CLAP... thoughts?

[Markus Krause]

I plan to add CLAP support to our plugins and maybe also to NanoHost.

CLAP does NOT suffer from the problems that Steinberg's proprietary VST3 format has:
- Arrogant and unhelpful developers that ignore the demands of the industry
- No proper Midi support
- A bloated, messy and buggy SDK that fails compiling in many places

I think that CLAP is a well-designed and elegant plugin format with a developer friendly licensing. It should be easy to add to our existing projects.

[jules99]

Bitwig tutorial YouTuber Polarity made a video of the new modulation features, CLAP brings to the table. That really is MPE on steroids and then some.

[Biome_Digital]

Bitwig tutorial YouTuber Polarity made a video of the new modulation features, CLAP brings to the table. That really is MPE on steroids and then some.

Although I wonder if I could do most of that "POLY" modulation in Hive itself, What effect would this have on me as a preset designer. The line between DAW and SYNTH has been blurred now. Actually, I think Reaper has some LFOs that can be used like this to control parameters too.

[Urs]

Are Clap plugins designed to use MTS-ESP? Or rather is that something easily added? Seems like with all the individual voice access it's something that could be added almost as a standard. I'm very interested in seeing where this goes.

Two devs from Oddsound - makers of MTS-ESP - have recently joined our little CLAP chat and of course they're part of our announcement, if that is an indicator.

Does anyone know if there is active development for supporting CLAP in environments like PD, and Supercollider?

That's a little early to say. We're two days in... I did read something about PD a few days ago, but I can't remember of the context was CLAP. Best to ask them directly!

Anyway. I'm interested in CLAP and mostly because developers are interested in it . I've never heard a single developer say "I'm so excited to see what my code can do as a vst3/ AU!". I'm also curious how hard it would be to embed a simple clap host for Bela or RPi. Not being a real coder leaves me not having a clue whether this is something that would be possible (obviously for relatively simple synths and FX), but all the talk of how flexible it is makes me optimistic for some less orthodox uses.

Writing a host for CLAP is almost as easy as writing a plug-in. Unlike a plug-in which can decide to be MIDI only or just stereo or whatever subset of CLAP, a host should probably support as many features as possible.

[Urs]

Bitwig tutorial YouTuber Polarity made a video of the new modulation features, CLAP brings to the table. That really is MPE on steroids and then some.

Although I wonder if I could do most of that "POLY" modulation in Hive itself, What effect would this have on me as a preset designer. The line between DAW and SYNTH has been blurred now. Actually, I think Reaper has some LFOs that can be used like this to control parameters too.

Well, of course one can not save those modulations as presets, so there'll still be ample demand for presets. I see them rather as an extension to classic automation - Bitwig only happens to offer synthesis building blocks for automation, hence in Bitwig Parameter Modulation looks like a sound design tool rather than an option for expression/arrangement. Parameter Modulation however addresses the issue that automation overwrites the parameter value and thus alters the preset. And polyphony, if synth/host developers support that (initially many won't, but they'll gradually add it).

But yes, Reaper has similar concepts which I hope will go that way, too.

[Arashi]

Logic used to support VST plugins. Apple dropped VST support for Logic when they bought Emagic.

[...]

They did so because they wanted total control over Logic's plugin interface. It's that plain and simple. And it's why they will never adopt CLAP either.

More specifically, they dropped VST in the Mac OS X version of Logic. Logic 5 and 6 could still load VST plugins in Mac OS 9, but never supported VST in Mac OS X, because Apple wanted a system-wide standard for audio plugins in their new OS (as you said, probably for control). At the time, those of us who were transitioning to Mac OS X could use a VST-to-AU wrapper (there was one made by FXpansion) to load VST plugins that didn't have AU versions yet.

I agree that it's unlikely Apple will ever support CLAP. Even if it gains widespread adoption and some of its features prove compelling enough for people to consider switching DAWs, Apple's answer (if any) will probably be to (someday) add similar features to the AUv3 spec. And I wouldn't hold my breath for that either. AUv3 is being developed at an absolutely glacial pace. It was introduced in 2015, was only available on iOS for the first few years, and almost nobody is making AUv3 plugins for the Mac yet.

That said, while CLAP has a long, uphill battle ahead, I think the technology is interesting, and there's nothing wrong with trying something new. Maybe the benefits will be compelling enough to make it worthwhile for DAW and plugin makers to adopt it. Performance improvements alone (if significant) could give it a pretty good push.

[GeneralQ]

I do see a Problem here:
Until the point where every DAW including Cubase, Protools and Logic are supporting CLAP as a Plugin format, developers don't benefit from this at all. Because now they have to support AAX,AU and VST/VST3 already. So now they are just adding one more format that they need to support, which increases the workload even more. Protools could probably just be ignored but apple users are sadly to big of a share of the market.

[phazed]

If you ask me what's currently missing: I would like to see a proper website under http://cleveraudio.org, not just a link to GitHub. That would give a more professional touch to the whole project. As an example, have a look at https://bevyengine.org/ which is a game engine, so completely unrelated, but it shows how it can be done nicely.
Also, I would suggest to open a server on Discord or Matrix for discussion between users/developers.

[Ah_Dziz]

Are Clap plugins designed to use MTS-ESP? Or rather is that something easily added? Seems like with all the individual voice access it's something that could be added almost as a standard. I'm very interested in seeing where this goes.

Two devs from Oddsound - makers of MTS-ESP - have recently joined our little CLAP chat and of course they're part of our announcement, if that is an indicator.

Does anyone know if there is active development for supporting CLAP in environments like PD, and Supercollider?

That's a little early to say. We're two days in... I did read something about PD a few days ago, but I can't remember of the context was CLAP. Best to ask them directly!

Anyway. I'm interested in CLAP and mostly because developers are interested in it . I've never heard a single developer say "I'm so excited to see what my code can do as a vst3/ AU!". I'm also curious how hard it would be to embed a simple clap host for Bela or RPi. Not being a real coder leaves me not having a clue whether this is something that would be possible (obviously for relatively simple synths and FX), but all the talk of how flexible it is makes me optimistic for some less orthodox uses.

Writing a host for CLAP is almost as easy as writing a plug-in. Unlike a plug-in which can decide to be MIDI only or just stereo or whatever subset of CLAP, a host should probably support as many features as possible.

Good to know. I'm sure someone will write an external for it soon enough. Hope you are having a good day.

JJ

[Urs]

I do see a Problem here:
Until the point where every DAW including Cubase, Protools and Logic are supporting CLAP as a Plugin format, developers don't benefit from this at all. Because now they have to support AAX,AU and VST/VST3 already. So now they are just adding one more format that they need to support, which increases the workload even more. Protools could probably just be ignored but apple users are sadly to big of a share of the market.

We're internally seen CLAP plug-ins wrapped as VST2 and VST3, support for AU is in the pipeline. These wrappers will be free for anyone to use. CLAP will run in Cubase and Logic, no matter what those companies decide.

The major problem CLAP addresses is that an estimated 90% of plug-ins which are older than JUCE have VST2 as their core codebase and they use wrappers to bind to VST3 and AU. I reckon that more than a quarter of newer plug-ins still do today. As VST2 becomes unusable for legal and technological reasons (e.g. it does not support NoteExpressions), these plug-in will need a new core codebase. CLAP ticks all boxes. It can save those projects from becoming obsolete.

Once such a plug-in is ported to CLAP (in a matter of days, really), all modern features of VST3 can be implemented gradually as well. Finally, more people will implement NoteExpressions and sample accurate automation - which as of now, really, zilch developers do.

I can back this up with testimony from several companies I spoke to, but I do not wish to speak on their behalf.

[Biome_Digital]

Bitwig tutorial YouTuber Polarity made a video of the new modulation features, CLAP brings to the table. That really is MPE on steroids and then some.

Although I wonder if I could do most of that "POLY" modulation in Hive itself, What effect would this have on me as a preset designer. The line between DAW and SYNTH has been blurred now. Actually, I think Reaper has some LFOs that can be used like this to control parameters too.

Parameter Modulation however addresses the issue that automation overwrites the parameter value and thus alters the preset.

Yes that's a big difference.

[noiseboyuk]

Urs - can you describe what the difference (features / performance) will be between the following, from the user perspective?

CLAP wrapped as VST2
CLAP wrapped as VST3
CLAP

...and bonus question, would there be any benefit to either of the first two scenarios in a non-CLAP host?

Thanks!

[jamcat]

We're internally seen CLAP plug-ins wrapped as VST2 and VST3, support for AU is in the pipeline. These wrappers will be free for anyone to use. CLAP will run in Cubase and Logic, no matter what those companies decide.

Are these wrappers external files that call a separately compiled CLAP plugin, or are they something along the lines of a C++ class in the CLAP SDK that handles all of the tedious VST3 code and calls the counterpart functions in the VST SDK if a ‘VST3WRAP’ flag is set?

[Tj Shredder]

We're internally seen CLAP plug-ins wrapped as VST2 and VST3, support for AU is in the pipeline. These wrappers will be free for anyone to use. CLAP will run in Cubase and Logic, no matter what those companies decide.

As Bitwig supports CLAP already, I am more interested in the other way round. A CLAP plugin as AU/VST host to load AU into Bitwig for example. Best would be a wrapper app, that creates a Clap plugin out of a AU/VST you throw at it... Of course that can't do more than the underlying plugin could do in the first place...

[Urs]

Urs - can you describe what the difference (features / performance) will be between the following, from the user perspective?

CLAP wrapped as VST2
CLAP wrapped as VST3
CLAP

...and bonus question, would there be any benefit to either of the first two scenarios in a non-CLAP host?

Thanks!

First off, simply every plug-in that comes in more than one flavour is already wrapped. The performance overhead of wrappers and adapters was next to nothing in 2002, and with today's computers it will be very difficult to reliably measure *any* performance impact.

Featurewise CLAP can not magically add anything when there's nothing there. A VST wrapped to CLAP will only ever offer what a VST offers. Also a CLAP wrapped to VST will also only ever offer what a VST offers. Unless of course, maybe, VST gets updated with a feature that CLAP does not offer, and then we'd first have to add the corresponding extension to CLAP and the wrapper to get back to what I said before.

The main benefit of CLAP-as-SomeFormat is that CLAP has inherent stability testing for many common bugs. Hence, fewer crashes are a very likely outcome of using CLAP somewhere in the development process of a plug-in or host.