the OP was just wondering if he had to redo some work. he's dealing with dozens of songs. perfectly OK question to ask.
as for the tech side, it's fun to know what's up. the ones who are relying on techshit and not using their ears are in the 'Spectrum screws up my mixes' threads.
Tarekith wrote:Tone Deft wrote:Tarekith wrote:32 bit files have the same dynamic range as 24bit files, the extra 8 bits are just there in case you go over 0dBFS. Quick and simple answer.
I'd doubt ANYONE can hear the difference between 24 and 32 bit.
don't think so Mr. T.
by your logic 24 bit files have no headroom?
proportionally they have the same amount of headroom.
even by your definition 32 bit files have more dynamic range, headroom is counted in dynamic range because it's the range of values the audio level can have.
I was under the impression from various reads that a 24bit file and a 32bit both use 24 bits of precision (ie, approx 144dB), however the 32bit file uses the extra 8 bits as part of the floating point calculations. It's these extra 8 bits of FP processing that can allow you to technically go over 0dBFS without clipping. I'm basing this off of information from threads like this:
http://www.gearslutz.com/board/masterin ... float.html
http://www.gearslutz.com/board/masterin ... jects.html
I fully admit I'm no math whiz though (the word cosine makes me run for the hills), so if my laymans explanation is lacking somehow, I'd love some clarification.
the guy who said they use the same 24 bits on KVR was wrong, a lot of the other posters had good info.
first off, it's all arbitrary, there is no standard to adhere to and Live's headroom and technical specs in that area aren't published. even their white papers leave this out, they're not really white papers, they're more tutorials. good info, not what I was hoping for.
aaaanyway... the straightforward way to design something is to say that it has 'x'dB of headroom, regardless of bit depth. suppose that's 10dB of headroom. with 24 and 32 bit words that's going to be a different number of bits but it's not just the top 8.
Live probably normalizes for just a few dB of headroom, maybe 3dB or less, I don't know.
is that enough of an explanation without getting into the math and stuff? cool thing is, once you do the math a few times the concepts become clear and you don't have to do the math anymore. I know you know the concepts, does this help? FWIW my whole college experience was math, literally thousands of pages of math homework, shit, I can't do ANY of that anymore, but the concepts stuck.
GAFM ***
