Steinberg Discontinuing VST2 Support in its products
-
Music Engineer Music Engineer https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=15959
- KVRAF
- 4378 posts since 8 Mar, 2004 from Berlin, Germany
there, it's even explicitly stated that midi plugins are not possible at all with vst - which is a much stronger statement than "some may have problems with vst3"
edit: i'm refering to this quote from Arne_Scheffler:
edit: i'm refering to this quote from Arne_Scheffler:
so, apparently, vst was never supposed to support the creation of midi plugins but vst2 kinda allowed it by accident - and vst3 doesn't anymoreyour frustration comes from the misunderstanding that you can build MIDI plug-ins with VST. VST describes an audio plugin API. That you could misuse version 2 for building MIDI plug-ins was not intended.
Last edited by Music Engineer on Mon Feb 07, 2022 6:15 am, edited 1 time in total.
- KVRAF
- 14435 posts since 16 Feb, 2005 from Planet Earth, Somewhere
Exactly..Music Engineer wrote: Sun Feb 06, 2022 7:33 pm there, it's even explicitly stated that midi plugins are not possible at all with vst - which is a much stronger statement than "some may have problems"
edit: i'm refering to this quote from Arne_Scheffler:so, apparently, vst was never supposed to support the creation of midi plugins but vst2 kinda allowed it by accident - and vst3 doesn't anymoreyour frustration comes from the misunderstanding that you can build MIDI plug-ins with VST. VST describes an audio plugin API. That you could misuse version 2 for building MIDI plug-ins was not intended.
Which is what I had referred to previously as a different paradigm.
SB position is that vst3 is working as it was designed and some developers are saying but it is not working like vst2 did.
I believe that is why some have no issues and some do, depends if you accept the paradigm or not.
rsp
sound sculptist
-
- KVRAF
- 1985 posts since 14 Mar, 2006
Arne made that statement hastily in my opinion. VST3 clearly supports midi-driven instrument plugins.....so its not completely only audio. But his view is that plugins should not ever need to PRODUCE midi output, only audio. Plugins should be fed either audio, or non-audio performance data from the host DAW (Cubase)...and produce audio output.
Arne's point of view is that raw MIDI as a serial protocol, should only be handled by the host DAW...and that plugins should never see raw midi. His point of view is that plugins should be using some kind of abstracted view of data coming from the DAW that represents notes and expressions, keyswitches, etc.. and never raw midi. That abstracted view of data from the host...whether its audio or other kinds of performance data...can be rendered into audio by the plugin...
That's a pretty limiting myopic view in my opinion, but anyway, that appears to be the view they took when they undertook architecting the VST3 api around that mindset. That is why it is particularly limiting in certain situations with how midi is handled. That definitely excludes midi plugins, which Arne has basically said, Steinberg doesn't support at all....and also some use cases for instruments if and when the raw midi stream being fed to it is mangled by that abstraction process and makes it impossible for a plugin to exactly handle what was intended by the composer. There are holes in this design...and particular midi use cases where the VST3 abstraction loses ordering information. There have been and continue to be some problems with outputting certain kinds of midi events...extreme work arounds have been needed... midi learn often can't be done with VST3..the extreme work arounds are actually extreme.
Listen, Steinberg is perfectly within their rights to make their proprietary format more limiting if they want to. The reason people aren't using it is because of that. VST3 is more limited then VST2. That is a fact. That is part of why some people want to keep using VST2. In theory this should only affect Cubase users that love Cubase so much that they are forced to use VST3 with its limitations (and benefits). Everyone else should have no reason to be concerned about it...but unfortunately the rest of the industry made VST2 a defacto standard..which has worked fine for quite a while...but now Steinberg will be taking that away....and VST3 is not the right replacement for most of us.
long live CLAP.
Arne's point of view is that raw MIDI as a serial protocol, should only be handled by the host DAW...and that plugins should never see raw midi. His point of view is that plugins should be using some kind of abstracted view of data coming from the DAW that represents notes and expressions, keyswitches, etc.. and never raw midi. That abstracted view of data from the host...whether its audio or other kinds of performance data...can be rendered into audio by the plugin...
That's a pretty limiting myopic view in my opinion, but anyway, that appears to be the view they took when they undertook architecting the VST3 api around that mindset. That is why it is particularly limiting in certain situations with how midi is handled. That definitely excludes midi plugins, which Arne has basically said, Steinberg doesn't support at all....and also some use cases for instruments if and when the raw midi stream being fed to it is mangled by that abstraction process and makes it impossible for a plugin to exactly handle what was intended by the composer. There are holes in this design...and particular midi use cases where the VST3 abstraction loses ordering information. There have been and continue to be some problems with outputting certain kinds of midi events...extreme work arounds have been needed... midi learn often can't be done with VST3..the extreme work arounds are actually extreme.
Listen, Steinberg is perfectly within their rights to make their proprietary format more limiting if they want to. The reason people aren't using it is because of that. VST3 is more limited then VST2. That is a fact. That is part of why some people want to keep using VST2. In theory this should only affect Cubase users that love Cubase so much that they are forced to use VST3 with its limitations (and benefits). Everyone else should have no reason to be concerned about it...but unfortunately the rest of the industry made VST2 a defacto standard..which has worked fine for quite a while...but now Steinberg will be taking that away....and VST3 is not the right replacement for most of us.
long live CLAP.
MacPro 5,1 12core x 3.46ghz-96gb MacOS 12.2 (opencore), X32+AES16e-50
-
Music Engineer Music Engineer https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=15959
- KVRAF
- 4378 posts since 8 Mar, 2004 from Berlin, Germany
both of which may be true. the fact that people used vst2 for creating midi effects, even though that was never intended, is probably an occurrence of:SB position is that vst3 is working as it was designed and some developers are saying but it is not working like vst2 did.
https://www.hyrumslaw.com/
...sort of. however: i guess at this point, it should be fairly uncontroversial, that the creation of midi plugins with vst3 is more problematic than it was with vst2. even though people were not supposed to create midi effects with vst2, they did. why? because, they can!
Last edited by Music Engineer on Mon Feb 07, 2022 8:06 am, edited 2 times in total.
-
Music Engineer Music Engineer https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=15959
- KVRAF
- 4378 posts since 8 Mar, 2004 from Berlin, Germany
yeaah...i wonder how even something as common as voice-stealing in a synth is supposed to work properly, when the order of the note events is getting lostDewdman42 wrote: Sun Feb 06, 2022 7:59 pm There are holes in this design...and particular midi use cases where the VST3 abstraction loses ordering information.
-
- KVRAF
- 1985 posts since 14 Mar, 2006
well its more subtle than that. Notes do not lose their order compared to other notes. Its really only certain unique use cases. a lot of cases are "fine". Its more about the ordering between notes and CC's when there are notes and CC's interwoven on the same timestamp.
But there are other situations related to getting midi learn to work right... getting program change events to work right in a plugin, etc.. and the outputting of various non-note midi events from a plugin is basically still problematic.
Anyway, don't want to get into the weeds on these specifics, it has been documented on other threads in the past. its just fundamentally a design flaw stemming from the fact that VST3 does not pass the raw midi to the plugin...its abstracted..and when its abstracted the queue ordering of the original raw midi queue is lost....and in some cases it may be even that not all midi events would be passed in perhaps... There is a translation by the host DAW form midi to VST3-abstraction....and in that translation...some exactness is lost. In many cases it won't be a problem...the vast majority of cases it won't be a problem. But sometimes it is a problem...and its fundamental in the design of VST3...which is not meant to handle midi apparently.
But there are other situations related to getting midi learn to work right... getting program change events to work right in a plugin, etc.. and the outputting of various non-note midi events from a plugin is basically still problematic.
Anyway, don't want to get into the weeds on these specifics, it has been documented on other threads in the past. its just fundamentally a design flaw stemming from the fact that VST3 does not pass the raw midi to the plugin...its abstracted..and when its abstracted the queue ordering of the original raw midi queue is lost....and in some cases it may be even that not all midi events would be passed in perhaps... There is a translation by the host DAW form midi to VST3-abstraction....and in that translation...some exactness is lost. In many cases it won't be a problem...the vast majority of cases it won't be a problem. But sometimes it is a problem...and its fundamental in the design of VST3...which is not meant to handle midi apparently.
MacPro 5,1 12core x 3.46ghz-96gb MacOS 12.2 (opencore), X32+AES16e-50
-
Music Engineer Music Engineer https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=15959
- KVRAF
- 4378 posts since 8 Mar, 2004 from Berlin, Germany
yeah - ok - neither do i want to hear themdon't want to get into the weeds on these specifics
-
Music Engineer Music Engineer https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=15959
- KVRAF
- 4378 posts since 8 Mar, 2004 from Berlin, Germany
- Banned
- 4491 posts since 8 Jul, 2008 from UK
The more I read about this, the more I understand why CLAP is being talked about so favourably. I wasn't sure at the beginning , but it does seem like Steinberg are hell bent on ignoring everyone's pleas for MIDI CC support etc.
I don't understand the coding, but I suspect unless they start to listen and work with people, more and more dev's are going to be supporting CLAP. Which might not be a bad thing.
I don't understand the coding, but I suspect unless they start to listen and work with people, more and more dev's are going to be supporting CLAP. Which might not be a bad thing.
Don't trust those with words of weakness, they are the most aggressive
- KVRian
- 823 posts since 27 Aug, 2020
Reaper, which is admittedly not likely to drop VST2 support any time soon, but let's face it, the VST2 purge is coming sooner or later.
It's obvious St*inberg want to play the role of legalistic dickheads in this situation, so VST2 is bound to become extinct and it's only a matter of time when that's gonna happen exactly. I have zero trust in what Steinberg do at this point, it's a goddamn debacle.
VST2 is dead for all intents and purposes and VST3 has always been more or less of a failure, especially when MIDI and backwards compatibility are concerned. CLAP must be successful, otherwise St*inberg will continue to keep all plugin devs and users on a tight leash.
- KVRAF
- 6208 posts since 25 Dec, 2004
they can pry my FeelYourSound plugins out of my cold dead hands
sketches... http://soundcloud.com/onesnzeros
some artists i support... https://bandcamp.com/spectraselecta
some artists i support... https://bandcamp.com/spectraselecta
-
- KVRist
- 233 posts since 19 Aug, 2021
Serious question: Is that going to impact me at all since I'm on a Mac and using AUs? Even if everyone drops VST2 support across the board, is that even relevant to me then?
- KVRian
- 823 posts since 27 Aug, 2020
Probably not. The success of CLAP, however, might be.flori89 wrote: Mon Feb 07, 2022 8:16 am Serious question: Is that going to impact me at all since I'm on a Mac and using AUs? Even if everyone drops VST2 support across the board, is that even relevant to me then?
-
- KVRist
- 233 posts since 19 Aug, 2021
Oh but if I understood that correct CLAP is open source. Also u-he built it and if I trust anyone to get it right, it's Urs and his team.crickey13 wrote: Mon Feb 07, 2022 8:18 amProbably not. The success of CLAP, however, might be.flori89 wrote: Mon Feb 07, 2022 8:16 am Serious question: Is that going to impact me at all since I'm on a Mac and using AUs? Even if everyone drops VST2 support across the board, is that even relevant to me then?