Welcome Guest! To enable all features please Login or Register.

Notification

Icon
Error

Forum Jump  
VEPRO and Apple Silicon M1
Last post Fri, Nov 25 2022 by ArasB, 186 replies.
Options
Go to last post
10 Pages<12345>»
Posted on Sat, Mar 19 2022 13:55
by dmansfield
Joined on Tue, Sep 21 2004, New Jersey, Posts 17

+1. I am waiting for this to transition to M1, especially urgent because my DAW is DP11 and there is a MIDI timing bug with DP11 on older Intel machines, so until I switch I'm using annoying workarounds :(

Working under Rosetta 2 sounds too dicey, though if any of you have had good experiences, please post.

No ETA yet??

Posted on Sat, Mar 19 2022 15:11
by evang42
Joined on Fri, Apr 08 2011, Posts 30
I also use DP11 and am a beta tester and can report issues directly to the MOTU devs. What are the steps/circumstances that cause the midi timing bug w DP11/VEP you’ve found on intel???? This needs to be fixed asap regardless of AS transition.
Posted on Sat, Mar 19 2022 15:24
by dmansfield
Joined on Tue, Sep 21 2004, New Jersey, Posts 17

Originally Posted by: evang42 Go to Quoted Post
I also use DP11 and am a beta tester and can report issues directly to the MOTU devs. What are the steps/circumstances that cause the midi timing bug w DP11/VEP you’ve found on intel???? This needs to be fixed asap regardless of AS transition.

With higher buffer settings (e.g. 1048), if you are routing a MIDI track to a VI (any VI, even a MoTU VI), MIDI records about 60 ms late. Also "Fine-tune Audio I/O Timing" does not work on these machines. The only workaround is nudging MIDI back into place after recording. I've been working extensively with Matt Batson in tech support, and the engineering team has been made aware of the bug. They have reproduced it on 2 different Intel machines, and confirmed that it does not happen on their AMD PC or their M1 Mac Mini. I don't know that it is a priority for them though.

There is an extensive thread about this on MoTUNation, "MIDI recording late", where I document what's going on in detail (though you may have to wade through some off-topic or uninformed posts to see it). If you have access to the engineering team and can help nudge them toward supporting the older Intel machines, that would be great!

Posted on Sat, Mar 19 2022 19:53
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

Originally Posted by: dmansfield Go to Quoted Post

+1. I am waiting for this to transition to M1, especially urgent because my DAW is DP11 and there is a MIDI timing bug with DP11 on older Intel machines, so until I switch I'm using annoying workarounds :(

Working under Rosetta 2 sounds too dicey, though if any of you have had good experiences, please post.

No ETA yet??

 I think it's been stated elsewhere that M1 support would come sometime this year. 

Have you tried the AU version of VEP with DP? As you may know there is no Rosetta version of the VEP MAS plugin, but the Rosetta 2 version of the AU seems fine here. However, I've never tried to record MIDI at buffers hight than 512. It's hard to imagine attempting a performance with 1048ms latency - I don't think many users do, so fixing it may not be a priority for MOTU

I've checked out the MOTUnation thread, to be clear here, you've found this issue to be VEP MAS on Intel specific, correct? 

Posted on Sat, Mar 19 2022 20:32
by evang42
Joined on Fri, Apr 08 2011, Posts 30

WOW.  mind absolutely BLOWN.  Thank you so much for bringing this to my attention.  This has long plagued me too!!  But I had no idea until 20mins ago that this was an actual bug effecting only DP users.  I incorrectly thought this was just the very nature of latency and buffer settings etc and all DAWs experienced this.  Just confirmed with Logic and Cubase using friends that those DAWs will compensate for higher buffers when live recording MIDI.  DP needs to fix this RIGHT AWAY.  I'm going to ping the hell out of this to the dev team.  Thanks again for bringing to my attention!

Posted on Sat, Mar 19 2022 20:40
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

Originally Posted by: evang42 Go to Quoted Post

WOW.  mind absolutely BLOWN.  Thank you so much for bringing this to my attention.  This has long plagued me too!!  But I had no idea until 20mins ago that this was an actual bug effecting only DP users.  I incorrectly thought this was just the very nature of latency and buffer settings etc and all DAWs experienced this.  Just confirmed with Logic and Cubase using friends that those DAWs will compensate for higher buffers when live recording MIDI.  DP needs to fix this RIGHT AWAY.  I'm going to ping the hell out of this to the dev team.  Thanks again for bringing to my attention!

Wait, what? I've used DP exclusively for composing with MIDI for years. Are you saying other DAWs have found a way to eliminate buffer latency when recording MIDI (IOW the interval between the time a key/pad is pressed and the VI sounds)? If so, I missed a memo, because it was not always so. Back in the oughts, when I used Logic, latency was definitely a thing.  

Posted on Sat, Mar 19 2022 21:00
by evang42
Joined on Fri, Apr 08 2011, Posts 30

Originally Posted by: Dan Wool Go to Quoted Post

Originally Posted by: evang42 Go to Quoted Post

WOW.  mind absolutely BLOWN.  Thank you so much for bringing this to my attention.  This has long plagued me too!!  But I had no idea until 20mins ago that this was an actual bug effecting only DP users.  I incorrectly thought this was just the very nature of latency and buffer settings etc and all DAWs experienced this.  Just confirmed with Logic and Cubase using friends that those DAWs will compensate for higher buffers when live recording MIDI.  DP needs to fix this RIGHT AWAY.  I'm going to ping the hell out of this to the dev team.  Thanks again for bringing to my attention!

Wait, what? I've used DP exclusively for composing with MIDI for years. Are you saying other DAWs have found a way to eliminate buffer latency when recording MIDI (IOW the interval between the time a key/pad is pressed and the VI sounds)? If so, I missed a memo, because it was not always so. Back in the oughts, when I used Logic, latency was definitely a thing.  

No no, theres a very clear distinction that needs to be made here.  I'm not saying other daws found a way to eliminate the buffer latency of - the interval between the time a key/pad is pressed to the time you hear the sound - that would basically be impossible.  What Im saying is other DAWs have a way to place the actual MIDI notes that you play in real time down in the midi track at the EXACT time you physically touch the keyboard, regardless of buffer.  Pretend your speakers are muted for a sec, this whole thing has nothing to do with actual sound.  Set up a metronome on your iphone or something.  Now, If you record arm a track in Logic or Cubase, play a midi passage to the click at a buffer of 64, the resulting recorded midi will be in time to the same extent that if you then change your buffer to 2048, record arm the track and play to the same click.  That is currently NOT the case w DP.  DP's buffer settings are having a negative delay/impact on where the actual midi notes get put while recording and it gets worse the higher the buffer goes.

Posted on Sat, Mar 19 2022 22:24
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

Got it. What you describe is what I always assumed. Yeah, I guess other DAWs would need to have implemented time-travel to eliminate live buffer latency!  

So higher buffers causing "negative-delay" in DP is the operative here. That is new info for me. By keeping my buffers at 512 I guess I've never noticed the discrepancy - mainly, probably because I'm not much of a player. At 512 that negative-lag is the least of my problems.  

But is this a VEP issue, evang42? Or do all VI exhibit this in DP? 

Posted on Sun, Mar 20 2022 01:14
by dmansfield
Joined on Tue, Sep 21 2004, New Jersey, Posts 17

This problem exists in DP with any VI. For example, we tested it with a basic MoTU VI, in the actual sequence on the host computer (no VEPro, Bidule, etc.), and it is still a problem. Take the VI out, and record into a MIDI track that's not connected to anything, and MIDI is recorded correctly, regardless of buffer setting. By the way, I can set Pro Tools up with the same parameters and it does not exhibit this problem. All modern DAWs compute where to record the MIDI note regardless of buffer size, latency, etc..

Finally, this problem does not exist with AMD PCs or M1 Macs. You can record MIDI with a VI and a 1048 buffer on a M1 MacBook Pro and the MIDI notes will be recorded exactly as you played them. Matt Batson and I did a lot of testing on this, and the engineers confirmed it.

Posted on Sun, Mar 20 2022 07:10
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

Originally Posted by: dmansfield Go to Quoted Post

This problem exists in DP with any VI. For example, we tested it with a basic MoTU VI, in the actual sequence on the host computer (no VEPro, Bidule, etc.), and it is still a problem. Take the VI out, and record into a MIDI track that's not connected to anything, and MIDI is recorded correctly, regardless of buffer setting. By the way, I can set Pro Tools up with the same parameters and it does not exhibit this problem. All modern DAWs compute where to record the MIDI note regardless of buffer size, latency, etc..

Finally, this problem does not exist with AMD PCs or M1 Macs. You can record MIDI with a VI and a 1048 buffer on a M1 MacBook Pro and the MIDI notes will be recorded exactly as you played them. Matt Batson and I did a lot of testing on this, and the engineers confirmed it.

Now I see. Hence, you're hoping for an AS native version of VE Pro ASAP, so you can make the switch to an M1 Mac. As I mentioned the AU version of VEP works fine with DP. I haven't investigated, but presumably it wouldn't have the timing bug. It's cumbersome, and not nearly as convenient as the MAS version but it might work as a stopgap for you, so you could get going on an M1 Mac, until there's a native MAS VEP. 

Posted on Mon, Mar 21 2022 18:00
by fatis12_24918
Joined on Sat, Dec 16 2006, Posts 344

 

...and now iLok is done... hopefully beside debugging it (just last little inconsistencies, the overall job was quite brilliant, well done VSL) a VEP7 universal macOS will be released... 

Posted on Mon, Mar 21 2022 18:11
by dmansfield
Joined on Tue, Sep 21 2004, New Jersey, Posts 17

From your mouth to God's ear 

Posted on Thu, Mar 24 2022 19:52
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

Originally Posted by: fatis12_24918 Go to Quoted Post
...and now iLok is done... hopefully beside debugging it (just last little inconsistencies, the overall job was quite brilliant, well done VSL) a VEP7 universal macOS will be released...
I reckon it could be awhile still. Creating a UB version of VEP must be a lot more complex than creating one for an audio plugin....many of which are still in development almost 18 months after the arrival of M1

Posted on Fri, Mar 25 2022 14:46
by santamorphic
Joined on Wed, Jun 04 2008, Posts 189

I'm assuming VEP7 would stay on Rosetta and VEP8 would be the software to make native. Seems like a heck of a lot of work to have to do it fro VEP7 when VEP8 is probably on the horizon. It would also be a little nudge to encourage an upgrade :-)

DAW Client: Cubase 12, Nuendo 12, Ableton Live 11, iMac Pro 10/20 Core, 64GB RAM
VEPro Server: Windows 10 Pro x64, 4/8 core, 256GB RAM, SSDs
Posted on Fri, Mar 25 2022 19:06
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35
Originally Posted by: santamorphic Go to Quoted Post
I'm assuming VEP7 would stay on Rosetta and VEP8 would be the software to make native. Seems like a heck of a lot of work to have to do it fro VEP7 when VEP8 is probably on the horizon. It would also be a little nudge to encourage an upgrade :-)
We’ll, I sure don’t like this idea. I’d like it better if VEP worked with Rosetta for me, but it does not. The MAS version of the plugin doesn’t work under Rosetta. If they updated the MAS plugin to work with Rosetta I’d be okay waiting a bit for the privilege of paying for a UB compatible upgrade.
Posted on Sat, Mar 26 2022 23:25
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

...along these lines, I'm reading that MacOS 12.3 seems to have fixed a lot of plugins that weren't working in Monterey. I'm holding back updating for a couple of reasons, but I'm curious if, by some miracle, the update makes the MAS plugin usable on M1 Macs.

Posted on Sat, Apr 02 2022 19:31
by Michael Canavan
Joined on Sun, Jun 25 2017, Posts 25

Originally Posted by: Dan Wool Go to Quoted Post

...along these lines, I'm reading that MacOS 12.3 seems to have fixed a lot of plugins that weren't working in Monterey. I'm holding back updating for a couple of reasons, but I'm curious if, by some miracle, the update makes the MAS plugin usable on M1 Macs.

It does not. Basically unless you're running Cubase or Bitwig, VSTs don't run in Rosetta. Bitwig sandboxes the VSTs so they run in Rosetta in Bitwig native. As Far as MAS is concerned MOTU would have to implement this themselves and I would bet they decided that the 8-12 companies that make MAS plug ins would be coming to native soon enough. 

  So yeah, I miss the MAS VEP plug in, and hope that Vienna get going on this soon enough. 

Posted on Thu, Apr 07 2022 07:47
by evang42
Joined on Fri, Apr 08 2011, Posts 30

To be clear, DP11 itself is fully AS native, but as you point out VEP MAS won't work in Rosetta.  I actually spoke to someone at MOTU about this and they said that Rosetta could only translate AU plugins (and not MAS) so unfortunately it's up to VSL to go AS native and deliver a MAS version to get this working again and it seems there's nothing more MOTU can do.  

Posted on Thu, Apr 07 2022 07:52
by evang42
Joined on Fri, Apr 08 2011, Posts 30

Originally Posted by: dmansfield Go to Quoted Post

Originally Posted by: evang42 Go to Quoted Post
I also use DP11 and am a beta tester and can report issues directly to the MOTU devs. What are the steps/circumstances that cause the midi timing bug w DP11/VEP you’ve found on intel???? This needs to be fixed asap regardless of AS transition.

With higher buffer settings (e.g. 1048), if you are routing a MIDI track to a VI (any VI, even a MoTU VI), MIDI records about 60 ms late. Also "Fine-tune Audio I/O Timing" does not work on these machines. The only workaround is nudging MIDI back into place after recording. I've been working extensively with Matt Batson in tech support, and the engineering team has been made aware of the bug. They have reproduced it on 2 different Intel machines, and confirmed that it does not happen on their AMD PC or their M1 Mac Mini. I don't know that it is a priority for them though.

There is an extensive thread about this on MoTUNation, "MIDI recording late", where I document what's going on in detail (though you may have to wade through some off-topic or uninformed posts to see it). If you have access to the engineering team and can help nudge them toward supporting the older Intel machines, that would be great!

Hi dmansfield!  I found an answer/fix for you on this!  In DP's preferences, go to midi options and UN-check the box called "Sync Recorded Midi to Patch Thru".  This SHOULD totally fix your issue (on ANY machine, apple silicon, intel, pc etc).  It's a setting that tells DP to place midi either 1) exactly where you PLAYED/intended it (un-checked) OR 2) where you would normally HEAR that midi (which of course is dependent on buffer size/latency).

Posted on Thu, Apr 07 2022 19:02
by Dan Wool
Joined on Wed, Jun 17 2015, San Francisco, Posts 35

Originally Posted by: evang42 Go to Quoted Post

To be clear, DP11 itself is fully AS native, but as you point out VEP MAS won't work in Rosetta.  I actually spoke to someone at MOTU about this and they said that Rosetta could only translate AU plugins (and not MAS) so unfortunately it's up to VSL to go AS native and deliver a MAS version to get this working again and it seems there's nothing more MOTU can do.  

Okay. Not great news, but this kind of clarity is useful. 

I know developers rarely divulge timelines, but it sure would be nice to get a general sense of how long it will be until there's a UB VE-Pro and MAS plugin (weeks? months? years?) - I need to restart several huge projects from 2019 that rely on VEP with MAS. Converting them to AU will be a time-consuming pain. 

10 Pages<12345>»
You cannot post new threads in this forum.
You cannot reply to threads in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.