In message , Paul
writes:
[]
https://en.wikipedia.org/wiki/Mp4a#A...ments_over_MP3
[]
In contrast to MP3's hybrid filter bank, AAC uses the
modified discrete cosine transform (MDCT) together with the
increased window lengths of 1024 or 960 points.

That doesn't strike me as an exact match. Transforming from

No, me neither. I'd thought maybe they were the same underneath, as so
many audio and video formats are these days, but sounds as if .m4a is
more modern (and thus almost certainly more efficient) than .mp3 is.
[]
A proclamation of "Try extract original audio stream" may be a
conversion from a compressed format to .wav, where the "try" is
certainly true. It still sounds like music. Then, running it

No, Pazera offers lots of different formats, including both
little-endian and big-endian WAV; "Try extract original" is at the
bottom of the list, so I am inclined to believe it does what it says. It
doesn't specify a type for that option, but most - possibly all - files
it has produced when I have selected that option have been .mp4 .

through the destination compressor, throws away content just
as the original conversion did. Neither is remotely similar
to the digitally recorded original PCM. You have two different
mutilations.

I grok that.
[]
So much for faithful reproduction.

Paul

Viewing on a spectrogram can be most revealing! Especially the cutoff: a
surprising number of files have a brickwall cutoff; 16 kHz is common,
13, 11, 10½ ... that's files I've received in digital form. Ones I've
made myself, e. g. by digitising from (analogue) records or tapes, I've
also found - especially for older material - there's often no _content_
above a surprisingly low frequency, other than tape hiss or (for 78s)
surface noise. (If in doubt, _high_ pass it and listen!) But that's
something of a side path from format conversion discussions.
--
J. P. Gilliver. UMRA: 1960/1985 MB++G()AL-IS-Ch++(p)Ar@T+H+Sh0!:`)DNAf

"OLTION'S COMPLETE, UNABRIDGED HISTORY OF THE UNIVERSE
Bang! ...crumple." - Jery Oltion