Yesterday i showed this thread to some professional russian audio engineers (they all work on Protools hardware, Pyramix, Sadie and other exotic platforms), they confirmed this topic - cpu bit and os bit has nothing to do with audio precision.
As a professional programmer, I can chime in and add my own voice to the chorus of people saying that the host has nothing to do with dynamic depth. I code 32-bit applications all the time, and there are absolutely NO (practical) limits regarding my use of 64-bit variables or data streams. Bit_Arts’ confusion seems to come from not understanding that 64-bit data types can exist and be tossed around in a 32-bit OS, and that this has been possible for ages.
In other words, a 32-bit VST with 64-bit audio processing will both use 64-bit audio internally, in its communication with the host, and within the host itself (unless the host deliberately downscales the signal to a different bit rate, but that’s a different topic entirely). I guess it’s entirely possible that someone would be able to present to me an application that fails to qualify in that it somehow delivers a 32-bit signal to the host despite using 64-bit processing internally (on a 32-bit OS). That would be the result of a bug within the VST (or host application), or even a design decision. Whether or not the OS is 32- or 64-bit doesn’t even factor into the equation and would have no effect on the process.
not gona lie, im a massive audiophile. Ive mixed certain moggs down to 64 bit wav before and it was like cocaine in audio form. i know it would be completely useless in terms of serious prodction, as when you make it a 16 bit mp3 it gets rid of all the quality anyways, but for personal uses, it would be awesome.
I guess i can just render all the tracks in my song as separate files then render them together in 64-bit wav with Audacity or something.
There is no difference between 32-bit and 24 bit, if there is no clipping in either file. 32-bit = 24 bit with clipping info. (which means that destructive data is stored with a 32-bit file whereas a 24-bit file would require a new remix/remastering if it is rendered with clipping)
Interesting, didn’t know that. So what is the difference between 32bit and 64bit audio then?
Anyway, my point was that one is not able to hear the difference at all. I mean really no one will be able to tell in a proper double blind test!