I was sent a .wav file which I wish to edit in Audacity. When loaded, the file identifies itself as "32-bit floating’. However, its .wav header simply gives the sample depth as ‘24 bits’. By editing in Audacity I changed the format to 16 bits. If I compare the lengths of the two ,wav files, I find that the total length of the one that claims to be ‘32-bit floating’ is exactly 1.5 times the one that claims to be 16 bits. All in all, it seems that the one that claims the floating format is actually simple 24 bits. What am I missing?
By default Audacity converts uncompressed audio files to 32 bit float format when they are imported. This is a totally lossless conversion and is useful as it allows processing to be far more accurate. When you export your completed project you can set whatever (supported) file format you like in the file export options.
Yes, 32-bit floating has a 24-bit mantissa, but as a number it would be meaningless without a sign bit and a seven-bit exponent. That’s why a wave file would need to allocate the full four bytes to each sample, making it 33% larger. Also, its header would have to specify that that’s what it is so that a playing program could decipher it correctly. Hence my original puzzlement. Now, however, my problem has been resolved: the “32-bit floating” refers to Audacity’s storage method and not to any file.
I tried as an experiment, Exporting something as 32 bit to my Desktop, but when I looked at the Properties, I didnt see 32 bit, I did see 16-bit though. How can I save something as 32 bit and be sure? I saved it as “Other Uncompressed Files,” but did not see any Options box I was told of also—???
. But what is the difference between 32-bit Float and 32-bit PCM???