Page 2 of 3

Posted: Wed Mar 11, 2009 6:47 pm
by abZ
24 bit no dither is the standard.

I am honestly not sure what you are talking about in this thread. Ableton uses 32 bit floating. I don't think you can change that can you? If I was to record via mic or whatever (which I should do more of but I am lazy) then you just want the highest quality that your soundcard can do. As far as samples go, I don't really give a shit. I have used ripped real audio for samples before. If it sounds alright fuck it. I am not going to turn up my nose at a sick sample just because it is in mp3 or whatever but that is me. I have had a lot of comments from people telling me my productions sound really clean fwiw.

Posted: Wed Mar 11, 2009 6:52 pm
by -drix-
24 bit all the way

Posted: Wed Mar 11, 2009 7:07 pm
by Sharmaji
there's no point in working at 32 bit, fixed-point. most daw's process at 32-bit, floating point, so that you have a massive amount of internal headroom, but the final mix bus is down to 24 bit-- which still has a dynamic range beyond the human ear.

to get down to 16bit without ugly, underwater/noisy/warbly artifacts, yes, dither is a good solution. there's a bunch of different dither algorithms out there; i tend to like pow-r #3 for the majority but if you have the option, try 'em out.

there's really no reason to ever work at 16 in a daw. it really, honestly and truly, doesn't sound as good. all the discussions we've had about gain staging and headroom aren't really relevant in 16bit, where you want to work as close to 16 bit as possible (drop 6db and you're working at 12 bit, etc).

so really-- stay at 24 bit.

though bitcrush for fx!

Posted: Wed Mar 11, 2009 7:20 pm
by paradigm_x
abZ wrote: I have had a lot of comments from people telling me my productions sound really clean fwiw.
definitely.

Posted: Wed Mar 11, 2009 7:38 pm
by Pallms
24/44.1

Posted: Wed Mar 11, 2009 10:08 pm
by fattyknuckles
Botched wrote:16 bit is fine, :D
:o 16 bit sounds like ass.

Posted: Wed Mar 11, 2009 10:23 pm
by macc
Ben Freeman wrote:Oh, and if you send it out to be mastered, you would recommend 44100 24bit then right?
Give out 24bit (32 if you like but generally it is superfluous) at whatever sample rate you made it at, no SR conversions :)

@ alvin - dither is applied the last stage after all other processing immediately before bit depth reduction. Only use it when going from one bit depth to a lower one.

:)

Posted: Wed Mar 11, 2009 10:27 pm
by martello
Cheers Macc.

Posted: Wed Mar 11, 2009 10:54 pm
by -dubson-
errr i just use the reason default, probaly why my tunes always come out wrong :wink:

Posted: Thu Mar 12, 2009 2:14 am
by future one
24/48

Posted: Thu Mar 12, 2009 3:50 am
by manicmckanic
44.1/24bit. 96 with live recordings through me cakewalk sps-25 thingy

Posted: Thu Mar 12, 2009 4:50 am
by hd4000
only problem is i know some digital download labels dont accept 24-bit tracks.( i know of one for sure.)

plus with me i sequence everything in the mpc-4000.then record it into adobe audition.and audition gives me
8-bit
16-bit
or
32-bit float

no 24 -bit....

Posted: Thu Mar 12, 2009 9:39 am
by macc
Use 32-bit float if it is going for mastering or further processing. It is essentially 24-bit with an extra 8bits for information over 0dB.

If your audio isn't going over zero (ie clipping, which it shouldn't be) then there's no reason to use 32-bit FP over 24-bit other than canng your bandwidth/hard drive space.

:)

Posted: Thu Mar 12, 2009 9:44 am
by macc
Depone wrote:Just finished a 8 hour full band recording with a Protools Hd rig running 22 tracks I/O at 192khz/24. Sounded amazing with the studio PMC's and the Dynaudio near-feilds, But worried about the conversion stage when I come to bounce to a 44.1/24 .wav. And also converting to Red book standard (44.1/16) for CD.

At what stage should I be doing the conversion? And whats the best method of translating the audio from the 192 to 44.1?

Thanks in advance
I keep missing your AIMs man, so sorry. Been working beyond hard, even for me. feel fukt today!

I dunno why anyone bothers with 192 though, tbh... Anyway, it should go like this;

24/192 in the mix

24/192 stereo mixdown

mastering @ 24/192 (limit to around -0.5dB because....)

SRC to 24/44.1 (.....this process will affect your peak levels slightly, and you may wish to VERY (!) gently limit again)

dither to 16/44.1

:)

As for the best way to get from 192 > 44.1, well there are a million SRCs out there. Voxengo r8brain (full paid pro version) and the free SOX utility are my weapons of choice for that.

Posted: Sat Mar 14, 2009 3:00 am
by kizza2435
24Bit and 48kHz...just in case you get it pressed to wax :wink:

Posted: Sat Mar 14, 2009 9:44 am
by macc
You'll get bonus points for the 'peak levels changing during resampling' thing ;)

Let me know when my degree is ready :6: