question... quality comparison between Opus 256 and 512 about opus

[tdaede]

Subtracting the compressed and original audio isn't useful to compare audio codecs for many reasons. As an example of one reason, Opus (and all good encoders) take advantage of masking - a psychoacoustic property where loud sounds hide quiet sounds. If you subtract the two, you lose this property.

A much better methodology is the one done by this test, for example: http://listening-test.coresv.net/index.htm

And the results: http://listening-test.coresv.net/results.htm

For very high rates, you can also do a much simpler ABX test, which just tells you whether you can hear a difference or not.

You might also be interested in the HydrogenAudio listening test forum: https://hydrogenaud.io/index.php/board,40.0.html