Sample rate/ bit depth?
Home > Home Recording Forum > Recording Techniques > Sample rate/ bit depth?
Posted on Nov 26, 2009 07:34 am
Fragile
Member Since: Sep 30, 2009
I've been curious for some time now, what should i set the sample rate and bit depth for? the standard is 44100hz and 16bit. For some reason late one night i decided to switch to 24 bit until i was downloading a free piano sampler (CVPiano, not real impressed). When the UX2 was set on 24 bit, major static "distortion" kept going on. But i was wondering if there's any other disadvantages to higher bit depth besides large file sizes? Also, is there much advantage to the higher depth or am i wasting space here?
[ Back to Top ]
pjkPrince CZAR-mingMember
Since: Apr 08, 2004
Nov 26, 2009 08:19 am I posted a link on here a while ago, relating some people's ideas on that subject, and the results weren't highly in favor of higher rate/depth.
Many (me included) have used 24bit, and 48khz. I think the 24 bit gives you more headroom to work with, but i'm thinking 48khz may not be worth the dither.
Hopefully others with greater wisdom will chip in some info.
Nov 26, 2009 11:31 am The *delivery* standard is 44/16. The *recording* standard is 24-bit, usually at the target rate or a multiple of the target rate.
If you look at the typical polls - Online, professional and othterwise, you'll find that (depending on the poll) anywhere between 70 & 80% of full-time industry professionals record at the target rate. About 95% record in 24-bit. And yes, the damage that can be done by simply resampling the signal can more than destroy any frequency response advantage that you may have had before.
24-bit offers more *downward* headroom** ('foot-room') and ridiculously higher resolution (theoretically, 256x the resolution of 16-bit with 16.7 million possible points of resolution vs. 16-bit's scant 65,000 possible points).
** 0dBFS is still 0dBFS. 1.23vRMS is still 1.23vRMS. Line level does NOT change - A converter calibrated to -18dBFS=line level is still going to be the same whether that converter is converting to 16 or 24 bit data. But one of them is going to capture that data with 256 times the resolution of the other and 48dB more theoretical dynamic range.
Why you were having issues at 24-bit, I'm not sure. You can almost always mix word-lengths - even in a single project (as most systems are going to throw calculations in 32 bit - or higher, floating integers, etc.). Maybe the 'hiccup' was due to switching while something was active...?
Nov 26, 2009 01:55 pm Nah, it wasn't a hiccup :\ full blown distortion on every note i played, higher velocity meant more distortion. Pretty ugly.
But it doesn't really matter, that was certainly CVPiano's fault. Somehow the programmers had a glitch with 24 bit data, or maybe in my installation a file got corrupted. Either way, no biggie. I didn't like the program much anyway. A free sampler wasn't likely to be as good as i was hoping. Too many annoying pops and hisses or several out of tune notes. I'll use the line out on my yamaha or other synth-based piano.
Anyway, i don't quite know enough to understand much of what you just said lol. But it sounds like, in time, my ears will learn an appreciation for the more options in a higher bit depth, so i might as well start training them now. :] thanks!
ps. my post earlier was stated poorly. I meant the "standard" setting on my UX2 was 16bit, and i changed over to 24 out of curiosity. but changed back to 16 when i got CVPiano, and i'm now changing back to 24.
