24 bit 96khz best to work with?

[TheElement]

I've been looking at different file formats and understand 24 bit 96khz is a popular lossless file format.

We have the option of working inside our DAW and also exporting mixdown in this 24 bit 96khz file format. And also the 32 bit float 96khz.

Would this be considered the best file format to work with?

Thanks!

[Len911]

I've been looking at different file formats and understand 24 bit 96khz is a popular lossless file format.

We have the option of working inside our DAW and also exporting mixdown in this 24 bit 96khz file format. And also the 32 bit float 96khz.

Would this be considered the best file format to work with?

Thanks!

There is no benefit in re-sampling for the sample rate of 44.1 khz or 48khz from what your sampled instruments or VI's are, to 96khz. To my knowledge, no vi's or samples are 96khz.

32 bit float gives you a few more bits when you have several 16 or 24bit tracks, more decimal places for accuracy in adding effects when the math is performed??

I would say 32 bit float 44.1 or 48khz, that's if you might need to import it for further processing, otherwise if it's finished, dithered down to a 16-bit.

I suppose if you recorded with a mic or analog to digital conversion into a 96khz project.


http://www.trustmeimascientist.com/2013 ... n-it-isnt/

[mobster85]

For televisión they want 24 bit and 48khz. Thats What the library i signed with wants.

[TheElement]

Thanks guys. some real good points there. Guess I'll stick with what I've been doing.

[mojobone]

Theoretically, 24-bit, 48kHz is all you'll ever need, and can represent evenly and adequately all the frequencies a human can possibly hear, but as we all know, the difference between theory and practice is that theoretically there's no difference. Certainly, higher sampling rates and greater bit depths can't hurt anything and if you remember to dither only once and from the highest resolution files available, you get to keep much of any extra detail the extra file size buys you. When your DAW operates at 32-bit float, that's the length of a word, and float means as many decimal places are used as are necessary-it's more accurate when calculating, because there's less rounding off-some folks claim they can hear digital truncation, which is most likely poppycock. Better to take good care of your signals going in, and let the dither fall where it may. Tracktion has been 64-bit float since version 2.0, so I never got angst-y about this stuff; I can generally deliver music in the formats most often required and like Joe Meek said, if it sounds good, it is good.

[mojobone]

In the interest of giving a more complete answer, I gotta add a couple items; one, that there are times when upsampling/higher sample rates are worthwhile. It's been reported that some DAW plugins work better/sound better at higher sample rates, and I consider it theoretically possible, verging on plausible. If you're archiving/futureproofing masters that are of actual value, consider storing and backing up at the highest possible sample rate, but you probably don't need more than 24 bits at any juncture for PCM digital files. Direct Stream Digital, or DSD is another matter; it's 1-bit, with exponentially higher sample rates, and generally only mastering DAWs can use it.

Two, when shopping for an interface, (or a DAW, for that matter) you want one that's capable of operating at 192kHz, because that buys you greater precision at all the other sample rates, though some engineers will continue to claim otherwise. You also want word clock I/O, cuz that tells you it's not a consumer model.

Three, there are sample libraries that were captured at higher sample rates, but it's difficult to discover which ones, (unless it's in the promo materials) because of proprietary data compression formats .

Four, if you're interested in all the gory details, pick up a copy of Bob Katz' masterpiece, Mastering Audio; The Art & The Science.

[garywhite]

48kHz is the highest sample rate I've ever been asked for, whether for TV, Film, Library, whatever. I work in 24bit as standard, but if I know that my client definitely only wants 16bit, why waste processor power and drive space? I'll just work in 16bit 44.1kHz. I've never noticed any real difference anyway. My MOTU interface is capable of 192kHz, but I've never needed to or been asked to run at 192.. Good converters at 44.1 will probably sound sweeter than crappy ones at 192. That's my take, Gavin. All the best

[Len911]

http://bcove.me/b4fep7ct

This is one of the best videos or papers I've seen on ADC's. The Dan Lavry white papers is another one.
http://www.lavryengineering.com/pdfs/la ... _audio.pdf

[TheElement]

Thanks for all that info guys!

When I upgrade my DAW I'll start running it at 32 bit float 96khz internally.

[cardell]

For televisión they want 24 bit and 48khz. Thats What the library i signed with wants.

That's what I use as well.

Stuart