Delay compensation issue

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
Post Reply
linusminus
Posts: 22
Joined: 21 Oct 2020

21 Oct 2020

I've been using more and more third-party plugins and recently noticed on my vocal tracks that automations don't respond properly. This is also the case with anything related to time, such as Synchronous, PanMan, LFO Tool etc. It's generally not noticeable when it's just a few samples delay, but when you start stacking plugins with lots of look-ahead, it suddenly becomes very apparent. What's scary is that even the mix channel's volume fader is affected. Basically anything that comes after "heavy" plugins will be timed wrong; I tried automating the gain using kHs Gain put last in the chain but that didn't respond correctly in the sequencer either.

Delay compensation is complex and most DAWs have some strange things going on, but I feel like you should at least expect individual channels--there's no interaction between tracks with various delays--to work. When I can't even draw sidechain on my own and expect it to be in time, something's not right.

Example #1 (automating mute):

Example #2 (LFO/envelope with Synchronous):
Last edited by linusminus on 21 Oct 2020, edited 1 time in total.

Le Boeuf
Posts: 91
Joined: 16 Mar 2017
Location: Copenhagen
Contact:

21 Oct 2020

Following here.
Ive had The same issue with pitch bends

User avatar
Loque
Moderator
Posts: 11236
Joined: 28 Dec 2015

21 Oct 2020

AFAIK the problem comes from the developers of the plugins, because they do not report any or the wrong delay. Often they just forget things to report too, like an fx on a synth.

In Reason you may have problems with complex routings, especially with loopbacks and high buffer size.
Reason12, Win10

linusminus
Posts: 22
Joined: 21 Oct 2020

21 Oct 2020

Le Boeuf wrote:
21 Oct 2020
Following here.
Ive had The same issue with pitch bends
Yes, pitch bends might also be affected. All automation that takes place after the delay is introduced will be off. It's quite annoying because after using Saturn 2 (linear phased turned on), Pro-DS and soothe2 on vocals, there's enough delay introduced that isn't compensated for for it to be noticeable when using creative effects afterwards. I can't even sidechain it at that point using "horizontal" plugins like LFO Tool because the signal will be way off. Heck, not even manually drawn sidechain will be aligned correctly.
Last edited by linusminus on 21 Oct 2020, edited 1 time in total.

linusminus
Posts: 22
Joined: 21 Oct 2020

21 Oct 2020

Loque wrote:
21 Oct 2020
AFAIK the problem comes from the developers of the plugins, because they do not report any or the wrong delay. Often they just forget things to report too, like an fx on a synth.

In Reason you may have problems with complex routings, especially with loopbacks and high buffer size.
If there were complex routings, I'd expect issues :D Not, however, when it's a stack of plugins on a single channel. It's nerve-wrecking that it's off even at channel level (like volume and mute automations) :?

linusminus
Posts: 22
Joined: 21 Oct 2020

21 Oct 2020

Just need to point out that I don't normally stack limiters :lol: It's for demonstration purposes! The issue is usually not to this extent, but it's an issue nevertheless. Throw a trance-gate, auto panner or other rhythmic effect after a few plugins using look-ahead and it will be noticeable. With complex routing, I'd understand if there were issues, but this...

User avatar
Boombastix
Competition Winner
Posts: 1929
Joined: 18 May 2018
Location: Bay Area, CA

21 Oct 2020

On the mixer you can see total delay comp for each channel. Helps with debugging the situation. There is a pass through VST that creates delay comp but you can also adjust it on the back of the mix channel.

And there is a RE that let's you ping the delay time so you know exactly how much to compensate for. Keep in mind that some plugins have a lag in the automation to avoid digital zipper noise. You need to always offset manually for that.

I have reported the delay compensation bugs in Reason when recording from one channel to the next. This was for v10 and 18mo ago, still no fix and it exists in v11 too.

Your problem is probably a plugin, but if it is a VST it could be Reason's handling of VSTs as it is known to not follow typical VST standards.
10% off at Waves with link: https://www.waves.com/r/6gh2b0
Disclaimer - I get 10% as well.

linusminus
Posts: 22
Joined: 21 Oct 2020

21 Oct 2020

Boombastix wrote:
21 Oct 2020
On the mixer you can see total delay comp for each channel. Helps with debugging the situation. There is a pass through VST that creates delay comp but you can also adjust it on the back of the mix channel.

And there is a RE that let's you ping the delay time so you know exactly how much to compensate for. Keep in mind that some plugins have a lag in the automation to avoid digital zipper noise. You need to always offset manually for that.

I have reported the delay compensation bugs in Reason when recording from one channel to the next. This was for v10 and 18mo ago, still no fix and it exists in v11 too.

Your problem is probably a plugin, but if it is a VST it could be Reason's handling of VSTs as it is known to not follow typical VST standards.
Thank you! I'll try to deal with it manually for now, though I wonder how you'd go on about doing it with something like Synchronous or LFO Tool placed last in the chain; you can't really do anything to compensate there, as far as I know. Tried the Time Slider extension which could only delay it further but not the other way around. I've tried several plugins that adds delay from FabFilter and they all seem to cause the same problem.

User avatar
Boombastix
Competition Winner
Posts: 1929
Joined: 18 May 2018
Location: Bay Area, CA

21 Oct 2020

This is what you can do. https://www.voxengo.com/product/latencydelay/

But some cases can never be supported if you break the Compensating Rules.
And check what Pro-L reports on the mix channel. What number for DLC do you see?
If that is wrong then Pro-L has an issue or the Reason VST holder.

Also note, the Reason sequencer play-head is NOT compensated for and it WILL lag considerably when having a large latency, so unfortunately, I do not think they ever will fix that (based on my conversation with them). Super annoying if you still are composing/writing with large latency plugins as the offset can easily be 1/16th. Not sure if other DAW do this or not.
10% off at Waves with link: https://www.waves.com/r/6gh2b0
Disclaimer - I get 10% as well.

linusminus
Posts: 22
Joined: 21 Oct 2020

21 Oct 2020

One instance of Pro-L 2 creates 2875 samples delay and it's correctly reported according to the plugin you linked. Again, there's never a problem with the "sound source" even after stacking multiple instances; it always arrives on time, as you can see in the posted videos.

However, things like rhythm effects (LFO/envelopes) and automations of anything appearing after these plugins aren't properly aligned. It's like they operate on the un-delayed timeline even though the mix channel itself is delayed. Do you understand what I'm getting at? Check video #1 for example: the piano itself is on tempo in both versions, but the mute automation operates as if there was no delay. Same goes for video #2, where the Synchronous device operates as if there was no delay added even though it's placed after Pro-L and the delay is correctly reported.

The Latency Delay plugin can't seem to help me with this as the delay is reported correctly, just ignored. It feels like if you place, say Synchronous after a plugin that introduces delay, it should operate based on that and not as if no delay was introduced. This is how Reason treats it, I think.

Maybe I'm misunderstanding it all. None of my friends in other DAWs have these issues.

User avatar
Boombastix
Competition Winner
Posts: 1929
Joined: 18 May 2018
Location: Bay Area, CA

21 Oct 2020

If you noticed, I mentioned the sequencer play head gets out of sync at high latency comp.

Maybe this includes the timing that is sent to the track units that use this info.

If so then this is a major bug and not just a nuisance with the moving player head line being off.

You should be able to replicate this with the Ozone RE or Neptune. And then have others confirm it.
10% off at Waves with link: https://www.waves.com/r/6gh2b0
Disclaimer - I get 10% as well.

linusminus
Posts: 22
Joined: 21 Oct 2020

22 Oct 2020

So I tested the same thing with Neptune and Ozone Maximizer using a rhythm effect in the form of Synchronous and mute automations. What's weird is that while Synchronous is timed all wrong when delay has been introduced, automations seem to align properly--unlike my previous examples using VST plugins. Basically, with REs and stock units introducing delay you can at least expect automations on following devices to be timed correctly, but rhythm effects (such as LFO-based sidechain) will still be off. I also tried automating Pro-Q 3 instead of kHs Gain and that also works properly.



How odd.
Last edited by linusminus on 22 Oct 2020, edited 2 times in total.

User avatar
Boombastix
Competition Winner
Posts: 1929
Joined: 18 May 2018
Location: Bay Area, CA

22 Oct 2020

Yeah, I think they screwed up Delay Comp when the re-wrote the engine in v10.3 to fix the buffer size issue. What is mostly odd is that they haven't fixed it, since 18mo I believe...
10% off at Waves with link: https://www.waves.com/r/6gh2b0
Disclaimer - I get 10% as well.

linusminus
Posts: 22
Joined: 21 Oct 2020

22 Oct 2020

That's not good at all. I make music full-time and rely on Reason to work properly. From now on I'll bounce tracks that introduce a good amount of delay before I automate things or add effects such as PanMan afterwards. It's usually not noticeable when it's say 100 samples--I think RC-20 is around there--but when it starts approaching 1000, I can't take any risks.

User avatar
Boombastix
Competition Winner
Posts: 1929
Joined: 18 May 2018
Location: Bay Area, CA

22 Oct 2020

Send in a bug report.
10% off at Waves with link: https://www.waves.com/r/6gh2b0
Disclaimer - I get 10% as well.

linusminus
Posts: 22
Joined: 21 Oct 2020

22 Oct 2020

I've done so :thumbs_up:

exxx
Posts: 154
Joined: 12 Sep 2016

22 Oct 2020

Was the strangeness I felt in my work result due to delay compensation...

linusminus
Posts: 22
Joined: 21 Oct 2020

22 Oct 2020

To summarize after new findings

Although the delay itself is reported correctly, rhythmic effects such as Synchronous, PanMan or LFO Tool still operate based on the un-delayed sequencer position when placed after delay-inducing plugins or rack devices.

Similarly, when using VST plugins, any sort of automation will be misaligned. This, however, is not the case when the delay comes from stock devices or rack extensions, where automations will align properly but rhythm effects will not. It's quite peculiar, isn't it?

Post Reply
  • Information
  • Who is online

    Users browsing this forum: alterEe and 13 guests