Are there any free standalone concatenative synthesis programs?

Have you tried the IRCAM catart / mubu options? Of course they require you use Ableton / M4L / MAX.




I am afraid I don’t think there actually are any standalone packages.

Maybe join the IRCAM Forum and ask there. Diemo Schwarz is the concatenative synthesis expert that would know.

I’ve found that exploring academic papers published in the appropriate music science journals can help send you on an exploratory journey in this field. I’ll add two good reads here.

Especially reading though the references literature list has given me new insights. That said, I think in terms of practical applications Skatart may be the current best implementation (which is macOS only at the moment, I think). A lot of solutions seem to be written from an academic Proof-of-Concept point of view and not so much with actual musicians in mind. Somehow, conceptually the principles seem to make a lot of sense, while in practice they render ”granular synthesis” type results at best.

I'd also recommend these Ircam projects. CataRT-Mubu and the legacy CataRT CLassic are free, you may need to join the forum. But you'll need MAX to work with them as they are really a collection of patches that you can customise and combine to build your own 'instrument'.
https://cycling74.com/products/max
https://cycling74.com/products/max

CataRT standalone and SKataRT require an Iracm forum subscription.

I used to perform with a patch built from the CataRT classic and used live contact mic input to trigger the corpus audio and use midi controllers mapped to various parameters such as pitch, duration etc. CataRT standalone will do this although it is a bit of a legacy product now as CataRT-Mubu has superseded it.

@Breve Yes, unfortunately the standalone versions require subscriptions as you've mentioned.
Do you mind linking this classic version of Catart from the forum?
I find it strange that there is no windows standalone program other than the Catart one out there.

-I'm uncertain whether audiostellar can do it or not

I should probably say what im after...because maybe there are alternatives.
I have some midi being sent out from a program and I want to trigger fragments from a piece of audio that match the length of the input midi note.
The midi notes would be consistent (as in they wont be changing notes), but I want to be able to randomly shuffle new fragments to be assigned to these incoming midi notes.
If there is some simple windows program that fits this bill...

This page has links to the various versions but the link to CataRT Classic seems to be broken. I'd imagine that it will still be available somewhere, but you may have to request it on one of the CataRT forums. I was in an email group for a while that Diemo Schwartz ran (he built the original CataRT).

If you are plannning on trying it out in Max I'd probably recommend going with the free CataRT Mubu objects and patches as this has replaced the classic version and you will find better tutorials and support.

Either way you'll need to pay for MAX (or trial it) and learn how to use it if you don't already. It's quite a steep learning curve. You may be better subscribing to the forum and going with one of the ready made options, CataRT standalone or SkataRT (if you have Ableton Live). These should work off the shelf and not require any programming and building in Max. Subscription cost is not insignificant but I think you get to keep all the premium software after your sub expires, just without access to updates and other perks.

I'm on Mac so I'm afraid I haven't looked for anything specifically for windows. But it think all the ircam versions/ Max are cross platform.

There was one hobbyist who used the MAX catart code to build their own M4L device I think. No idea if it actually worked properly and / or if it’s still available on the MAX website.

Yes, there may be M4L devices or Max patches available from the community for free. It's worth a look.

There is an implementation that doesn't require Max - it's written in Python and can I think use csound for rendering. But for Windows, you're going to have to do some assembly as it was originally written on a Mac, like much of the Ircam stuff (which is pretty much why you don't see Windows versions). And forget real-time MIDI input. This involves putting together an input file in something that itself looks a bit like a chunk of Python code and what you get out you then need to feed to csound (ie the free option) or Max/MSP (the not-free option).

There was one hobbyist who used the MAX catart code to build their own M4L device I think.

Click to expand...

This implementation looks like it has a lot of limitations as far as analysing the samples going in the corpus and how they are triggered. I'm not sure it does audio mosaicing, which is what you need if you want to replace input audio (your midi notes) with grains from the corpus.

See the video @doctoremmet posted for a better overview of the possibilities.

Breve said:

This implementation looks like it has a lot of limitations as far as analysing the samples going in the corpus and how they are triggered. I'm not sure it does audio mosaicing, which is what you need if you want to replace input audio (your midi notes) with grains from the corpus.

See the video @doctoremmet posted for a better overview of the possibilities.

Click to expand...

Just throwing some unresearched options out there. I know the possibilities; I’ve worked with catart / mubu myself and posted that IRCAM video myself this morning - which I still think is OP’s safest bet. But of course it does require Ableton Suite and / or MAX.

doctoremmet said:

Just throwing some unresearched options out there. I know the possibilities; I’ve worked with catart / mubu myself and posted that IRCAM video myself this morning - which I still think is OP’s safest bet. But of course it does require Ableton Suite and / or MAX.

Click to expand...

Yes, I agree. Just pointing out that the Corperous Concativus may not do what the OP is looking for.

Breve said:

Yes, I agree. Just pointing out that the Corperous Concativus may not do what the OP is looking for.

Click to expand...

I am afraid you’re right.

Thanks all.
For instance if I have a C note, a D note and an E note being sent, and I want to "shuffle" (I mean literally like a shuffle button) what fragment is used for each of these input notes, is concatenative synthesis needed?
Or can granular synthesis or some other method work too?

SaxOnTheBeach said:

Thanks all.
For instance if I have a C note, a D note and an E note being sent, and I want to "shuffle" (I mean literally like a shuffle button) what fragment is used for each of these input notes, is concatenative synthesis needed?
Or can granular synthesis or some other method work too?

Click to expand...

To be honest concat is probably overkill just for this. You could build something like this quite easily in MAX. But a granular synthesis engine should be able to do this. If you have control over the number of grains you could set it to 1 or a very slow grain rate. You would usually also have control over the grain size/duration, pitch volume, pan etc.

Maybe something like this?

Didn’t that Unfiltered Audio granular reverb plugin respond to MIDI as well? SILO? But that would imply you’d have to feed in live audio into it separately, so it wouldn’t help achieve your objectives either I guess.

@Sampleconstruct Simon Stockhausen has a lot of experience with Arturia’s EFX_Fragments as well as lots of other sample manipulation tools. Maybe he can chime in here as well?