My professor stated that although digidesign and some others say this doesn't happen. Most professionals find it better to mix down to a two track device instead of bouncing because of some loss of quality. I think I didn't clarify what I was saying and the word "ruined" is a little to intense compared to the real extent of the damaged done.
bouncng down ruins mixes.
- Thread starter DAS19
- Start date
My professor stated that although digidesign and some others say this doesn't happen. Most professionals find it better to mix down to a two track device instead of bouncing because of some loss of quality. I think I didn't clarify what I was saying and the word "ruined" is a little to intense compared to the real extent of the damaged done.
Ethan Winer
Acoustics Expert
If it really happens, and can be heard reliably with headphones, then - again - something is broken. I mix and render Wave files in SONAR all the time, and the file renders sound exactly like what I heard while mixing. It makes no sense for the "bass to get lost" in a Wave file render. This is also easy to prove by nulling files etc.
Acoustics issues are real and need to be addressed if you intend to do pro quality work, but perception is a big problem too. If you play the same file three times in a row, it can sound different each time just due to the frailty of our perception. This is why you need a blind test to know for sure.
--Ethan
noisewreck
New member
I have never heard any difference on mixed down tracks in Cubase.
Waffleness
New member
Me neither. How about mixing it down in real time, then comparing what you just heard to what got exported?
mshilarious
Banned
Very simple reason for this phenomenon:
When you bounce a final mix to disk, are we talking 16 bit, 24 bit, or 32 bit float mixdown file? If you are bouncing down to 32 bit float, and it sounds different than playback of the mix, then your DAW is broken.
However, if we are talking a 16 bit mixdown, of course it's going to sound different than live playback of the project. It has to. Your interface probably receives 24 bit data, and outputs that as best it can. But a bounce to 16 bit is either dithered or truncated. Either option will change the sound vs. the 24 bit playback you're used to (where you probably can't hear the truncation or dither).
Try this: take a mix and route it to a stereo subgroup. Make sure any bus processing is on that subgroup and not on the master. Bounce to 16 bit and then bounce to 32 bit float. Take the 16 bit stereo file, invert polarity, insert that into your project and route to the master (that is, not to the mixdown subgroup). Listen to the result. You should hear some very quiet noise, with your track even quieter underneath that.
OK, now do the same with the 32 bit float mixdown. If you hear anything at all, you either haven't matched levels properly, or your DAW is broken (or one of the plugs attached to it, or something).
If your 32 bit float test is a dead null, but you think it still sounds different, I'm afraid you are imagining that.
When you bounce a final mix to disk, are we talking 16 bit, 24 bit, or 32 bit float mixdown file? If you are bouncing down to 32 bit float, and it sounds different than playback of the mix, then your DAW is broken.
However, if we are talking a 16 bit mixdown, of course it's going to sound different than live playback of the project. It has to. Your interface probably receives 24 bit data, and outputs that as best it can. But a bounce to 16 bit is either dithered or truncated. Either option will change the sound vs. the 24 bit playback you're used to (where you probably can't hear the truncation or dither).
Try this: take a mix and route it to a stereo subgroup. Make sure any bus processing is on that subgroup and not on the master. Bounce to 16 bit and then bounce to 32 bit float. Take the 16 bit stereo file, invert polarity, insert that into your project and route to the master (that is, not to the mixdown subgroup). Listen to the result. You should hear some very quiet noise, with your track even quieter underneath that.
OK, now do the same with the 32 bit float mixdown. If you hear anything at all, you either haven't matched levels properly, or your DAW is broken (or one of the plugs attached to it, or something).
If your 32 bit float test is a dead null, but you think it still sounds different, I'm afraid you are imagining that.
NL5
Unpossible!
Aren't most people listening to 24 bits during playback, and bouncing 24 bit files? (the defacto standard for mastering) 32 bit float has to be truncated before it hits the converters anyway, right?
mshilarious
Banned
Strictly speaking, it needs to be dithered before truncation, but it doesn't really matter due to noise already present in the source.
I don't know exactly how people are working; one could argue you should export 32 bit float mixdowns for mastering. Again, probably doesn't matter, but if you are trying to minimizing processing steps that's what you'd do.
I don't know what people are doing that causes their bounces to sound different than their mixes, but a 16 bit bounce is going to be measurably different. Audibly different? Perhaps, depends on the dithering routine.
I can easily get Wavelab 5 to null 32 bit float bounces . . .