Re: 32 bit float truncated to 24 bit int with no parameters

On Sep 12 22:45:27, [email protected] wrote:
> However, when I try to save my file as 32 bit float, the audio is truncated
> to 24 bit int.  Even if I have no parameters and just input output a 32f
> file, the audio is truncated to 24 bit int.  Here's the command, where
> in.wav is a 32 bit float file with content more precise than 24 bit int:
> 
> sox "in.wav" "out.wav"

I cannot reproduce this:

$ sox -n orig.wav synth 1 sin gain -6
$ sox orig.wav copy.wav
$ soxi orig.wav copy.wav

Input File     : 'orig.wav'
Channels       : 1
Sample Rate    : 48000
Precision      : 32-bit
Duration       : 00:00:01.00 = 48000 samples ~ 75 CDDA sectors
File Size      : 192k
Bit Rate       : 1.54M
Sample Encoding: 32-bit Signed Integer PCM

Input File     : 'copy.wav'
Channels       : 1
Sample Rate    : 48000
Precision      : 32-bit
Duration       : 00:00:01.00 = 48000 samples ~ 75 CDDA sectors
File Size      : 192k
Bit Rate       : 1.54M
Sample Encoding: 32-bit Signed Integer PCM

Total Duration of 2 files: 00:00:02.00

> To be clear, the output file IS a 32 bit float file,
> but the audio is truncated to 24 bit.

Meaning what, exactly? A 32bit float is a 32bit float.

If the files are small enough,
can you post them somewhere to be looked at?

> I have been using Bitter VST in iZotope RX to view the actual depth of the
> audio file.  One more thing, saving a 32 bit float as a 32 bit int file is
> truncated as intended.  It's only when saving as 32f that the audio is
> truncated to 24 bit int.

I don't know what you mean by "truncated"?
Please post an example file.

> Here are some screenshots of Bitter to ease your understanding (note that
> foo.wav is out.wav): > https://imgur.com/a/wWlwk6T

I have no idea what those images are supposed to show.

Jan