Audiocave
01-29-2009, 04:38 PM
Although it's irrelevant as to how it came up, and currently not a major issue as I like the mixes I get with Cubase I think we may have a problem with PDC that could affect overall depth or quality. I'd like someone to try to duplicate this test to see if I took a wrong step somewhere.
Load a stereo mix and leave it a unity gain.
Create 8 stereo group busses and send from the source track to those groups at unity.
Load one plugin on each group that have different latency values. I used a couple of Steiny plugs that report nothing and some others like Waves and Voxengo. Try to choose plugs that can zero out all of their internal values and just pass a raw latency compensated signal with no dsp processing into the group busses.
Engage the phase reverse switches on all of the group channels and mute them.
Zero out all of the plugin parameters. Make sure their I/O's aren't adding any gain or trim.
Play / unmute each group against the source and tell me what you see and hear.
You shouldn't see any metering on the master and you shouldn't hear anything.
Now the theory here is that an engaged plugin that requires xxx samples of latency compensation will always require that no matter what it's doing if it's powered on and the stream is passing through it. If it's 2000 samples it's always 2000 samples, even if all the eq bands are switched off. Cubase should compensate for all 8 of those varying latencies individually.
When I did this test 7 of the 8 group signals didn't come close to canceling against the source track. Power off the plugs and they all canceled. I repeated this test in Reaper with many of the same plugs and they all (except one, exception noted) cancelled with the plugins in-line and powered on and their parameters zeroed out. I reset the audio engine a couple of times in Cubase, same thing. The timing of those streams were obviously not close to sample accurate as some combinations of signals that should be identical streams 180 degrees out of phase actually increased the gain into the master.
This is not exactly an indication of well done PDC in Cubase and could have obvious implications for mixes as it relates to overall depth of field and space.
Load a stereo mix and leave it a unity gain.
Create 8 stereo group busses and send from the source track to those groups at unity.
Load one plugin on each group that have different latency values. I used a couple of Steiny plugs that report nothing and some others like Waves and Voxengo. Try to choose plugs that can zero out all of their internal values and just pass a raw latency compensated signal with no dsp processing into the group busses.
Engage the phase reverse switches on all of the group channels and mute them.
Zero out all of the plugin parameters. Make sure their I/O's aren't adding any gain or trim.
Play / unmute each group against the source and tell me what you see and hear.
You shouldn't see any metering on the master and you shouldn't hear anything.
Now the theory here is that an engaged plugin that requires xxx samples of latency compensation will always require that no matter what it's doing if it's powered on and the stream is passing through it. If it's 2000 samples it's always 2000 samples, even if all the eq bands are switched off. Cubase should compensate for all 8 of those varying latencies individually.
When I did this test 7 of the 8 group signals didn't come close to canceling against the source track. Power off the plugs and they all canceled. I repeated this test in Reaper with many of the same plugs and they all (except one, exception noted) cancelled with the plugins in-line and powered on and their parameters zeroed out. I reset the audio engine a couple of times in Cubase, same thing. The timing of those streams were obviously not close to sample accurate as some combinations of signals that should be identical streams 180 degrees out of phase actually increased the gain into the master.
This is not exactly an indication of well done PDC in Cubase and could have obvious implications for mixes as it relates to overall depth of field and space.