Is there a point to export to 24bit 96khz before mastering?

[peteman]

Ok, here's the situation:

I'm finishing a hour long ambient style mix and I'm close to the point where i need to export it to wav and start teh mastering process in a separate audio editor. The mix consists ot 10 songs that has been made earlierr and all of them are in 16bit 44khz wav format and I've mixed them together and added in some synths and effects from variouse VSTi's.

Question is, does it matter if I export the mix to 24bit 96khz wav or 16bit 44khz? What's the benefit of 24bit audio in a situation like this (if any)?

[dubbelQ]

There's no point export them as 96khz. You dont win anything soundwise.

[peteman]

ok, but how about the 24bit?

[C00kie]

What's the benefit of 24bit audio in a situation like this (if any)?

Ever heard of quantization errors?
They do build up if you follow a scenario like this:
1) rip original CD track as 16bit audio
2) lower volume by a few dB to match other tracks
3) save as 16bits (least significant bits are truncated or rounded off)
4) import for mastering
5) apply some compression / limiting (which raises volume by some dB again)
6) save as 16bits

If you had saved as 24bits in step 3, then no truncation would have occurred. So in mastering you'd get something closer to the original.

Weather you can realy hear it, that's a different story ofcourse. But as a general rule: never save as 16bits until it's finally finished. That is to preserve the tiny little details.

[jupiter8]

In theory the stuff you add (synhts and effects) will benefit from going 24 bit and even 96 kHz but don't be surprised if you don't notice any difference. At all. Or it could be quite a noticeable difference depending on the plugins you use and how much you use them.

[peteman]

What's the benefit of 24bit audio in a situation like this (if any)?

Ever heard of quantization errors?
They do build up if you follow a scenario like this:
1) rip original CD track as 16bit audio
2) lower volume by a few dB to match other tracks
3) save as 16bits (least significant bits are truncated or rounded off)
4) import for mastering
5) apply some compression / limiting (which raises volume by some dB again)
6) save as 16bits

If you had saved as 24bits in step 3, then no truncation would have occurred. So in mastering you'd get something closer to the original.

Weather you can realy hear it, that's a different story ofcourse. But as a general rule: never save as 16bits until it's finally finished. That is to preserve the tiny little details.

Never heard of the quantization error, but thanks for the link. Everyday I learn something new in the audio world

I'll go with the 24bit. Even though I'm pretty sure I DONT here the difference, but at least there's no risk losing the details.

thanks.

[peteman]

In theory the stuff you add (synhts and effects) will benefit from going 24 bit and even 96 kHz but don't be surprised if you don't notice any difference. At all. Or it could be quite a noticeable difference depending on the plugins you use and how much you use them.

ok. thanks.

[Sequent]

I would say definitely go with the higher bits. Some editors like Wavelab, for example, can even handle 32-bits. Depending on the plugins you've used and also your monitors you may indeed hear the difference.

For one... if you have a large number of tracks on mixdown, the extra bits should offer more headroom and clarity.

Secondly, during the mastering process... if you're running at 16-bits, you only have a certain number of operations before you begin to degrade your audio. If you go for 24-bits then you shouldn't have to worry about it.

Consider also that even though you might not hear the difference now, it's entirely likely that at some point in the future you may or may not (probably will) have better monitors (or a better soundcard) than you have now. Or you may have to hand off your files to someone like a pro mastering house that probably will have a better system and then the things that you're not hearing now, might suddenly be an issue.

So, I think it's best to always go for the higher rate. Unless your computer can't handle the files. But for a stereo file for mastering it's shouldn't be an issue.

Well... that's just my opinion and also based on my own experience.

[bduffy]

Yeah, I'd say go with at least 24-bit whenever you can, even when you were mixing those old tracks. As said, the extra bits will help preserve clarity and headroom. Export 16-bit only at the final mastering stage.

[Drillmouth]

You will gain nothing. As everything started at 16bit you can't turn them into 24bit. Same goes for sample rate. You can only ever reduce these things. You can perform the action but it will make no changes as the audio is already quantized.

[timobrien]

Tell me which of these numbers is more accurate:

1

000000000000000000000000000000000001

They're both 1

All you do when upsampling is add zeros in front of the number.

Most audio programs natively work their math in 32bit numbers anyway
so the result is you still get: 1

You SHOULD be working in 24bit to start with anyway
and there is little reason to go above 44.1khz unless you work in video (16bit/48khz final files) or with hardware that is 48k native.

Lock everything down on 24bit/44.1khz and forgetaboutit.
Convert to 16bit/44.1khz at the FINAL step.

[tenshin111]

You will gain nothing. As everything started at 16bit you can't turn them into 24bit.

You're wrong here. Most (if not all) modern audio interfaces record in 24bits.

[tenshin111]

Tell me which of these numbers is more accurate:

1

000000000000000000000000000000000001

They're both 1

All you do when upsampling is add zeros in front of the number.

Most audio programs natively work their math in 32bit numbers anyway
so the result is you still get: 1

Ever heard of floating point (!) numbers and rounding errors?

[jupiter8]

You will gain nothing. As everything started at 16bit you can't turn them into 24bit.

You're wrong here. Most (if not all) modern audio interfaces record in 24bits.

And this is relevant to this discussion,how exactly ?

The mix consists ot 10 songs that has been made earlierr and all of them are in 16bit 44khz wav format

[annode]

The best way to get better audio quality on the rendering is to import your 16/44 to 32/96 then work on your project...assuming you have the CPU strength to do it.
(There's no truncation errors added at 32bit as well...although it's the sample rate that will give you improved VST/VSTi quality, and not the bit rate in ambient music.)

Many or maybe most VST(i) run their background processing in 32bits by default...you will get an upgrade in audio performance when the VST/VSTi is run at 96khz.

I know if you change the sample rate in your project from 44.1 to 96, for example,all your midi tracks and audio parts will be moved and displaced. If you render a 44.1 project to 96khz, I don't know if it is the same as importing 44.1 tracks into a 32/96 project from the start.

It's worth searching your manual for that answer.

PS: and remember to place a dithering vst in the last slot of your master out buss.
Cubase offers the Apogee UV-22.
http://www.apogeedigital.com/products/uv22hr.php