Null test with bounced audio file weird!

I've reported this to Apple (about 4 months ago), and they are likely working on it.

It a corruption of a lower bit and it ONLY happens in Logic 8.. (the one that people feel compelled to say 'sounds better than 7" .. well, it doesn't..!)

The null even with a 24 bit file is only -90.3 dB


In logic 7 I get null of greater than -215dB

You can see a thread here:

http://recforums.prosoundweb.com/index.php/mv/msg/24947/0/0/5734/

zmix wrote:
I've reported this to Apple (about 4 months ago), and they are likely working on it.

It a corruption of a lower bit and it ONLY happens in Logic 8.. (the one that people feel compelled to say 'sounds better than 7" .. well, it doesn't..!)

It does sound better when playing back a mix... but the bounced file sure doesn't
In fact, on my studio monitors the bounced file from Logic doesn't very much at all like the playback mix... and I've been at this a long time. Version 7 doesn't behave like this. I've switched to Cubase in order to finish an album I'm mixing for a client.

pancenter-

This is the result of my null-test :



It is audible too, like a transistor picking up a faint radiosignal.

regards, Erik.

It seems that this problem is related to the various pan laws. If the pan law is set to 0db then the bounce nulls as expected. -3db and -3db comp produce the problems. Also, UTM off with a pan law setting of -3db works as expected. Can someone test the various pan laws in 7.2.3 to see if this problem exists ? I don't have it installed at the moment to compare.


-GK

Just tested Logic 7.2.3. Same results as Logic 8.02 when changing the pan laws. If you want a true 1 for 1 bounce then you need to use the 0db pan law. The -3db setting works correctly with UTM off. The -3db comp setting alters the sound whether UTM is on or off.


-GK

Hi,

I get the 'loudest' readouts with panlaw set to -3dB comp and UTM off.

The 'quitest' is panlaw 0dB and UTM on:



panlaw -3dBcomp and Universal Track Mode off:



regards, Erik.

Grassrootking wrote:
It seems that this problem is related to the various pan laws. If the pan law is set to 0db then the bounce nulls as expected. -3db and -3db comp produce the problems. Also, UTM off with a pan law setting of -3db works as expected. Can someone test the various pan laws in 7.2.3 to see if this problem exists ? I don't have it installed at the moment to compare.


-GK

But the pan law theory wouldn't explain the fact that I get 215dB nulling in Logic 7, but under identical conditions I only get 90.3dB in Logic 8.

It's a new problem, and it exists only in Logic 8 and Waveburner 1.5...

zmix,

How are you performing the test for the null ? I would like to try it so that I can see the same results you are getting. Here is what I am doing to test the pan law theory.


System Settings: Logic 8.02, Pan Law = 0, UTM = on


1. I import a track, in this case "Heartless" by Kanye West, to audio track 1 at bar 1.

2. Set audio track 1 to stereo.

3. Select the region and then click on Set Locators to set the cycle for the region length.

4. Click on the Bounce icon.

Settings for Bounce: 16bit, PCM, Interleave, no dither

5. Bounce and import new file to audio track 2 at bar 1 and set to stereo.

6. Insert Gain plugin and click the Left and Right phase buttons.

7. Insert Multimeter plug and adjust for finer resolution up to -120.

8. Hit play to analyze spectrum.

My results show nothing audible or visually detected at -120 or beyond even if I add 2 Gain plugins set at +24.

This happens in Logic 7.2.3 as well.


When performing the same steps with a pan law of -3db or -3db comp the results are drastically different.

Please let me know if I am in error here.


-GK

Yep, that's exactly what I did to get the results on the screens (+9. Take Screenshot) in the earlier message (except I used another track), with the notable difference that UTM off was 'quiter' than UTM on with all three panlaw settings. Weird that that behaviour appears to be reversed on my machine.

My results show nothing audible or visually detected at -120 or beyond even if I add 2 Gain plugins set at +24.

Well, you've seen my results... there must be some difference other than the ones we covered... Maybe it is best to perform this test with an Apple Loop everybody has, to eliminate possible file encoding differences. I used 'Academy Viola 1' from the Symphony Orchestra Jam Pack for my tests.

regards, Erik.

Erik,

Just reproduced the same results using Academy Viol 01. I didn't even have to play it to see that there is a difference between to the two files. Since the Apple Loops are in the CAF format I tried bouncing to that format first using the 0db pan law. The resulting file does not match the original at all. If I use the bounced CAF file as the source and bounce again, the files match and they null. Very strange behavior. I wonder if this is just exclusive to Apple Loops.

-GK

I tried the 'perfect' test: nulled the same file. The file is a 16/44.1 stereo interleaved AIFF bounce from the Test Oscillator, at a sinus of 439.90 (which is what it jumps to when you type in 440...) and it's own out level at -12 dB (default).
Panlaw OdB, UTM on. *+ 72 dB extra gain* via 3 Gain plugs.



The bottom line? RMS seems to be well under -170 dB here... The different harmonic content of files can give different results, I think.

I can only say that in the real world of composing and mixing music I use the pan law of -3dB compensated and no client has come back to me and said that they tried to null files out and mine did not null.

Maybe one of you will educate me, but I do not see this so much as a problem, more an anomaly.

I see on my System that Logic start by default as UTM mode off...
My songs are UTM-off mode in most of case (for the DAE TDM compatibilty)... and when I use UTM on, I try to don't forget to change pan law at 0dB (sometime I forget)

after this post I will put much care about that...

Anyway: my workflow about using Pan-pot was changed from many years...
I always use Direction Mixer plugin instead pan-pot trim on the Logic mixer! (in that case there is no reason to use pan law set as - 3db because Direction Mixer plugin gives the best "PanLaw" that you can find 🙂 )

.... however a noise floor at - 100 dB is normal on Digital Mix... there are converter/Mics/preamps/ AND-DONT-FORGET-SAMPLES that comes with a more highest noise!

only one question...
Have you listen the noise floor of the EWQLSO samples?
Only VSL samples comes with a decent S/N ratio but it reach about 80 dB in the best condition!

So, forget this kind of self-HOME-made-TEST! (they are totally unuseful when you mixdown a REAL RECORDING song! (this test must be done with a mono pure sine-wave.. NOT WITH A STEREO FILE!!!)

Clearly there are people who have time to waste ... This is normal!

But Logic is the most powerful DAW... that's way many people talk about it!
Logic gives you the best sound as possible with ONLY 500 Dollars...
the other DAWs are OBSOLETE if compared.

Just my 2 cents

G