What is "24 bit fixed-point padded to 32 bits"?

Apr 30, 2004 at 5:14 PM Post #3 of 10
The driver deals with 16 and 32 not 24 so that's why you must pad it. Padding means adding 0s. A 24bit DAC takes 24bits of input and gets rid of the extra 8 zeros.
 
Apr 30, 2004 at 5:21 PM Post #4 of 10
So if I want my 44.1/16 music to be as unaffected as possible in foobar2k, and my ASIO driver only accepts 32bit output, I should chose 16bit padded to 32bit?
 
Apr 30, 2004 at 6:23 PM Post #6 of 10
Quote:

Originally Posted by Haribo
So if I want my 44.1/16 music to be as unaffected as possible in foobar2k, and my ASIO driver only accepts 32bit output, I should chose 16bit padded to 32bit?


If it works, go for it.

Quote:

Originally Posted by mshan
Also, is padding supposed to improve sound quality or is it simply a trick to make foobar function properly with various sound cards?


Nothing to do with sound quality. I think it makes things work with ASIO mainly. I don't think the other output methods complain.
 
Apr 30, 2004 at 6:23 PM Post #7 of 10
It puzzles me too, seeing that the au2 is promoted as a 24 bit DAC, yet I use 32bit fixed point mode in foobar, and it works great using KS...
confused.gif
 
May 1, 2004 at 3:53 AM Post #8 of 10
The thing to keep in mind is that you don't get any additional increase in precision so the sound quality is still limited to what 24 bits will provide. Padding up to 32 bits with zeros does not actually add actual sound information, just zeros to make the data format compatible with the output device.
 
May 1, 2004 at 9:57 AM Post #10 of 10
well, i dunno how it's applied to your soundcard... but when you're doing lots of processing, internal 32-bit processing can help with generation loss. that is, as you apply more and more effects, you are able to save a couple more bits of information ultimately. but if you're just talking about your driver... yeah, it's probably more for compatibility purposes.
 

Users who are viewing this thread

Back
Top