utilizing 24 bit in CEP 1.2

[secretorchard]

If I go to Options:Settings (F4) before opening a new file in edit mode, I can see that my sound card supports up to "24 bit packed," as well it should. Now..when I go to open a new file by File:New, I get a small menu of options for the sample rate (44.1 kHz), whether I want it to be a stereo or mono track, AND the resolution, under which I have the option of 8 bit, 16 bit and 32 bit float. Ok folks...how do I record in 24 bit, as my sound card supports? Pretty elementary question, huh?

 

[dobro]

Select 32 bit float. Cool operates in that (which is better for internal processing than 24 bit), and if you need to send a 24 bit wav out, you can convert it before you do so. Otherwise, you'll convert your 32-bit files to 16 bit before burning to CD.

The 24 bit packed option is what Cool will convert the 32-bit file to *if* you do the conversion. Otherwise, it'll happily operate at 32 bit float. I bet your soundcard won't have a problem with that - mine doesn't.

 

[camn]

this is whats happening.. if your card is anything like mine...

INPUT... a 24-bit word comes in... CEP adds 8 bits of zeroes on the end. At this point the file is technically still a 24 bit sample....Then.. once you make any adjustment in CEP (volume, whatever...) it suddenly becomes 32bit... with any rounding errors in the last 8 bits.. keeping your 24 bits pure.

OUTPUT.. if you are playing back an UNMODIFIED 24-bit sample... it is unmodified... if you are playing back a 32-bit sample.. your card TRUNCATES the last 8 bits.. delivering a 24-bit sound to your amp (which then degrades it to, like 4-bits j/k)... and there you go. The chopped 8 bits generally represent all of your error.. but some is introduced since youre chopping it. (for instance a .....49 becomes ....40 instead of .....50, only in binary.. not base10)

OUTPUT PART 2... when you dither down 16-bit... normal dithering problems apply... but generally its better than chopping.

xoxo