master buss levals not accurate??

[pie55]

I was doing some research into noise floors in my system. its an hd 3 accel system with protools 8.1.1with mac os 10.6.8
In a nutshell I had a track with a pink noise signal peaking @ -22.2 dbfs. but in the master buss the peak was -24.7 dbfs.


when the track is panned L the leval is right @ -22.2dbfs in the master buss but when panned R the master leval is -23 dbfs.


there is nothing in this session. 2 tracks, no plugins, a masterbuss.


I"ve tried the master assigned to c1 and 2 (my spdif outs) as well as analog 1 and 2 with the same measurment differences.


i've tried rebooting the system. I can't figure it out

[JFreak]

Maybe it is the stereo panning law? Surround Mixer has it at 2.5dB and Stereo Mixer has it at 3.0dB. Try swapping surround mixer to stereo mixer and expect half a decibel difference on your test results. It you get it, it works as expected.

[BScout]

Originally Posted by JFreak

Maybe it is the stereo panning law? Surround Mixer has it at 2.5dB and Stereo Mixer has it at 3.0dB. Try swapping surround mixer to stereo mixer and expect half a decibel difference on your test results. It you get it, it works as expected.

It was the other way around. Stereo Mixer was 2.5dB, Surround is 3dB for the old Pro Tools. But panning law has no effect on hard pans.

As for the discrepancy, are you using pink noise generated by the TDM version of the signal generator? It is phase incoherent (so sums of signals like hard panning a stereo left or right will have issues.) Switch to the RTAS version.

I have pointed this out and demonstrated it to Avid multiple times and they acknowledge it (but haven't fixed it.) On current systems, AAX-DSP has this problem. AAX native does not.

[BScout]

To do a check on meter levels, use a sine wave and not pink (or white)

[JFreak]

Originally Posted by BScout

It was the other way around.

Yes, my bad. Was thinking something else while typing. I stand corrected.

[pie55]

thanks everyone! Will try tomorrow

[pie55]

yes with rtas 1k sign wave the panning level is the same. and just to confirm, it''s supposed to be 2.5 db quieter panned centre?

[Darryl Ramm]

It's doing whatever you tell it to. The magic words were already mentioned above: pan law. What pan law settings are in your session setup?

[pie55]

there is no option called pan depth (pan law) in my session set up in version of Protools. hd8.1.1.
i assume that 2.5 makes the track panned to the centre the same perceived level as the same thing panned left or right?

[Darryl Ramm]

Originally Posted by pie55

there is no option called pan depth (pan law) in my session set up in version of Protools. hd8.1.1.
i assume that 2.5 makes the track panned to the centre the same perceived level as the same thing panned left or right?

It's not about making something "the same", it's how you want a signal to sound as it's panned across the sound field. e.g. making the level decrease as it pans hard one way can appear to make it seem more distant/wider.

Oh right, that was not an option way back then, it was a 2.5dB pan law. There are really good articles online online explaining this, and lots of past debate on DUC and elsewhere what it means, if it matters for particular workflows, and if it does what law is best. It's a debate best had while consuming beer.