/or don't use audio files in your production, you probably won't find a great deal of use for it.

For me though, it's the single most consistently inspiring and surprising piece of software I own. The other day I figured out how useful it was for making wavetables. Import any interesting sound, discard all wavesets above and below say, 200hz works well with my voice, and bang! A minute or so of me speaking may yield a <1 second sound file containing around 100 or so cycles which are exactly 200hz, perfect for dropping straight into Serum. The rabbit hole is endless, it really is.

I suppose you could do worse than listen to my track "No I" that was made entirely with CDP and SPEAR (bar audio sequencing and mix processing like compression/EQ etc), with short sample of me singing the "No I" from an abandoned cover of Chris Isaak's Wicked Game as the sound source. The background harmonic fuckery in the opening minute is SPEAR and everything else is CDP. In particular, look out for the 'bouncing' vox sounds (clearly audible from about 2.20 onwards) made with CDP's granular tools (what CDP calls granular anyway). Starting with the raw vocal sample, I inserted random moments of silence to create what CDP calls grains for it to work with, randomly rearranged those grains, then used the grain repeat function with a breakpoint (i.e. automation) file to repeat each grain around 10 or so times in a speeding up, 'bouncing ball' type rhythm pattern. I then stacked three copies on top of each other to create a minor chord with CDP's stack function (which works and sounds the same as loading into a sampler and playing a chord there). Hopefully gives you a rough idea of the kind of workflow that goes into making a complex transform in CDP. So many happy accidents to be found...

https://soundcloud.com/charityqueen/no-i

Post

Oh, and since I forgot to address what you can do with CDP that you can't with real-time stuff, the answer is kind of in the question: non real-time processing. You can grab a segment from the end of the audio file (or indeed any number of files simultaneously) whereas you're kind of restricted to wherever the play cursor is when doing real-time mangling. Amazing for granular work among other things cos you can whip around the timeline totally freestyle (as you can with Hourglass). Can't look into the future with real-time processing (unless you're using a plug which loads audio files of course). If CDP ever got an interface like Hourglass, rather than the clunky breakpoint .txt automation files you currently have to work with, I think I'd wet myself :lol:

Non real-time processing is a great way to generate shitloads of new material quickly too. I prefer Soundshaper as the CDP front end, but the Sound Loom front end is great for batch processing. Stick a folder of 300 hi-hats through a process (or queued list of processes), one or 2 seconds later and you have 300 new hi-hats.

Post

Ooh, and check out Tongues of Fire by CDP developer Trevor Wishart. Made entirely (including mixing and arrangement - nutter) in CDP in 1996, all from the two-second vocal sample heard twice at the start of the track.

Honestly, despite all the virtuoso sound transformation it's the fact that he flippin mixed and arranged it in CDP that blows my mind. I'm not that much of a masochist :lol:

Post

:tu: :tu: :tu: cron!

That's what I'l talking about! Nice input, clear, honest and straight to the point. Thanks for that. Nice audio example too, cool experimental still in there. The major let down for me is really its workflow, I'm not a genius but normally I still can get my head around a new software in an hour or so for the tough ones. This one, ouch! Not fun at all... Even with the GUI (called Soundloom) :(

I'm not afraid of non real-time audio processing, I find those tools pretty interesting as a matter of fact, while they push you to really work on one sound at a single time.

Thanks again for you input, I'll still keep experimenting with it and be patient, who knows it might click with me at some point. I just feel my Max For Live totally crushes it tough, so I'll have to see it CDP can do things M4L can't....

Following your on SC by the way! :wink:

Post

FWIW I never really clicked with Sound Loom at all. I kinda have to refresh my memory how to even work it whenever I use it. I ought to dig in though (been saying that for years!) since it's by Trevor Wishart himself, the guy who developed most (IIRC) of CDP's processes. I fire it up for batch processing once in a while, but just remembering the fundamentals (e.g. how the 'workspace' works - aren't there like, 3 lists of files or something?) means I reach for a guide almost every time. I find it hard to even get audio into a processor without. I find Soundshaper much more straightforward. Load a soundfile, choose a process and some parameters, hit process then hit play. Shame Soundshaper is dire when it comes to data management (particularly the automation stuff). Soundshaper is PC only unfortunately, so you're stuffed if you're on a Mac (I noticed you mentioned Metasynth). Yeah, I've probably only used Soundloom for a combined few hours compared to the solid weeks of time I must have accumulated in Soundshaper.

Must check out Xenakios' work on making CDP accessible inside Reaper. I've had the Reaper demo ready to go for over a year but can't face leaving Live behind. Time to shell out for the upgrade to Live 9 or make that jump to Reaper methinks...

Post

Ah yeah, if you can get hold of a copy of Trevor Wishart's book 'Audible Design', that has a load of interesting stuff in it. It's mostly CDP focused in terms of the transformations he describes (along with some excellent diagrams visualising various CDP processes) but there's loads of interesting stuff on the aesthetics of sound that's interesting even if you and CDP don't get on. An excellent CDP primer though...

There's one really interesting transform in there that doesn't seem to be possible with CDP though, waveset stretching - forcing each waveset to a certain length - which I'd love to play with. Has me wondering if the command line progs have a few undocumented processes lurking in there...
Last edited by cron on Wed Aug 30, 2017 3:55 pm, edited 1 time in total.

Post

What exactly is it?
I found a link. http://www.unstablesound.net/cdp.html#

Post

Unstable Sound hosts the download, but the main site is http://www.composersdesktop.com/

It's a tool for experimental sound processing dating back as far as the mid-80s IIRC. It's not modular or generative or anything like that, very much sound in/sound out (although some processes can have randomness involved). The closest analogy to using CDP is doing destructive editing in a wave editor using your wave editor's menu functions I suppose (except without the wave editor attached!) It originally ran on Atari ST as it was then the only home computer powerful enough to play back a 44.1kHz/16 bit sound file in real time so you could hear your results! Since then it's been ported to PC and Mac and embellished with loads of new processes and whatnot, but sadly the usability has... not... kept up with modern standards.

If you grab the download it comes with an HTML manual describing all (?) the processes available in the software. Alternatively the Soundshaper (a CDP graphical front-end, CDP itself is command line software) manual is spread over a few pages here and gives a good insight into the types of processes possible. http://www.ensemble-software.net/cdp.html

Post

Cool. Will try it out tomorrow!

Post

Thanks again cron. Yeah I'm on OS X here and I really don't like that Soundloom thing. Definitely not for me, too bad I can't try Soundshaper.

Good luck arkmabat, keep us updated here once you've tried it, see how to like it.

Post

I've used CDP for number of years, though, sporadically compared to cron up there..
Started with the XP version (paid!) and enjoyed using it with the SoundShaper tool, it's kind of obvious how you chain commands together, it made a lot of sense! Then I moved to mac.. yup, no SoundShaper, only SoundLoom, which I hadn't used on XP either.

However, it isn't as painful-an-experience as you might imagine, from the posts above. I was thinking (as I just looked and didn't find any) about possibly doing a couple of CDP vids, for SoundLoom. If you have good process and discipline, it's actually quite usable, and easier to use than the command line for the processes I was performing (lots of FFT stuff, formant extraction/replacement..)
And even though I'm technically a nerd (use command line in my profession), I just never got round to using the command line for CDP.

All I'm trying to say is that we shouldn't dismiss technology (esPECially audio tools, surely?) if it doesn't immediately function how we expect. Could you be enticed into CDP if there was an video guide?

Post

CinningBao wrote:I've used CDP for number of years, though, sporadically compared to cron up there..
Started with the XP version (paid!) and enjoyed using it with the SoundShaper tool, it's kind of obvious how you chain commands together, it made a lot of sense! Then I moved to mac.. yup, no SoundShaper, only SoundLoom, which I hadn't used on XP either.

However, it isn't as painful-an-experience as you might imagine, from the posts above. I was thinking (as I just looked and didn't find any) about possibly doing a couple of CDP vids, for SoundLoom. If you have good process and discipline, it's actually quite usable, and easier to use than the command line for the processes I was performing (lots of FFT stuff, formant extraction/replacement..)
And even though I'm technically a nerd (use command line in my profession), I just never got round to using the command line for CDP.

All I'm trying to say is that we shouldn't dismiss technology (esPECially audio tools, surely?) if it doesn't immediately function how we expect. Could you be enticed into CDP if there was an video guide?
Totally agree with you, no dismiss at all here, I was just questioning CDP relevancy since it was made in another era. With the powerful sound design tools that we have now, I'm challenging the fact that it is still a useful tool. I'm a Max For Live & Metasynth user, so trust me I know what it is to not always have immediate results :P

It just seems impossible to find any tips or tutorials on the workflow of Soundloom, readings or videos. If you know some, would be nice of you to share! I just had my loads of codes and programmation with Max For Live, I don't feel like diving again into something zero-visual, all command-lined, not interested. So at first sight Soundloom seemed interesting, although I still have NOOOO idea (yet) how to make it growl 8)

Post

Using the CDP programs from SoundLoom is IMHO worse than running them from the command line! The last time I was even able to get SoundLoom running, I had absolutely no idea what to do in it. SoundShaper is a bit more easier in that regard but yeah...The developer has written it with Delphi and probably has no real plans to ever port it to OS-X.

And to be honest, I never liked SoundShaper much either. So, I ended up doing those programming projects of my own to create CDP GUI frontends, 3 in total...Unfortunately none of those are that ideal for "public consumption" either. I hope I will get the latest development attempt going again. It will feature a modular patching type of GUI to connect the CDP programs together etc. The current plan is to make it work as a standalone software, a VST plugin (though with very obvious limitations and workflow problems) and a Reaper extension plugin, for Windows and OS-X. It currently isn't anywhere near ready to be used by anyone but me.

For cron, the Reaper plugin based frontend (my CDP frontend Mark II) probably has nothing particularly interesting for you. It only contains a limited set of the CDP programs implemented and I have no plans to add more...(That frontend was supposed to be a standalone application too, in addition to a Reaper plugin, but I am not sure how well that can be built and run anymore.)

As far as the CDP sound processings still being relevant, that's completely subjective of course. Personally I got quite nice things done with them for my new composition. Total mangling done to water sounds... ;) The CDP programs are definitely for the kind of stuff like the Trevor Wishart composition linked above. If that's not your "cup of tea", then the CDP programs probably won't do much for you.

Post

Searched and this is the most recent thread, lots of good info in here so i thought it was worth bumping.

Soundshaper frontend has been updated this year. They have dropped Lite / Pro editions, and its now free. If you previously purchased it, you still get email support.
Image

KVR Audio Forum Signature is currently unavailable for your web browser.

Post Reply

Return to “Effects”