Drambo + Pivot = why am I staring longingly at the DN2?
@Ess one thing that I’m really noticing in Drambo is the importance of resizable UI for good AUv3 guest behavior. I know that you’ve put a huge amount of time and thought into Pivot’s UX, but guests that don’t scale are really challenging (for me), and I suspect that it might affect adoption. I think Pivot’s good enough that it will be extremely successful no matter what, but I think it’s worth considering how it’s going to work in a host if you decide to go down that path.
Hey everyone ![]()
Since the latest Ableton update, I can’t drag samples into Slate - Opal just disappears.
Is anyone else having the same problem?
I believe I’ve encountered the same issue. I attempt to drag a sample from either Ableton or Finder, but none of the samples load. I’ve tried different file formats, but I thought it was just me.
If it’s disappearing it was a bug in the beta, if that’s it they know about it and I presume are working on it. I’d assumed that would be fixed before release but maybe not. If you have two screens you can put opal on a different screen or drag the live app over and opal will be behind it. There was a bug with things loosing focus, it was vsts and m4l but they may have only got vsts fixed at the moment.
To be clear I’m talking about Ableton not fors, it was happening with lots of m4l devices where you’d drag a sample from the browser.
This happens to me when I drag from finder, removing focus from Ableton ‘hides’ Opal, so I can’t drag to it.
I can drag to it fine from within Ableton though.
Unfortunately, I can’t load a sample into Opal either - dragging from both finder and Ableton doesn’t work for me. I’m pretty sure it’s an issue caused by the latest Ableton update.
Thanks a lot, that makes things clearer. Hopefully it gets fixed soon. I guess I should turn off auto-updates in Ableton and pay more attention to the beta reports.
Ah I see that 12.3 has finally dropped, I haven’t updated yet! So I’ll be on the previous version.
Probably a Max thing as the bundled version has been updated with 12.3
Yeah it’s a good idea to stay on previous version before they make a fix
I don’t think I was a bug a ton of people were having so I guess it was oddly specific to certain systems, is it just that the window disappears or samples can’t be loaded at all? I could load samples once I put things on another monitor or slide windows around. Sometimes Max won’t like a certain file type, if you bounce the clip in live can you load the bounced clip into opal?
Another Tela piece
Been getting dropouts for Pivot and checked back what’s going on with auval. Looks like Pivot is doing a ton of things under the hood that aren’t realtime safe. (same for Sala)
Can be easily reproduced with “auval -v aumu FrPl FORS -real-time-safety”.
Dozens of:
Realtime-safety violation:
libsystem_pthread.dylib`pthread_mutex_lock
libc++.1.dylib`std::__1::recursive_mutex::lock()+0x10
Pivot`0x0000000105730da0+0x2dc
Pivot`0x00000001057361a8+0x38
AudioToolboxCore`AudioUnitRender+0x1b0
auvaltool`0x0000000104c825e4+0xd8c
auvaltool`0x0000000104c84794+0x1af8
auvaltool`0x0000000104c8ad10+0x64c
auvaltool`0x0000000104c78d00+0xc4c
auvaltool`0x0000000104c78a5c+0x1a0
dyld`start+0x17bc
...
Real-time Safe Test FAILED!
Kinda dissapointing to see that the Fors team doesn’t have this locked down.
Oh damn, I’ll check this out asap!
FWIW I haven’t gotten any reports of dropouts, and all plugins run through max strictness in pluginval, but it could be something that hasn’t been caught there of course.
Appreciate you having a look!
Of course! Thanks for highlighting this.
So, there was one violation on my end where a vector wasn’t properly handled in Pivot-Lite (and Sala), and this is actually something I’ve already fixed in Pivot but unfortunately missed to correct in Pivot-Lite as well. I’ll build a new version with the fix shortly.
The other violations actually come from JUCE, which makes more sense to me – I don’t do any locks or mutexes my own code specifically to avoid these kind of issues and was surprised/confused to see that in the trace!
Will dig into this a bit more but even building a “blank” AU with JUCE exhibits these issues on my end, along with a bunch of other plugins from manufacturers authored in JUCE. Will keep digging/report to them either way.
(However I should also point out that auval is extremely strict and points out things that are technically violations but might not pose real-world issues)
If you have any more details or examples of the drop-outs together with system information that’d be very interesting to see since I’ve really gotten no other reports about this. Feel free to shoot over an email to hi@fors.fm
shallow symptom report: with Sala, CPU usage goes up the roof after going idle out of a long reverb. Hinting to a leak or race condition. When the device is removed, replaced all back to normal. Most prevalent it doesnt happen so quick with short reverb value (below 50%). I will feedback more precise when i am in studio using it again - soon.
Hmm ok, interesting! The decay/size affecting this is super odd. Can’t repro on my end though, just behaves like expected. Regardless will of course look into it. (It’s odd that I haven’t heard of this issue before, but maybe people are less likely to report issues on free plugins??)
And yeah, that’d be very helpful! Again would be best to send an email since that’s where I usually do support stuff. (hi@fors.fm)
EDIT: actually I have a very good hunch why this could happen and why it would only happen on a small set of systems. Should be an easy fix and I’ll do that along with whatever issue that I can adress in auval pops up. ![]()
Thanks for reporting, it sucks that there are issues but I’m glad I can improve things!