Darn it I think I found a serious bug - please help confirm!

[Boombastix]

Darn it, I think there is a bug in the Reason audio engine how latency/delay comp is managed?
If anyone have time please confirm so I can report. I had to run 18! cases, but you can perhaps pick the highlighted ones that I had issues with to see if you can replicate.
Besides all the description below I also included a picture with all the audio tracks, it is very visible. Pink track marks problem.
Different VST apparently generates non-identical results.

I have used two long latency VSTs such as:
Acustica Ceil (delay comp = 2688)
IK Steahlth Limiter (delay comp = 4871)
It seems the IK Steahlth still reports 4871 in VST Bypass, but 266 when using the Audio Track Bypass.
Acustica Ceil in both Bypass modes gives the normal 266.

"Audio Track BYPASS" refers to the bypass button on the Audio Track.
"VST BYPASS" refers to the bypass button on the VST rack unit.

Doing Bounce In Place:
Bounce in place w IK Steahlth ON works fine.
Bounce in place w IK Steahlth Audio Track BYPASS works fine.
Bounce in place w IK Steahlth VST BYPASS do not work. Drums hit EARLY.
Bounce in place w Acustica Ceil ON works fine.
Bounce in place w Acustica Ceil Audio Track BYPASS works fine.
Bounce in place w Acustica Ceil VST BYPASS works fine.

Recording live from one Audio Track over to a new Audio Track:
Recording live w IK Steahlth ON does not work. Drums hit LATE.
Recording live w IK Steahlth Audio Track BYPASS works fine.
Recording live w IK Steahlth VST BYPASS works fine.
Recording live w Acustica Ceil ON does not work. Drums hit LATE.
Recording live w Acustica Ceil Audio Track BYPASS works fine.
Recording live w Acustica Ceil VST BYPASS works fine.

Playing two tracks simultaneously, one with the VST and one without, same drum loop in the sequencer:
Playing w IK Steahlth ON works fine.
Playing w IK Steahlth Audio Track BYPASS works fine.
Playing w IK Steahlth VST BYPASS do not work. Drums hit EARLY.
Playing w Acustica Ceil ON works fine.
Playing w Acustica Ceil Audio Track BYPASS works fine.
Playing w Acustica Ceil VST BYPASS works fine.

The setup for "live" recording from one Audio Track to another Audio Track:
Drum loop on an Audio Track with one of the above VSTs on the insert. Activate "Rec Source" on this Audio Track channel. Create a new/empty Audio Track and record the audio into this new track.
"Playing" means I just play the track with the drum loop and the VST and compare it to my reference track playing its drum loop but with no VST.

Reason issue w latency testing IK Stealth.gif (97.3 KiB) Viewed 1844 times

[Loque]

Did you tried switching back to old audio render settings? To me it sounds like a general problem.

[Boombastix]

Did you tried switching back to old audio render settings? To me it sounds like a general problem.

Do you mean in the preferences?
I used the 10.3 setting "Render audio using audio card buffer size setting" in my OP testing.

Now, I switched this setting off and made a new quick test of three cases and still same issue (see below picture). These:

Playing two tracks simultaneously, one with the VST and one without, same drum loop in the sequencer:
Playing w IK Steahlth ON works fine.
Playing w IK Steahlth Audio Track BYPASS works fine.
Playing w IK Steahlth VST BYPASS do not work. Drums hit EARLY. (Delay comp stays the same...)

This video illustrates the issue:

R v0.3 pref.gif (26.05 KiB) Viewed 1704 times

[Loque]

Did you tried switching back to old audio render settings? To me it sounds like a general problem.

Do you mean in the preferences?
I used the 10.3 setting "Render audio using audio card buffer size setting" in my OP testing.

Now, I switched this setting off and made a new quick test of three cases and still same issue (see below picture). These:

Playing two tracks simultaneously, one with the VST and one without, same drum loop in the sequencer:
Playing w IK Steahlth ON works fine.
Playing w IK Steahlth Audio Track BYPASS works fine.
Playing w IK Steahlth VST BYPASS do not work. Drums hit EARLY. (Delay comp stays the same...)

This video illustrates the issue:

R v0.3 pref.gif

Yea, i meant this. That is weird. IF PH implemneted everything well, the switch should turn everything back to <10.3. Are you really sure, ther wasnt a problem before 10.3?

[Boombastix]

Yea, i meant this. That is weird. IF PH implemneted everything well, the switch should turn everything back to <10.3. Are you really sure, ther wasnt a problem before 10.3?
[/quote]

I never used IK Stealth on an insert in 10.2. So I can not say. And when I discovered it I just tried another big latency unit (Acustica) to see if it is Reason or the VST that seems to have a problem. Since IK and Acustica both had problems, but manifested in different ways, I guess it has to be Reason, but like i said Reason 10.2 or earlier, I do not know. This is on my Windows 10 machine by the way.

Would be good if someone can replicate this, say older Reasons, Mac, 10.3/Win (what I have). Just make sure you only use one plugin on the insert and one with a big latency so you can hear it. IK Stealth is almost 1/16 not off, so yeah, you'll hear that

[ShawnG]

I only have stealth limiter, but other than the live wiring of one audio track to another, which I didn't do, I can confirm that this happens. I also tested it by exporting a song with Stealth Limiter in the Mastering Section both live and bypassed, and the behavior is the same. The Stealth Limiter plugin itself has a bypass button, and if you use that to bypass the stealth limiter, that works with the PDC just fine (unless you also bypass the Reason VST wrapper)

I'll have to check some of my other DAWs to see if its better there, I'm not sure at this point if this is a huge bug, or just the way it has to work because the fix might do other weird things to the track, but If you need to bypass Stealth limiter, you can do so just fine with the plug in's own bypass button.

[Boombastix]

I only have stealth limiter, but other than the live wiring of one audio track to another, which I didn't do, I can confirm that this happens. I also tested it by exporting a song with Stealth Limiter in the Mastering Section both live and bypassed, and the behavior is the same. The Stealth Limiter plugin itself has a bypass button, and if you use that to bypass the stealth limiter, that works with the PDC just fine (unless you also bypass the Reason VST wrapper)

I'll have to check some of my other DAWs to see if its better there, I'm not sure at this point if this is a huge bug, or just the way it has to work because the fix might do other weird things to the track, but If you need to bypass Stealth limiter, you can do so just fine with the plug in's own bypass button.

Thanks for checking. Yeah another DAW would be great to know about, then we know if it is Reason isolated or not. I agree that we can work with Reason until it gets fixed, so not tooooo bad.
But I do things from time to time where I want to do the "Recording live from one Audio Track over to a new Audio Track" and I get all wacky offsets. So I guess we get things delayed when recording from one track to another, and AFAIK I think I had this issue with Neptune recently as well, but did not investigate further, so maybe any RE or VST with lots of latency will show this behavior.

[ShawnG]

I only have stealth limiter, but other than the live wiring of one audio track to another, which I didn't do, I can confirm that this happens. I also tested it by exporting a song with Stealth Limiter in the Mastering Section both live and bypassed, and the behavior is the same. The Stealth Limiter plugin itself has a bypass button, and if you use that to bypass the stealth limiter, that works with the PDC just fine (unless you also bypass the Reason VST wrapper)

I'll have to check some of my other DAWs to see if its better there, I'm not sure at this point if this is a huge bug, or just the way it has to work because the fix might do other weird things to the track, but If you need to bypass Stealth limiter, you can do so just fine with the plug in's own bypass button.

Thanks for checking. Yeah another DAW would be great to know about, then we know if it is Reason isolated or not. I agree that we can work with Reason until it gets fixed, so not tooooo bad.
But I do things from time to time where I want to do the "Recording live from one Audio Track over to a new Audio Track" and I get all wacky offsets. So I guess we get things delayed when recording from one track to another, and AFAIK I think I had this issue with Neptune recently as well, but did not investigate further, so maybe any RE or VST with lots of latency will show this behavior.

To be honest, the reason I didn't test the recording live from one audio track to another is that the behavior is what I would expect from that scenario, you wouldn't notice it too much with most plugins, because the delays aren't that huge, maybe just a phasing effect when you play both tracks at the same time. Putting an enormous look ahead mastering limiter on a single track insert wouldn't be common engineering practice, and because of how the thing works, that's what you get.

Now, I did fire up Cubase Pro 9.5 and tried the same render in place scenario, and Cubase did render the track correctly with the insert effect enabled and disabled. but Cubase also had a more pronounced audio dropout when toggling the effect on and off while the loop was playing (Reason's audio was glitchy when you do this, but playback of the track does not stop entirely and timing seemed to be maintained; Cubase audio dropped out and restarted, with the timing interrupted.) obviously the programs handle that situation differently. If you use Stealth limiter's bypass button on the plugin, there were no dropouts or glitches on either program when switching the effect on and off live.

I then checked a couple other plugs, neptune and the VST version of the Ozone 8 Maximizer (on it's craziest mode) Both bounced to a new clip while bypassed on the rack perfectly.

So there appears to be a problem with the PDC correction for IK's stealth limiter, in Reason only. even in your examples, the only time the acustica plugin misbehaved was when live wiring one track to another, which is probably normal behavior. (If I'm wrong about that, I'd be more than happy to be corrected) The IK plug though is obviously different than other latent plugs when bouncing in place bypassed by Reason's rack control, when it is not in Cubase.

[ShawnG]

lol, I just opened Cubase and tried running 2 identical loops, one with IK SL as an insert and one dry. Setup the same in reason.

played them back in both DAWs and switched the limiter on while the track was playing. Reason had a glitchy sound when the effect was switched on, but the audio for both tracks remained in sync, and kept playing.

Cubase had a dropout, and then it was like a delay pedal had been switched on, and the beats played at different times.

edit: I played with it some more, and it does appear that cubase has both a on/off plugin button on the plugin window, and a separate effect bypass button on the inspector (been a while since I used it) It appears that if you use the on/off button on the plugin window border in cubase the PDC works ok with rendering. If You use the Bypass button on the track inspector, Cubase behaves badly with this plugin as well.

so then I googled, and found this: https://cgi.ikmultimedia.com/ikforum/vi ... 13&t=14904

[Boombastix]

Ok, see if another 18! test runs gives us more clarity. This time with Neptune and Ozone in IRC II mode. The two REs I have with enough latency to make it visible and audible if there is an issue (perhaps the only two that exists, since Polar seems unsuitable for this test with built in delay). New test results below:

But check this. Neptune has EXACTLY the same the issues as Acustica. Ozone has EXACTLY the same the issues as IK Stealth Limiter. Huh? Yes, for each and every one of the 9 tests, line by line, the problem matches... The plot thickens Sherlock
Acustica/Neptune latency values (2688/1117). IK Stealth/Ozone latency values (4871/3280). Far fetched to believe that latency > 3200 has anything to do with it (50x64=3200)?
But if our REs behaves like this as well - I think it just has to be Reason not managing latency in the different scenarios. Please share more thought if you have.

Doing Bounce In Place:
Bounce in place w Ozone RE ON works fine.
Bounce in place w Ozone RE Audio Track BYPASS works fine.
Bounce in place w Ozone RE RE BYPASS do not work. Drums hit EARLY.
Bounce in place w Neptune RE ON works fine.
Bounce in place w Neptune RE Audio Track BYPASS works fine.
Bounce in place w Neptune RE RE BYPASS works fine.

Recording live from one Audio Track over to a new Audio Track:
Recording live w Ozone RE ON does not work. Drums hit LATE.
Recording live w Ozone RE Audio Track BYPASS works fine.
Recording live w Ozone RE VST BYPASS works fine.
Recording live w Neptune RE ON does not work. Drums hit LATE.
Recording live w Neptune RE Audio Track BYPASS works fine.
Recording live w Neptune RE VST BYPASS works fine.

Playing two tracks simultaneously, one with the VST and one without, same drum loop in the sequencer:
Playing w Ozone RE ON works fine.
Playing w Ozone RE Audio Track BYPASS works fine.
Playing w Ozone RE RE BYPASS do not work. Drums hit EARLY.
Playing w Neptune RE ON works fine.
Playing w Neptune RE Audio Track BYPASS works fine.
Playing w Neptune RE VST BYPASS works fine.

[Creativemind]

I only have stealth limiter, but other than the live wiring of one audio track to another, which I didn't do, I can confirm that this happens. I also tested it by exporting a song with Stealth Limiter in the Mastering Section both live and bypassed, and the behavior is the same. The Stealth Limiter plugin itself has a bypass button, and if you use that to bypass the stealth limiter, that works with the PDC just fine (unless you also bypass the Reason VST wrapper)

I'll have to check some of my other DAWs to see if its better there, I'm not sure at this point if this is a huge bug, or just the way it has to work because the fix might do other weird things to the track, but If you need to bypass Stealth limiter, you can do so just fine with the plug in's own bypass button.

So you tried on 10.2 or an older version just to clarify?

[ShawnG]

I only have stealth limiter, but other than the live wiring of one audio track to another, which I didn't do, I can confirm that this happens. I also tested it by exporting a song with Stealth Limiter in the Mastering Section both live and bypassed, and the behavior is the same. The Stealth Limiter plugin itself has a bypass button, and if you use that to bypass the stealth limiter, that works with the PDC just fine (unless you also bypass the Reason VST wrapper)

I'll have to check some of my other DAWs to see if its better there, I'm not sure at this point if this is a huge bug, or just the way it has to work because the fix might do other weird things to the track, but If you need to bypass Stealth limiter, you can do so just fine with the plug in's own bypass button.

So you tried on 10.2 or an older version just to clarify?

No, I'm on 10.3 but I tried it on cubase, and in a lot of ways it's worse there with the IK plug (although it seems to bounce in place ok, it's way worse playing alongside other tracks in cubase than it is in Reason)

[ShawnG]

Ok, see if another 18! test runs gives us more clarity. This time with Neptune and Ozone in IRC II mode. The two REs I have with enough latency to make it visible and audible if there is an issue (perhaps the only two that exists, since Polar seems unsuitable for this test with built in delay). New test results below:

But check this. Neptune has EXACTLY the same the issues as Acustica. Ozone has EXACTLY the same the issues as IK Stealth Limiter. Huh? Yes, for each and every one of the 9 tests, line by line, the problem matches... The plot thickens Sherlock
Acustica/Neptune latency values (2688/1117). IK Stealth/Ozone latency values (4871/3280). Far fetched to believe that latency > 3200 has anything to do with it (50x64=3200)?
But if our REs behaves like this as well - I think it just has to be Reason not managing latency in the different scenarios. Please share more thought if you have.

Doing Bounce In Place:
Bounce in place w Ozone RE ON works fine.
Bounce in place w Ozone RE Audio Track BYPASS works fine.
Bounce in place w Ozone RE RE BYPASS do not work. Drums hit EARLY.
Bounce in place w Neptune RE ON works fine.
Bounce in place w Neptune RE Audio Track BYPASS works fine.
Bounce in place w Neptune RE RE BYPASS works fine.

Recording live from one Audio Track over to a new Audio Track:
Recording live w Ozone RE ON does not work. Drums hit LATE.
Recording live w Ozone RE Audio Track BYPASS works fine.
Recording live w Ozone RE VST BYPASS works fine.
Recording live w Neptune RE ON does not work. Drums hit LATE.
Recording live w Neptune RE Audio Track BYPASS works fine.
Recording live w Neptune RE VST BYPASS works fine.

Playing two tracks simultaneously, one with the VST and one without, same drum loop in the sequencer:
Playing w Ozone RE ON works fine.
Playing w Ozone RE Audio Track BYPASS works fine.
Playing w Ozone RE RE BYPASS do not work. Drums hit EARLY.
Playing w Neptune RE ON works fine.
Playing w Neptune RE Audio Track BYPASS works fine.
Playing w Neptune RE VST BYPASS works fine.

I can confirm the Ozone Maximizer RE does the same thing here. However, I wouldn't read all that much into it, That RE was released when Reason didn't even have plugin delay compensation at all, so I don't know how much work Izotope put into that end of the plugin's behavior, I think it's been abandoned basically since Reason 6.5.

The new Ozone Maximizer VST works fine in this situation, as have all the other non-IK plugs I've tested with it, and the IK plug has issues in the same situation in other hosts.

Live looping back and recording with latent plugins just isn't going to work, PDC can't account for everything, it isn't a magic tool, and I've had various instances with PDC in every DAW I've ever used, usually in live recording situations, that's even without doing bizarre things like using a mastering plugin as an insert on a track.

As for the other plugins and issues, you can see that the issue is plugin related, simply by going to the back of the rack, expanding the "Show Programmer" button you will see the plugin's reported latency, and a manual offset option. The plugins that are going to give you a problem, you will see that the latency stays the same when you put the plugin in bypass mode, whereas most plugins go to 0 latency in that situation, the plugin is still telling Reason to offset by that amount of samples, but then Isn't actually doing any processing to use that number of samples.

Simply put, I don't think this is a bug on Props' end. I would try to not use mastering type plugins on insert tracks, and find some other work around other than live looping back into the sequencer. (or manually compensate for that somehow)

https://ask.audio/articles/understandin ... mpensation

[Boombastix]

Can someone check to see if this got fixed in R11?

Here is how to test quickly.
1 Create an AUDIO TRACK and add a short transient sound, blipp, kick, hihat, something like that.
2. INSERT two (2) NEPTUNE as InsertFx
3 Set Neptune to: Low Freq (latency = 1702 each, for a total of sum of 3404)
4 Turn off: Pitch Adjust, Transpose, and Formant. (not necessary, but less audio artifacts if off)
5 ENABLE "Rec Source" on the channel.
6. Create a new Audio Tracka dn arm for recording with the first track as source.
7. Record (press star '*') and record over to this new track.

Look at the wave forms, there should be almost 1/16th of a note lag in the recorded material if this problem still exists.

[chimp_spanner]

Can someone check to see if this got fixed in R11?

Here is how to test quickly.
1 Create an AUDIO TRACK and add a short transient sound, blipp, kick, hihat, something like that.
2. INSERT two (2) NEPTUNE as InsertFx
3 Set Neptune to: Low Freq (latency = 1702 each, for a total of sum of 3404)
4 Turn off: Pitch Adjust, Transpose, and Formant. (not necessary, but less audio artifacts if off)
5 ENABLE "Rec Source" on the channel.
6. Create a new Audio Tracka dn arm for recording with the first track as source.
7. Record (press star '*') and record over to this new track.

Look at the wave forms, there should be almost 1/16th of a note lag in the recorded material if this problem still exists.

Hey! Just tested in R11. Results attached:

Screenshot 2020-01-07 at 15.53.06.png (12.74 KiB) Viewed 1090 times

[Boombastix]

Can someone check to see if this got fixed in R11?

Here is how to test quickly.
1 Create an AUDIO TRACK and add a short transient sound, blipp, kick, hihat, something like that.
2. INSERT two (2) NEPTUNE as InsertFx
3 Set Neptune to: Low Freq (latency = 1702 each, for a total of sum of 3404)
4 Turn off: Pitch Adjust, Transpose, and Formant. (not necessary, but less audio artifacts if off)
5 ENABLE "Rec Source" on the channel.
6. Create a new Audio Tracka dn arm for recording with the first track as source.
7. Record (press star '*') and record over to this new track.

Look at the wave forms, there should be almost 1/16th of a note lag in the recorded material if this problem still exists.

Hey! Just tested in R11. Results attached:

Screenshot 2020-01-07 at 15.53.06.png

Ugh, guess it is still unresolved. Thanks for testing. Funny thing was RS sent me a response a while after I reported it saying they wanted to close this issue. I said no. But I'm wondering what happened if it is still a bug since April last year. I'll see if I can get a new response.