Why does consolidation &or bouncing tracks=loss of sound quality

[PeeTee]

Nobody complained about bouncing tracks in analog.

[Doug Ring]

I'd like to raise an issue about Phase-invert comparisons.

Nobody ever says that they hear anything at all when they try this test. The silence is always taken as proof positive that the two files are identical in every respect except absolute phase. Because this would be the case with two analogue files, the assumption is made that it's also true for digital files.

I would contend that it may not be true.

If the files are bit-for-bit identical, you would assume that they should sound the same, and if there are any differences in the bits, they should sound different. But suppose the bit differences are at the very bottom of the signal's resolution: the bottom few steps of the sampling ladder. The phase inversion trick becomes a bit of a blunt instrument when trying to pick out -144dB signals amongst convertor noise. You'd be hard-pressed to hear them, in other words, especially if they were randomly distributed. Yet this very low-level information difference may be just what's contributing to a feeling that the files sound different.

Furthermore, it's now generally accepted that clock accuracy has a big influence on sound quality. There may be timing errors - jitter - introduced by bouncing to disk. These errors may not be made audible by the phase-invert test, but could make two files sound different.

In short, I would suggest that there could be many artifacts within a bounced file that simply do not show up with the phase-invert test. The only sure way to determine what's happening is to compare files on a bit-for-bit basis.

(Discuss, 30 marks)

[narcoman]

interesting doug...

[PeeTee]

Originally posted by Doug Ring:
The phase inversion trick becomes a bit of a blunt instrument when trying to pick out -144dB signals amongst convertor noise.

<font size="2" face="Verdana, Arial">If you can hear down to -144dB, then the molecules bashing into each other in the air must be driving you nuts! [img]images/icons/rolleyes.gif[/img]

[John Kurzweg]

Originally posted by doug ring But suppose the bit differences are at the very bottom of the signal's resolution: the bottom few steps of the sampling ladder. The phase inversion trick becomes a bit of a blunt instrument when trying to pick out -144dB signals amongst convertor noise. You'd be hard-pressed to hear them, in other words, especially if they were randomly distributed. Yet this very low-level information difference may be just what's contributing to a feeling that the files sound different.

<font size="2" face="Verdana, Arial">Ahhhh-so it would be hard to hear on one track-- but what about 32 or 48 or 64 or beyond?If this noise is cumulative then that might explain why when I print a mix of 32 or more tracks from original files and then the identical mix from consolidated files the original files sound more pleasing to me.

[Rich Breen]

Originally posted by Doug Ring:
I'd like to raise an issue about Phase-invert comparisons.
...But suppose the bit differences are at the very bottom of the signal's resolution: the bottom few steps of the sampling ladder. The phase inversion trick becomes a bit of a blunt instrument when trying to pick out -144dB signals amongst convertor noise....

<font size="2" face="Verdana, Arial">Doing the test with a bit scope is definitive.

best,
rich

[ashdot]

We're finishing our first 5.1 project and we started bouncing the mixes today.
I was shocked to listen to the resulting files. They sounded dead comparing to the original PT session, and worst of all, the 2 surround channels are about 5db lower in level!

Does anyone have any similar experiences with bouncing 5.1 mixes?

Izhar
PT HD3, 5.3.1.

[Doug Ring]

PT, I hate the sound of those molecules banging together, especially in the morning after a bottle of malt whisky [img]images/icons/smile.gif[/img]

John, I wouldn't doubt that the cumulative effect of minor distortions could be a problem over many tracks, in the same way that you can get a low-frequency energy build up if you don't roll the LF off on each track.

Can't comment on 5.1, I'm just a plain ol' stereo broadcast man!

[zboy2854]

Originally posted by ashdot:
We're finishing our first 5.1 project and we started bouncing the mixes today.
I was shocked to listen to the resulting files. They sounded dead comparing to the original PT session, and worst of all, the 2 surround channels are about 5db lower in level!

Does anyone have any similar experiences with bouncing 5.1 mixes?

Izhar
PT HD3, 5.3.1.

<font size="2" face="Verdana, Arial">Izhar,

What were you listening to your bounced 5.1 mix on? Did you just re-import the bounced files into a Pro Tools session to listen, or did you encode them and listen to the mix on another format or system? If you encoded them, you should be aware that many AC3/DTS encoders' default settings are set to encode the surround channels 3db down, which would explain some of what you're hearing. If you're not playing the bounced mix back inside Pro Tools, whatever player or system you're using to play back the mix should be checked to make sure it's not set with the surround channels at the -3db default.