Logic Pro Help

That's a good point, since the audio isn't being read from the aux track, the aux track can't do anything with the signal until it's received it. So Logic needs to instead move the audio source (in this case, midi) back in time to compensate for the plugin induced latency. With a multi-out instrument though this is very difficult as Logic doesn't know what midi will produce sound from what output. Still, there has to be a clever solution for this, and it certainly isn't to move audio on sidechain receiving tracks out of time. That's silly.

the side chaining is a problem, no argument. You're not the first to complain about it. But again, I want to understand exactly what are the conditions under which it fails so that we can determine whether its something that is even possible to solve and then we can all submit bug reports if we think so.

think about this. Lets say you have an AUX channel that is getting its live audio from somewhere. Another audio channel is getting its audio from a region. They both have latent plugins. Ok, so the audio region is queued early to make the audio channel sound earlier. Fine so far. Let's assume the audio channel is feeding itself through a send into an AUX channel that is adding more latency. OK, so the engine needs to delay the audio channel in order to compensate for the extra delay in the AUX channel, but in so doing, wil it also be now sending its own audio even later to the AUX channel through the send? Same thing kind of applies to side chains. And that is just a simple example, routing can get a lot more complicated...eventually...the PDC simply can't resolve it. You can't always get blood from a stone.

But in any case, if there are simple scenarios, and looking at them under a microscope we can identify that LPX could easily resolve the PDC, then we should identify that and report to to Apple. There are a lot of complaints about PDC on this forum and other places, and IMHO, a high percentage of them turn out to be user error or lack of understanding. So if we send a generic bug report to Apple that says more or less "Logic is worst DAW ever the PDC never works right". How much attention will that get from devs? IMHO not much. IF we give them very very specific information about where its failing and we have thought it through completely and think it could be resolved, then we can give them something they can pay attention to and maybe fix.

hope that makes sense what I'm saying

Instruments could be improved in LogicPro by not using AUX channels for them. They should have a dedicated Instrument extension which are technically not AUX channels, but just multiple instrument outputs on an instrument channel...and then all instrument channels would cause the input midi to be queued earlier as appropriate for any PDC on the instrument channel.

They could still probably still do that under the covers even though they call it an AUX channel, maybe a special kind of AUX channel since its not getting its input from a general bus, its getting its input specifically from the output of an instrument plugin.. So totally doable. I would have thought it was doing that already actually... But I haven't really tested for that to see exactly what is happening.

I think the automation PDC bugs are definitely a bigger deal IMHO and I can't think of any reason why they haven't fixed that other then people just aren't being assigned to the task somewhere deep inside Apple.

So looking at your project....

I agree, there is nothing that tricky about this routing and there is no reason I can think of why it shouldn't play in sync...

Now to try to analyze it.. what is going on..?

Theoretically, regardless of whether you have side chaining activated; when the EnVerb latency is added, the two audio channels with bass recorded on them should be delayed by the same amount, play in sync by the same amount as what Enverb is putting out. If they are both playing perfectly we should hear no bass (canceled out).

So if I remove the gate from the bass channel, that is exactly what happens. two channels cancel out. Put the gate back, and doesn't work.

I have no solution I can think of...its not handling it right.

the question is why. One thought I have is that LogicPro is confused by Instrument-AUX channels in some way..handling them like an AUX when it should be handling like an instrument...or visa versa...and meanwhile the side chain on the plugin is assuming its coming from an AUX.. something like that. Not entirely sure... But I agree its a simple case that Apple should have handled better!

Dewdman,
Thank you for the thought out reply. I'm with you man. That's why I made that example project that I posted. It's the simplest circumstance under which I could reproduce the exact bug I'm referring to, and I think it illustrates the problem more than adequately. Although it doesn't illustrate the other PDC issues mentioned in this thread.

...the engine needs to delay the audio channel in order to compensate for the extra delay in the AUX channel, but in so doing, wil it also be now sending its own audio even later to the AUX channel through the send?

Right, so that means for every downstream point where the audio signal is split (i.e. sent to an actual output), Logic would have to duplicate the source and entire signal chain thus far, adding a different latency compensation offset according to the latency added prior to that particular juncture. Okay, that's totally untenable.

Instruments could be improved in LogicPro by not using AUX channels for them. They should have a dedicated Instrument extension which are technically not AUX channels, but just multiple instrument outputs on an instrument channel...and then all instrument channels would cause the input midi to be queued earlier as appropriate for any PDC on the instrument channel.

Yeah, that would certainly help but I believe it would require plugin developers to code for that functionality. As in, the plugin would have to tell Logic what MIDI is triggering audio on which outputs in order for Logic to be able to properly move/retime the midi to account for downstream latency.

All of this aside, I am not a DAW designer, I am a musician. As fun a distraction as it is to ponder why it works or doesn't work, it's not our job to figure out how to make it work properly. It's my job to make music, and right now the DAW is hindering my ability to use what I've come to understand is a pretty common workflow. I've done as much as I can to illuminate the issue, short of submitting a bug report I suppose. So I guess that's the next step. Unfortunately it doesn't look like they let you attach a file though, which would be the quickest way to illustrate the issue. That, and they ask you to fill out a survey longer than an online job application.

Genuinely curious, do Apple engineers ever check these forums?

Dewdman42 wrote:

So looking at your project....

I agree, there is nothing that tricky about this routing and there is no reason I can think of why it shouldn't play in sync...

Now to try to analyze it.. what is going on..?

Theoretically, regardless of whether you have side chaining activated; when the EnVerb latency is added, the two audio channels with bass recorded on them should be delayed by the same amount, play in sync by the same amount as what Enverb is putting out. If they are both playing perfectly we should hear no bass (canceled out).

So if I remove the gate from the bass channel, that is exactly what happens. two channels cancel out. Put the gate back, and doesn't work.

I have no solution I can think of...its not handling it right.

the question is why. One thought I have is that LogicPro is confused by Instrument-AUX channels in some way..handling them like an AUX when it should be handling like an instrument...or visa versa...and meanwhile the side chain on the plugin is assuming its coming from an AUX.. something like that. Not entirely sure... But I agree its a simple case that Apple should have handled better!

Responded before I saw your post.
Yes, for sure it's weird. Neither bass track should be delayed or delay compensated at all. The one bass channel's only contact with the latent signal is via a sidechained gate plugin. No reason to change it's timing.

Try this for extra weirdness though, add an EnVerb to the main ultrabeat track and the bass tracks cancel again.

Kiriel wrote:

Right, so that means for every downstream point where the audio signal is split (i.e. sent to an actual output), Logic would have to duplicate the source and entire signal chain thus far, adding a different latency compensation offset according to the latency added prior to that particular juncture. Okay, that's totally untenable.

Well my only point is that AT SOME POINT, the routing complications could make it such that PDC simply can't work it out. I don't think that applies to your example project at all. I think there are certain routing configurations, if we sat here long enough we could think of one or two, that could not possibly play in sync after latency induced in the wrong places, but I also think 99% will never see that situation

Instruments could be improved in LogicPro by not using AUX channels for them. They should have a dedicated Instrument extension which are technically not AUX channels, but just multiple instrument outputs on an instrument channel...and then all instrument channels would cause the input midi to be queued earlier as appropriate for any PDC on the instrument channel.

Yeah, that would certainly help but I believe it would require plugin developers to code for that functionality. As in, the plugin would have to tell Logic what MIDI is triggering audio on which outputs in order for Logic to be able to properly move/retime the midi to account for downstream latency.

No actually plugin developers wouldn't have to do anything other then report latency as they always should be doing. A single instrument plugin reports its latency already and the instrument channel does the right thing. the instrument channel just needs to do the right thing with multi-outs. But I'm actually pretty sure the current solution is already doing that. What i think is maybe a little confused is the side chain input into the plugin. but who knows.

All of this aside, I am not a DAW designer, I am a musician. As fun a distraction as it is to ponder why it works or doesn't work, it's not our job to figure out how to make it work properly. It's my job to make music, and right now the DAW is hindering my ability to use what I've come to understand is a pretty common workflow. I've done as much as I can to illuminate the issue, short of submitting a bug report I suppose. So I guess that's the next step. Unfortunately it doesn't look like they let you attach a file though, which would be the quickest way to illustrate the issue. That, and they ask you to fill out a survey longer than an online job application.

if you submitted a bug report, they will often times contact you and ask for a project showing the problem. So hopefully they will. You submitted an official bug report right?

Kiriel wrote:

Responded before I saw your post.
Yes, for sure it's weird. Neither bass track should be delayed or delay compensated at all. The one bass channel's only contact with the latent signal is via a sidechained gate plugin. No reason to change it's timing.

Try this for extra weirdness though, add an EnVerb to the main ultrabeat track and the bass tracks cancel again.

well actually no, you want the bass track to compensate for the delay in Enverb. It should be doing that, thusly lining up itself with the delayed enverb output. that should feed into the sidechain in sync because the bass will have been delayed along with it. Both of those audio tracks should be PDC delayed to match the delay of Enverb on the AUX. And cancel themselves out. If you remove the gate plugin completely, that is exactly what happens.

And I also tried putting enverb directly on the instrument channel like you did and then the proper delay compensation happens. But in that case, Logic Pro feeds the UB region early so that the audio comes out not-late and thus lines up with the bass track that is waiting for side chain. So it works.

Its the AUX channel that has to be late, because its an AUX, I guess...and then the bass channels should ALL be delayed to match that..regardless of the side chain actually. but it seems like in this case, one of the two bass tracks is being delayed to match and the the other one isn't. My money would be on it being because its an instrument-AUX and LogicPro is supposed to be handling it differently but is missing some case.

So, if I'm getting what you're saying, PDC doesn't only play latent tracks early, it also shifts non-latent tracks to play late, in order to match??

EXACTLY. That is how AUX channels are supposed to be handled. AUDIO/INST channels have their regions shifted early. AUX and OUTPUT channels cause ALL OTHER CHANNELS to be shifted later to line up with their delay.

More food for thought.

If I add a GATE to the second bass track it doesn't work but, if I copy the GATE from the first bass track to the second bass track, everything works.


So internally the signal path is the same because of the copy, this might be the key to get things working.

More food for thought.

Thanks for making that vid. The reason that's happening is because any track with a plugin sidechained to an AUX with latency inducing plugins on it will be pushed out of time. So by copying the gate, you're just adding the same messed up timing to the phase-swapped bass track, so now they're both messed up. But cancelling!

By the way there's an existing topic related to this issue in the Bug forum : viewtopic.php?f=42&t=135038

great thread.

Seems like the work around for now would be to bounce the instrument-AUX channel to a track, then change that track to audio channel, then use it to feed side chain as a non-live-bounced instrument. A bit of a headache, but at least the production can move on until Apple fixes this horrendous bug that has apparently been plaguing us for at least 2-3 years...

The automation PDC problem I am not familiar with the exact problem or if there is a work around?

10.5.1 just released, haven't had a chance to see if this is fixed yet. Anyone else?

Still no fix for the completely messed up Plug-in delay compensation in Logic Pro X.

I'm not getting tired to ask all Logic Pro X 10.5 users to really report this long time issue of Logic Pro:



Just take 2 minutes and send a bug report to Apple about the missing plug-in delay compensation for automation, side chain signals and beat synced plug-ins. Thank you for your help!
https://www.apple.com/feedback/logic-pro.html

Is this PDC thing still going on in 10.6?

According to the other thread it still hasn't been fixed. Renders so many plugins useless.

rfpm wrote:

According to the other thread it still hasn't been fixed. Renders so many plugins useless.

I wonder why the Logic team can't get a handle on this? Is this an issue in Cubase and other DAW's?

Petition is here: http://chng.it/bfczxYTh9g
We need to post it in any ADC / PDC-related thread at any Logic forum.

Saw you post on Reddit too, this is what we need! These bugs are well overdue for fixing!