In Reply to: RE: I wonder... posted by audioengr on February 4, 2013 at 10:31:40:
"I believe that it has a lot to do with 192 versus 256. The test tracks were 256 right?"
The test tracks were actually targeted as being around 320kbs using the WavDiff variance measure. As I documented in the procedure, I did TWO PASSES through the LAME encoder using "free-format" and shutting off the normal low pass filter in order to maintain high frequencies (so it's not too easy for anyone to just pull up the files and look at the spectral changes from MP3, or running it through the TauAnalyzer and get MP3 verification). Imagine running a JPEG image through JPEG again and the errors that would cause.
Going through the 2-pass LAME transcoding process I believe made the audio sound WORSE than standard 320kbps because of temporal smearing to the frequencies plus handicapping the encoder by refusing to let it toss out inaudible frequencies. Again, imagine running an image through JPEG lossy compression *twice* and the distortion this would have on image quality.
Despite all this... Preference was still for the MP3 tracks!
-------
Archimago's Musings: A 'more objective' audiophile blog.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- RE: I wonder... - Archimago 02/4/1310:50:45 02/4/13 (2)
- RE: I wonder... - audioengr 11:55:53 02/4/13 (1)
- RE: I wonder... - Archimago 15:04:16 02/4/13 (0)