should i resample a 16 bit file for mastering?

[eugene]

when you guys master a 16/44.1 bit mixdown, do you resample it to higher 24 bit rate and\or at a higher sample before editing it in wavelab?

thanks in advance!

[bluelife]

mixdown in 24, resampling wont make it better.

cu

[unic]

That may already be adressed with your temp_file bit-depth settings, so you may not need to resample the original file. As always, RTFM is the best bet to avoid guessing, but I assume WL would do it's internal calculations at temp_file bit-depth.

If 16-bit file is your only source (i.e. if you want to remaster an old song), I think what I mentioned above will do o.k.

[pHz]

mixdown / render at the highest resolution your setup can handle ... edit ... THEN dither down to 16 / 44.1 for CD burning ...

slainte rob

[M'Snah]

I'm not sure if I understand your question.

If you have a 16 bit mixdown, e.g. the output of your host is a 16 bit file, then it's no use to turn it into a 24 bit file before mastering.
A better approach is to (always) save / freeze / render your files in 24 bit or 32 bit in the host. Even when you only use 16 bit samples, after processing them (even just changing gain) a higher bitrate makes sure that you don't lose any of the bits...

After I learnt about bitrates, everything I do is 24/32 bit, and only down to 16bit when mastering...

[Amberience]

I mixdown to 16bit... it really does not matter. All this technical crap is fine and dandy in the geeks realm, but the musicians? The artists?

Meh... dicussions over 16 or 24bit files are the reason shit music exists. Too much focus on doing it the "right" way and not enough focus on doing it the RIGHT way.

[respirator]

Some say that upsampling prior to mastering to a higher sample range, eg. 96, helps. Supposedly the mastering algorithms have more "headroom" or something. It makes sense in a mathematical way (higher resolution = better processing) but I'm not convinced that I can hear a difference (I do most in 88.2/24 btw). One thing that is very important though is dithering when reducing bit rate.

[unic]

The easiest for the wave-editor and the cpu is 32-bit, no dithering ... it isn't that good for your storage, esp with long, expressive, artistic, innovative etc. pieces ...

[dgkenney]

Too much focus on doing it the "right" way and not enough focus on doing it the RIGHT way.

Man...when you're right you're right!.

Dan

[M'Snah]

My personal experience is that music sounds a lot less flat when I mix and master it all in 24/32 bit and only go to 16 to create an mp3. And the additional diskspace: Twice as much as 16 bits could mean 40-80 Mb extra, on a 150 Gb disk...

And 24/32 bits ensures that you don't need to go to all kinds of toys to enhance / brighten / excite the final mix, because you didn't take the brightness away in the first place

[IIRs]

I mixdown to 16bit... it really does not matter. All this technical crap is fine and dandy in the geeks realm, but the musicians? The artists?

Meh... dicussions over 16 or 24bit files are the reason shit music exists. Too much focus on doing it the "right" way and not enough focus on doing it the RIGHT way.

I really don't understand this attitude. if your music sucks, it will suck just as hard at 16-bit as at 24 or 32. In what way does setting your record resolution to 24-bit affect your creativity?

[airon]

Mixdown to the highest bitrate your system spits out, which is usualy either 24 bit integer PCM or or 32-bit Float.

Compare this to the graphics realm. Are some of you nuggets actualy telling me, you'd rather hand in a 1 megapixel picture for editing/publishing/printing than a 7+ megapixel photograph? A rough analogy, shure, but you're sacrificing detail.

If so, you ain't pro, in part because you don't know when to listen to professionals, who do know why taking a highres path when you can is a good idea.

-edit-
Sorry for sounding so harsh, but this kind of aggravates me. Going for highres mixdown will not cost performance, and disk space is a non-issue.

Btw, if your recorded material is at a certain sample rate, it's usualy a good idea to keep your project and the mixdown sample rate at that value as well. Do as little resampling as you can, because that will cost performance and introduce more artifacts.

[cron]

I only use 24 bit if I know I'm well below my headroom and I need to normalise. I can happily sit a good 20 dB below the red at 24 bit and know that it's still going to sound as good as a 16 bit mix peaking at zero once I've normalised. Gives me more time to make music instead of anally watching my mixers for clipping.

It's not the reason 'professionals' mix at 24 bit, but whatever.

[Evan]

I think Acolmiztli means that many musicians nowdays seem to count music quality way to much based on numbers rather than the actual content. Engineers do that because it's their job, but I don't think it's ideal for a creative musican to go crazy over getting their maths perfectly right. That's what he means I think, and I tend to agree.

Musicians and engineers are two totally different things but you can have a bit of both nowdays. It's just that the one tends to interfere with the other. At least for me it does. I really wish I could temporarily forget all things related to engineering when I am trying to use my creative mind.

[Peel]

OK Eugene, you have ONLY the 16/44 mixdown, with no option to make another mixdown to a different format, correct?
I don't know about wavelab, but most hosts operate internally at a bit depth of 32. In other words, your host is converting your file to 32 bits on the fly. That's why there's no benefit to converting the file to a higher bit depth (although your file still has the quality of a 16-bit file -- nothing can change that).
With regard to sample rate, many (most?) mastering plugs have internal oversampling as well, so there may not be any benefit there either.
Personally, I would always try to work in 32-bit (but 24 would probably be fine) but I don't worry about going above 44.1.