[music-dsp] 32 bit floats >>> 24 bit integer

Dave Gamble signalzerodb at yahoo.co.uk
Fri Jan 20 09:19:04 EST 2006

On 20 Jan 2006, at 12:27, kernel wrote:

> On 20 Jan 2006, at 11:33, Dave Gamble wrote:
>> On 20 Jan 2006, at 08:11, Nigel Redmon wrote:
>>> That's why I hedged "in general" and "if you do you final  
>>> adjustment for audio level before the conversion". Almost always,  
>>> in a case where you need to do what the op is asking, you are  
>>> doing your computing in floating point, and converting so that  
>>> you can send it to a D/A for output. I suppose the "don't bother"  
>>> is good for 99% of the cases where this question is asked, and  
>>> for the other 1%, 99% of those cases no one could tell the  
>>> difference under the best of circumstances.
>>> I'm trying to think of a case where audio in general (not a  
>>> specific repeating signal) could get stuck on a limit cycle with  
>>> a truncated 24-bit signal versus a dithered version of the  
>>> same... but anyway, the answer was assuming that the conversion  
>>> wasn't followed by further processing.
>>> I have to think that the vast majority of music and recording  
>>> software, synths, and sound processors in the real world don't  
>>> bother--but hey, if you have time and motivation, it doesn't hurt.
>> And, of course, this is eventually why bits 25 and 26 will be  
>> needed...
>> The processing gets more complex and processing  chains get  
>> longer, so eventually, someone can hear that no-one is bothering  
>> to dither anywhere.
>> Multiply up by 128 tracks and wonder why things sound bad in a  
>> strange way you can't quite put your finger on.
>> Swap over to fully 32-bit float, and convert to integer right at  
>> the end, and the problem disappears...
>> Pro Tools TDM mixes vs Cubase/Logic mixes have been shown to be  
>> different...
> this is partly where I'm coming from.  my own software writes audio  
> tracks to 32 bit multichannel float files.  these files are going  
> to end up in my DAW (protools) which only takes 24 bit integers.   
> so there *will* be further processing via EQ, combfilters etc.   
> I've been known to dither each individual track before now but I  
> didn't much like the results.
Have you considered switching to RTAS? That should solve a lot of  
problems; that way you're passing through as float. As we know, this  
has better dynamic range,
and that's useful for your situation.

Another thing to bear in mind is that there's nothing for free -  
dithering in its simplest form means adding noisefloor. Even POW-r is  
a special type of distortion+noise system.
I'm not surprised that dithering every track didn't give quite the  
result you hoped for.


NEW Yahoo! Cars - sell your car and browse thousands of new and used cars online! http://uk.cars.yahoo.com/

More information about the music-dsp mailing list