Help with WAV bitrate encoding.

I’m editing some flac files in audacity and I’m exporting it as “WAV 32 bit signed PCM”.
But I’ve heard that the best format is 32 bit float.
Is there any data difference between these two?
Am I losing some quality using the signed PCM?

Audacity uses “32 bit float” internally, so working and exporting in 32 bit float format ensures that the exported audio is identical to the data in Audacity.
“Signed 32 bit” is a very high definition format but not “identical” to 32 bit float.

“32 bit float” format supports sample values that are over 0 dB. “Signed 32 bit” can not represent values over 0 dB.

Help with WAV bitrate encoding.

FYI - This is bit-depth (number of bits per sample) as opposed of bitrate (number of bits in one second of audio).

I’m editing some flac files in audacity and I’m exporting it as “WAV 32 bit signed PCM”.

What are you going to do with your exported files? If you are going to do further editing/processing, there may be an advantage to 32-bit floating point. For listening, 32-bits is total overkill. (16-bits is better than human hearing… I think there is only one company that makes a 32-bit audio digital-to-analog converter, there are no floating-point audio ADCs, and most 24-bit ADCs are only accurate to about 20-bits.)

Am I losing some quality using the signed PCM?

No. Not as long as you are not doing something to push the peaks over 0dB (which you shouldn’t be anyway).

I think FLAC supports 24-bits maximum and you are not gaining anything by converting to a higher bit depth than your original file.

Also bear in mind that most applications can’t play WAV files larger than 2 GB, and the absolute maximum is 4 GB. 32-bit files (floating point or fixed point) are twice as large as 16-bit PCM. For 44100 Hz 32-bit stereo WAV, you will only get 1 hour 40 minutes playing time within the 2 GB pseudo-limit.


Gale

Thanks for the help guys.

I’m using the wav format as a intermediate after editing. After I will reconvert it to flac.
I think I’m gonna use the 32bit float for now on as it is a more safe encoding.