On 10/25/2011 04:33 AM, Paul Davis wrote:
On Mon, Oct 24, 2011 at 10:00 PM, Al Thompson<althompson58@xxxxxxxxx> wrote:
MP3s take away a lot of "good" information.
Al, I have a huge amount of respect for your technical knowledge but
offhand remarks like this require some countering in an online forum.
Its true that low bit rate psycho-acoustic lossy compression does
remove some "good" information. But my impression is that at
"appropriate" bit rates, an overwhelming majority of humans cannot
differentiate between an mp3 and the original recorded audio. I'm not
entirely clear what "appropriate" means, but I believe that its hard
for most people to differentiate (in a double blind test) at 128kbps,
very hard at 192kbps and essentially impossible at rates above
256kbps. Should I update my understanding of this?
i'm relatively skeptical of any claims for audio steganography that
can survive arbitrary psycho-acoustic compression. in addition, random
permutations of the least significant bits of a PCM encoding will
almost certainly eliminate or at least reduce the confidence level
associated with the presence or absence of the watermark), in a way
that will be inaudible to more or less anyone.
While I don't know exactly what Al meant with "good information"
exactly, here's maybe a way to interpret it: The information that is
removed with lossy compression is in the best case exactly that
information that is redundant with respect to human perception. It would
be "good" in the sense that this redundant information would be a great
place to hide watermarking. With a "perfect" lossy compression scheme,
there's no bandwidth left to alter the signal without the human
perception noticing.. So any watermark would be audible.
Flo
Flo
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@xxxxxxxxxxxxxxxxxxxx
http://lists.linuxaudio.org/listinfo/linux-audio-user
_______________________________________________
Linux-audio-user mailing list
Linux-audio-user@xxxxxxxxxxxxxxxxxxxx
http://lists.linuxaudio.org/listinfo/linux-audio-user
