Logic Bounces Audio at -3dB?

Wow, Jim.

I never cease to be impressed and amazed with just how educational and informative this darn forum is.

You cats are AWESOME.

Thank you.

Javier

Oh . . . and by the way, as you may have guessed, going to the Pan Law setting was exactly what did the trick. It was set to, you guessed it, -3dB. 🙂

Wo. Weighty stuff here. Lots to think about and digest. I'm kind've sort've in the same boat steve is in, however. I did some level comparisons between my -3 and 0 dB Pan Law change bounces, and it seemed to be okay enough . . . but then again I wasn't even listening FOR certain things that have been brought to my attention here . . . So I'll go back and take another listen.

Thank you, Koros. So then what does this mean?

Does Logic have an automatic "attenuation curve" when switched to -3dB? Because if you drop the Pan Law to -3, then even though your center will now be at 0dB (or at least perceived 0dB), wouldn't your left and right pans be at -3dB? putting you right back where you started? - i.e. would you still have the disparity between the perceived loudness of the center verses the left and right pans whether you're at 0dB or -3dB?

Hmmm . . . I think I will read up on this. I hope the manual has something on it. Going to look it up right now.

Thanks

Just looked for Pan Law in the manual. Nothing. Didn't find anything. Anyone know where I can get some info on it?

Okay. So here we go. I did some personal pan law tests on a sequence I set up specifically for it, and here's what seems to be the case (which, I understand, has already more or less been stated, but I just wanted to voice it so I can make sure I grasp it outside of my own mind):

Just as Koros mentioned, a signal has an increase PERCEIVED loudness at center pan which, specifically, is between two and three decibels louder than when the signal is panned far right or far left. Now, note that the loudness is perceived only. It's not an actual decibel increase. It is only perceived as such by our ears. To deal with this, logic institutes the following compensators.

One, called -3db, does not change the decibel level of the signal(s) panned left or right, but, instead, decreases the level of the center signal(s) by about -3dB. This (again, as Koros mentioned) achieves a smoothess in the perceived loudness of signals by dropping the center signal to what appears to be an even level with the left and right signals (even though, again, it's in all actuality, about 3dB LESS).

The other parameter, called -3dB compensated, achieves the same effect of level smoothness, but instead of decreasing the decibel level of the center signal and leaving the left and right signals alone, it does the opposite: It leaves the decibel level of the center pan signal as is, and increases the decibel level of the left and right signals. Again, it achieves the same effect of level smoothness across the pan spectrum, but does it in a different manner.

What I've found is that -3db (that is WITHOUT the compensation) seems to keep your levels where you had them and . . .

Oh wait a minute . . . I just discovered a MAJOR pain in the rumpus. What you HEAR at a pan law of -3dB or -3dB Compensated prior to bouncing is NOT what you will hear AFTER you import that bounce back into the same sequence. If you want to hear what you bounced at -3dB back at that very same level (which, presumabely, one would) you actually have to change your sequence back from -3dB (or -3dB Compensated) to 0dB. Otherwise Logic will add a -3dB pan law to the bounce that was already made with the -3dB pan law on it in the first place. Ugh. That can be very, very confusing . . .

sigh . . .

Somebody enlighten me if there's something I'm not getting.

Javier

Well, one possible solution/remedy to this that I've discovered is simply putting a gain filter on the bounced sequence that you imported into the -3db Pan Lawed sequence and raising the gain level of the bounced region to 3dB. Doing some tests showed that the output levels and perceived volume dynamics between the original sequence signal at a pan law of -3dB and the bounced master at a +3dB gain are concurrent if not identical.