Hello Mark,
Sunday, May 14, 2000, 1:22:50 AM, you wrote:
Some more info: --noath : no noticable difference
--decode: no noticable difference
player: sonique 1.30.03 (no AE4)
MT> Does the chirp sound like an encoding glitch, or could it be explained
MT> by a lack of bits? At these high bitrates, there are two possible
MT> bugs you could be running into: There was a joint stereo bug
MT> introduced in 3.81 and fixed in 3.82 beta. If the version you are
MT> usings says "3.82 alpha", it means it comes direct from CVS and it
MT> might or might not have this bug.
I'm guessing no glitch, just lack of bits. Since 256 sounds ok to me
and 192 does not.
MT> The other bug might be this relaxed not-quite-ISO-spec
MT> framesizes that lame 3.82 can generate. So if you dont mind,
MT> can you try the following: (I would try this my self, but
MT> I'm guessing these chirps are too subtle for me to detect :-)
Now I got those ($40) Sennheiser HD490 and $22 SB128, I can with ease
detect artifacts. Decent headphones help alot. I plugged them in my
old opti soundcard, and the first thing I noticed was very much noise
in the R channel. When I used to listen to my stereo for things like
this (2 years!), I knew there was some noise, but I always thought it
came from bad cableing.
MT> CBR @ 128kbs. If this also has the problem, it is just the
MT> fact that these cymbals need more bits. It would also be
MT> interesting to report on how mp3enc at 128kbs compares.
MT> For preecho type stuff like this, mp3enc is usually a little
MT> better than lame.
I know it doens't represent much, but on my page I have some _very_
poor Q returns from mp3enc31.
mp3enc sounds a bit less harsh at 192, but in a different way also not
like the original wav.
LOL - 128 Lame JS -h sounds _better_ than the 191kbit average VBR file
...
Lame -b128 -mj -h -k confirmed: now the metal-like sharp tones are
back again...
Using polyphase lowpass filter, transition band: 15115 Hz - 15648 Hz
shows the higher spectrum isn't done very well at 192 :(.
MT> LAME382/ -V1 -mj -h -b128 --strictly-enforce-ISO
MT> If this fixes the problem, then we need to go back to enforcing
MT> the exact ISO buffer requirements.
No difference: this file and the original are byte-equal
MT> LAME382/ -V1 -m s -h -b128
MT> If this fixes the problem, it could be the jstereo bug.
No real difference: after a lot of hearing I think the first few
seconds sound better, but the last ones and the "Ki" still is too
sharp. I'm guessing the few extra bits help a bit.
MT> the last critical band is 16-22khz, and this band does not
MT> have a scalefactor, (and psycho acoustics are not computed
MT> for it either) so it is hard to control the noise in this band.
thanks, that explains _a lot_ of the weird freq analysings I get.
I hope I'm not too critical about this whole issue, but understand
that when you listen to the whole song in V1, it sounds really harsh
and unpleasant to my ears. The 7 seconds don't mean much, and they are
hard to distinguish, but the cymbals are constantly, for the whole
(album in that part) song sounding metal-like and too sharp (imho).
thanks for the response :)
Info2/r3mix.net/Roel
--
Best regards,
vdbj mailto:[EMAIL PROTECTED]
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
