Quality Loss When Mixing Down

  • Thread starter Thread starter laj35
  • Start date Start date
laj35

laj35

New member
Hello,

I did a search of the BBS and couldn't really find a definitive answer, so...

When mixing down (I use Cubase 5.1, if it matters) what type, if any, of loss in sound quality is there? Does it matter how many tracks you have? The # of insert/master/directx fx used matter? I've heard from some that you only want to mixdown once because there is a significant loss, others have said it's undecipherable, so whatcha think pros?!? Thanks in advance for your thoughts.



Laj
 
The "only mixdown once" thing is silly. You don't *lose* anything in your original tracks.

While it is possible for one application to sound better than another, typically a mixdown will be exactly what you hear when you're working in the application. The numbers in the mixdown file will be the same numbers generated by the mixing software while you're working in it, they're just "precomputed", if you will. Granted, it is possible to write an application where a mixdown is not of the same quality as the mix, but in real life that would be an *error*.

Slackmaster 2000
 
Thanks Slack2k, that's what I thought but I just wasn't really sure, thanks for clearing that up for me.



Laj
 
laj35 - it sounds like someone familiar with recording on analog tape was talking to you. When you are recording in analog and you "bounce" tracks from tape to tape and back again, you do introduce noise and degrade sound quality with each generation. But this has no bearing on recording digital info, unless there is a hardware problem or some sort of destructive compression is used (like in digital mini-disks) there should be no lose of information at all.
 
If you are mixing down with dither, you are introducing some noise with each successive mixdown. If you mixdown without dither, then you are losing some defenition unless you mixdown to 32bit files.
 
Yes, but that's not necessarily "quality loss", because you are already monitoring the dithered down audio mix when you're mixing (assuming your mixing application uses dither to go from 32bit to xbit). Therefore you're making your decisions based on the 24bit output, even though the internal processing is done at 32bit. You technically lose precision regardless.

Also, dither is indeed the process of introducing noise when decreasing bit resolution, but it's smart noise, not like "shhhh" kind of noise. Dither works on this simple principle:

Let's say your original samples look like this:

1.55
1.55
1.55
1.55
1.55
1.55

...and you drop one place of precision by truncation:

1.5
1.5
1.5
1.5
1.5
1.5

...or rounding you might end up with:

1.6
1.6
1.6
1.6
1.6
1.6

....but using dither, you instead might end up with:

1.5
1.6
1.5
1.6
1.5
1.6

which will sound closer to the original than the truncated or rounded samples. It's *GADS* more complicated than that, but that's the "why".

Slackmaster 2000
 
Yes, but that's not necessarily "quality loss", because you are already monitoring the dithered down audio mix when you're mixing (assuming your mixing application uses dither to go from 32bit to xbit). Therefore you're making your decisions based on the 24bit output, even though the internal processing is done at 32bit. You technically lose precision regardless.
Click to expand...


Yes, but we are talking about multiple mixdowns. If you import the mixdown back into a song, and mixdown again using dither, you will have dithered twice. Probably it makes very little difference, but technically, more noise is introduced with each mixdown, or if you don't dither, you lose a little bit of info.
 
Oh right, like bouncing...you're absolutely right. Hopefully you won't have to do that much on a good DAW, but it's a good point to consider.

Slackmaster 2000
 
You must log in or register to reply here.
Back
Top