CLAP: The New Audio Plug-in Standard (by U-he, Bitwig and others)

[EvilDragon]

Markus, if you're using JUCE's plugin wrappers and CMake, you can just https://github.com/free-audio/clap-juce-extensions

(Well, even if you use Projucer, that's fine too - thanks to effort from Jatin Chowdhury.)


With this, you could possibly be having CLAP versions of your stuff by the end of the week, or so.

[Markus Krause]

Yes. I definitely consider using this.

[baconpaul]

Thanks for your efforts. I already built a relieably working working vst2 to vst3 adapter in Juce. I will definitely try to add CLAP support to my existing products in the next months which should not big a big deal. Either by extending Juce or VST 2.4.

If it helps, I took a bunch of (zero gui) VST2s over the weekend and ported them automatically to CLAP as an experiment. Glad to share that code. Probably best to do that on some platform other than KVR though. PM me if you are interested.

[mystran]

I have a question: what is the modulation range supposed to be? As far as I can tell, the headers say modulation is additive on top of automation (which makes sense), but is it possible to allow modulation range beyond the actual parameter value range? Often times there is pressure to keep "knob ranges" in the sweet spot, even though internally there might not be much of a limit for modulation and I wonder if it is possible to express this in the API somehow?

[DJMaytag]

However:
*Cubase never adopted the AU format.
*Logic never adopted VST.
*Avid does not support anything besides AAX.

Each of those devs want to make their format standard. If Ableton supports CLAP, then there might be a dam breaking that could cause a lot of devs to stop supporting VST3. That could at least force Cubase's hand, since they're supporting the only format that a lot of people are active upset about with the switch from VST2 to VST3.

That said, this situation is VERY much akin to what Oberheim, Sequential, and others experienced in the late 70's & early 80's, with gear that each had its own protocol. Then SCI, Roland, and Yamaha got together and created MIDI. I'm not sure Apple, Avid, & Steinberg are ready to open up to new standards, but Ableton could be the Sequential in the modern day story.

The push would have to come from users, and users don't much care: most Cubase users don't even seem to care that much about using VST2 and possibly a bunch of older plugins no one is around to update.

I'm user, but am watching this with great interest as a Reason user who dumped Cubase as soon as VST support was added to Reason. I am hoping they adopt CLAP as a way to ditch being beholden to the whims of Steinberg. I am not holding my breath on this though.

I am a user who is keenly aware of the update process, having been helping with some beta testing of plugins in the past. I know all this stuff surrounding VST2/VST3 is a major PITA for a lot of smaller developers.

[soundmodel]

Is it reasonable to discard support for VST2/VST3 and produce CLAP-only plug-ins?

Or is this too fundamentalist or otherwise an impractical attitude?

[baconpaul]

Is it reasonable to discard support for VST2/VST3 and produce CLAP-only plug-ins?

Or is this too fundamentalist or otherwise an impractical attitude?

That would be wildly impractical today. Most users still make music in environments which cannot load a clap. The standard is 3 weeks past launch.

But for devs we are thinking that writing a clap and then projecting that into wrapped vst3 and au and so on could very soon be practical. So you would run an “audio unit” but it would just proxy through to the clap. If/when this works (and we have a lot of confidence that it will) a dev could choose to develop clap and then just project into other formats for their users using non-clap aware hosts

This is why you will hear some devs say things like “clap can be wildly successful even if host support is slow”. That internal dev cycle is very compelling

[Richard_Synapse]

We started dabbling with CLAP, so far it looks like a huge winner:

• Very easy to implement, probably the fastest way (other than VST2) to get a plug-in to at least experimental stage where it will show the UI and do some basic processing.
• Multi-threading is both unique and highly useful for multi-threaded plug-ins like DUNE 3. There is still hosts/systems where plug-in multi-threading is troublesome. This could fix this issue once and for all, which would be epic.

Overall I wouldn't be surprised if CLAP takes off like a rocket. So congratulations to Urs, the Bitwig team and all other contributors!

Richard

[Urs]

[No_Use]

If anyone's interested, I got my feet wet doing a simple (GUI-less) CLAP plugin, using JUCE and the clap-juce-extensions and having a GitHub CI/auto build system for it going.

Although there were some problems, it works rather nicely, spits out Win/Mac/Linux binaries (build artefacts) on pushing to GitHub.
https://github.com/nofishonfriday/Lauri ... reo-plugin

Starting point for CI was this template:
https://github.com/sudara/pamplejuce

Some things weren't quite working for me with it (auto builds not completing), so I had to disable some things in Cmake/build script in my forked version:
https://github.com/nofishonfriday/pamplejuce

Would be better to fix these issues of course rather just to disable, but that's beyond my capabilities (help welcome!) and at least the the auto builds complete with it.

[Tim Janik]

If anyone's interested, I got my feet wet doing a simple (GUI-less) CLAP plugin, using JUCE and the clap-juce-extensions and having a GitHub CI/auto build system for it going.

Although there were some problems, it works rather nicely, spits out Win/Mac/Linux binaries (build artefacts) on pushing to GitHub.
https://github.com/nofishonfriday/Lauri ... reo-plugin

Cool, sounds nice and works out of the box in Anklang. I'm just finishing off parameter support, here's how the GUI looks atm:

[No_Use]

Cool, sounds nice and works out of the box in Anklang. I'm just finishing off parameter support, here's how the GUI looks atm:

Wow, nice and thanks.

I'll admit I am (was) unfamiliar with Anklang though, it's a Linux DAW, right?
(the link in your sig to it seems down here btw.)
I'm usually not on Linux but I like to support it.
And what actually does "finishing off parameter support" mean lol? (sorry for the noob question...)

edit:
Ah you integrate your own GUI version of it?

[Tim Janik]

Wow, nice and thanks.

I'll admit I am (was) unfamiliar with Anklang though, it's a Linux DAW, right?
(the link in your sig to it seems down here btw.)
I'm usually not on Linux but I like to support it.
And what actually does "finishing off parameter support" mean lol? (sorry for the noob question...)

edit:
Ah you integrate your own GUI version of it?

Yes, Anklang is developed under Linux, it's a long way from feature complete though, versioning is still at 0.0.1

It is implemented as a DAW audio-engine in C++ and serves JS/CSS files so it's UI is rendered in Electron (or a browser like Chrome or Firefox). For plugins, I've recently written a tiny gtk+2 wrapper DLL, so they can display their own UI in an X11 window if they have one.
Regardless of plugins implementing their own UI, Anklang generates a device block/UI for each plugin based on the parameters exported by it.

The above screenshot shows the UI generated for its internal BlepSynth instrument and your LauridsenSchodderStereo plugin.

The website should be up, but isn't ;-(
Seems like Strato is having a hardware issues again, should be fixed in a few hours. In the meantime you can take a look ath the github repository
https://github.com/tim-janik/anklang

The last nightly doesn't have full parameter support yet, I'm completing that atm and will release a 0.0.2 once CLAP support is half way decent.

[No_Use]

Misunderstanding on my part, thanks for the explaination. Also found the GitHub page in the meantime.
Good luck with Anklang!

edit: Website is up again.

[soundmodel]

Any chance someone could produce a list of central points where VST2/VST3 design falls short, and how CLAP improves on these?

I've been surprised if/that Steinberg hadn't gotten things right, and I considered it interesting to see where they've thought "badly". It's counterintuitive if smaller entities can offer better solutions, but OTOH this is a bit similar to Linux vs Microsoft.