Plugin Latency Compensation PDC in Reaper?

[wilbur74]

Thought I should share this as although I've been using Reaper for nearly two years now, and this discovery has transformed the way I work with Reaper, improved the sound quality and and removed the crashes I used to have every 30mins.

Basically I found that Reaper's automatic plugin delay compensation (well, to be more acurate certain plugins) were not working as I expected.

I found this by creating a new session, loading a single audio file and duplicating it so it plays on two tracks. I then inserted a plugin , e.g. Compression. If the plugin delay is incorrect, there will be an obvious phasing of the two tracks.

Certain plugins that I was using were clearly not working! Digitalfishphone's Dominion, Isotope's Ozone, and a handful of others were introducing delay and hence phase issues.

By removing these from my mixes, or only using them on the master bus it's made a massive difference.

Hope this helps some Reaper beginners out there !

Cheers,
Joe

[pipelineaudio]

even after you hit stop and restarted playback they were still delayed?

[wilbur74]

yes, even after stopping and restarting playback. I just re-tested to be sure!

have you not heard of this before ?

[LOSER]

Originally Posted by wilbur74

yes, even after stopping and restarting playback. I just re-tested to be sure!

have you not heard of this before ?

Are any tracks record armed?
Also mind posting a simple project that demonstrates this? Because I don't see and never saw those problems (besides plug-ins reporting a wrong delay).

[nofish]

Isn't that a natural thing that when you play an unprocessed file against one processed eg with a compressor that same phasing differences occur?

Or are you saying the compressor (in this case) is just inserted, but doesn't do any processing?

edit;
Think I remember reading in another forum that (some?) Digitalfishphone plugs do indeed not report the correct latency to the host. Something which never got fixed. So that wouldn't be Reaper's fault then. (But not quite sure if I remember correctly). Don't know about other plugins though.

[LOSER]

Originally Posted by nofish

Isn't that a natural thing that when you play and unprocessed file against one processed eg with a compressor that same phasing differences occur?

Or are you saying the compressor (in this case) is just inserted, but doesn't do any processing?

I just assume he isn't doing any procession, because that would be stupid.
Furthermore does it not matter, because a compressor should only change the amplitude of a signal and thus not touching the phase, so it shouldn't cause any phase issues even if it does process the sound (not true for a multiband and some vintage/hw emulation plug-ins though).

[dub3000]

Originally Posted by LOSER

I just assume he isn't doing any procession, because that would be stupid.
Furthermore does it not matter, because a compressor should only change the amplitude of a signal and thus not touching the phase, so it shouldn't cause any phase issues even if it does process the sound (not true for a multiband and some vintage/hw emulation plug-ins though).

pretty sure that dominion does all sorts of weird EQ/filtering on the incoming signal. when there is hardware emulation or EQ, you *can* get weird phasing between clean/non-clean tracks.

for the original poster: LOSER's vst pack has a transient designer that is way better sounding than dominion, i reckon. (but i looooooove the digitalfishphones blockfish compressor though, and spitfish is a pretty great de-esser).

[jens]

Originally Posted by nofish

Think I remember reading in another forum that (some?) Digitalfishphone plugs do indeed not report the correct latency to the host. Something which never got fixed. So that wouldn't be Reaper's fault then. (But not quite sure if I remember correctly). Don't know about other plugins though.

yes, that's the thing - Rreaper's PDC works flawlessly, unless the plugins don't work flawlessly - it's not Reaper's fault at all, and the problem with the respective plugin (e.g. Dominion) will exist in any host.

[Greg M]

With Izotope Ozone there is a preference to set for delay compensation. It does not default to this setting. You must do it manually. I have the latest version.

Greg

[Ollie]

FWIW I just checked all my dynamics plugins and all the old Digitalfishphones plugs including THC behave like this, the only exception is Floorfish.

[xackley]

Dominion, I believe, is meant to mess with phase. It is an "enhancer". The problem has nothing to do with PDC.

[LOSER]

Originally Posted by dub3000

pretty sure that dominion does all sorts of weird EQ/filtering on the incoming signal. when there is hardware emulation or EQ, you *can* get weird phasing between clean/non-clean tracks.

Yeah, just tested it and dominion does some weird stuff. I was more thinking of a compressor like ReaComp, or also the mda_transient, which won't do anything else than alter dynamics, like it IMO should be.

Just to see why Dominion does what it does, I put it in Christian Budde's VST analyzer and now I can tell you that is uses (what appears to me to be a) resonant low-pass filter at around 50 Hz with 12dB/oct or something, so yeah it indeed messes with the phase quite a LOT. But like said hw emulations tend to do this (just like hw does) and afaik do the fish phone VSTs try to be some sort of analog emulations. Anyway there is nothing wrong with REAPER nor its PDC nor with anything else....

Originally Posted by dub3000

LOSER's vst pack has a transient designer that is way better sounding than dominion, i reckon.

Well it isn't that good.

[dub3000]

Originally Posted by LOSER

Well it isn't that good.

But... it helped save a horrible drum track!

[Wilkesin]

Not to derail, but i still can't figure out why people like those fish phones plugs. They were never interesting to me at all.

[planetnine]

Originally Posted by Wilkesin

Not to derail, but i still can't figure out why people like those fish phones plugs. They were never interesting to me at all.

cos they're quick and free and relatively easy for some guys to understand.

I record quite a few live shows for bands so that they can put themselves a demo together. this is done on the cheap and they take the files and mix themselves; I recommend REAPER if they've not got a setup, and blockfish is a comp i'll suggest if they can't get their heads around ReaComp or whatever... They're not perfect, but they'll get quick and easy results for beginners.

I've just tested two parallel tracks and there's definitely some HF loss unless blockfish's own bypass button is pressed -even with no apparent compression happening. This shouldn't be a problem to most live stuff as there's loads of spill anyway, but it would be nice to have it fixed...

>

[LOSER]

Originally Posted by planetnine

[...] but it would be nice to have it fixed...

Not gonna happen for 2 reasons, first I think the sound that they currently have is intended because they are supposed to be some sort of analog modeled devices as far as I know anyway and second Sascha has abandoned those plug-ins along time ago and is now working for image line. He also stated multiple times he is not gonna return, fix, nor change anything.

[wilbur74]

Many thanks for the comments! I've found that "delay compensation" tickbox in Izotopes' Ozone and it works fine..

[Alteregoxxx]

I have noticed the same strange behaviour with some plugins i use frequently, nasty CS(from bootsy),ElectriQ for example...i think that Reaper have some problems with PDC on plugins that use internal OVERSAMPLING,in fact the phase issues disappears if i use ElecriQ in digital and ECO(so no oversampling here...)mode and appears if i untick ECO mode(so Oversampling here)....obviously in analog mode the problem is there but in this mode this isn't a Reaper problem bacause this is related with the analog behaviour of ElectriQ----hope this will be fixed...

[GGB.C.]

Hello,

I just dig up this 12 years old post because I think I have the same problem and I don't understand what.

During sound monitoring I was using great free plugins from Tokyo Dawn Lab. And I noticed that on a microphone pair tracks the sound was changing if I put a plugin on only one channel. And if I put the same plugin on both tracks, the sound was "normal" again. Of course I checked PDC, but if the latency introduce by a plugin was written in the bottom of the FX windows, PDC didnt do what I expected (or what I badly understand). Put the same delay time of a plug on all the other tracks.

I notice the same problem, with ReaFIR for example. If I resize the analyzing window to 4096, there is 4096 samples of latency. And once again PDC, don't seem to change anything in monitoring.

What's the problem ? My understanding ? A general PDC disabled option hide somewhere ?

Thanks a lot.
Tschüss
GG

[siniarch]

Not sure if this is related, but ReaInsert suffers from the same thing. I think the automatic delay feature is waiting for the delay to be reported from the FX software in the case of the Plugin Delay feature, and it waits for the hardware to report the delay with the ReaInsert function. When either the Software or the Interface communicates the wrong delay, it introduces an issue.

It would be great if this delay could be handled in a brute automated way. Where Reaper actually send an audio signal and listened for it and calculated the delay. Perhaps that's what it's doing, but I can never get it to give me a proper delay setting. My interface has FPGA FXs in the unit and depending on which FX I place, the delay changes. So it's not like using an Analogue hardware that has a constant delay based on the length of the cables and circuits.

I wish this could be addressed.

[Elevate Audio]

Reaper's PDC is the worst of the major DAW's. Put a few high latency plugs on the master bus and bypass/unbypass them. It's doesn't take long before you get a obnoxious stuttering as PDC tries to keep up but can't. I know of no other DAW that handles this so poorly. It keeps Reaper from being a professional product.

[mschnell]

Originally Posted by Elevate Audio

Reaper's PDC is the worst of the major DAW's.

There have been multiple long discussions on that issue.

Result:

The only thing that is found bad on Reaper's site is that multiple PDC intriducing plugins in a row introduce at least one buffer of project latency EACH (i.e. the induvidual latency gets rounded up to the next full buffer) and not (as would technically be possible) the summed up latency gets rounded up to the next full buffer.
This has been acknowledged by the devs and will be mended in an upcoming version.

-Michael

[Giovanni_B]

Originally Posted by mschnell

There have been multiple long discussions on that issue.

Result:

The only thing that is found bad on Reaper's site is that multiple PDC intriducing plugins in a row introduce at least one buffer of project latency EACH (i.e. the induvidual latency gets rounded up to the next full buffer) and not (as would technically be possible) the summed up latency gets rounded up to the next full buffer.
This has been acknowledged by the devs and will be mended in an upcoming version.

-Michael

What does that mean in terms of audible problems?

In my templates have several busses with at least two or even three PDC introducing plugins one after another.

Are those tracks losing sync?

[mschnell]

The goal of the PDC handling is to avoid snyc problems between tracks, and AFAIK, if ll plugins clam the correct latency to the DAW, that works perfectly.

The issue discussed in these other threads is the project latency, e.g. the time from hitting the play button to hearing the sound that is supposed to be issued at the position of the current play cursor. When using Midi, this might result in midi event near the play cursor at this point to be not correctly handles or timed.

All this should be completely irrelevant when rendering a complete project.

-Michael

[Goldreap]

Originally Posted by Giovanni_B

What does that mean in terms of audible problems?

In my templates have several busses with at least two or even three PDC introducing plugins one after another.

Are those tracks losing sync?

My understanding:

No, those tracks are not losing sync.
The advantage of the new (forthcoming) option to sum the serial latencies then round up to the next block is to reduce the overall latency of what you hear when recording and monitoring through REAPER a live input.
This might not always solve anything, E.G. if the summed (then rounded up) latency was still more than 'comfortable' you would have to fall back on using 'direct monitoring' through the interface, or perhaps just disabling the culprit FX.
But it could be handy, E.G. block size 128 and 4 plugins with latency; 10 ms, 20 ms, 30ms, and 40ms. Sums to 100 ms... rounded up to 128... an extra 128 might be comfortable, whereas an extra 512 (each plugin rounded up) might not.