good to hear some clear thinking on the subject, helix t. i agree with
you mostly, and although i believe that if in a hypothetical situation
you would think that engineering was as good for one part of a given
program as for another (that cbr and vbr comment you made), it's a bit
of a gamble without more specific info than we have. it's a smart one,
so i like what you say allright. i can probably neither prove nor
disprove it, but i will say the following.
i like the r3mix site, which i read over in its entirety a while back,
but you may notice that many of the graphs show remarkable differences
between different versions of LAME. of the two versions of LAME included
in their semi-informal study, the earlier version of LAME that was shown
to work better with VBR than the later version (which shows higher
quality results on the graphs when using CBR) very likely does not have
as good psychoacoustics as the later; and i believe we are on yet later
versions of LAME even as we type now.
psychoacoustics is indicated to be one of the more important directions
the people working on this are headed towards, if not the most
important. so this would be an indication for anyone concerned with
encoding as much as we are to always try the updated versions and keep
tabs on the discussion of these new versions. the latest versions have a
tendency towards better ps.acoustics with the use of CBR, as is shown by
this idea, as well as by the graphs at www.r3mix.net, i would say (at
256kbps, which is also what i advocate for most electronic music, if i
want a high quality mp3. again, i will use 192kbps if i only need to
archive content that is not as important for me.)
thanks for reading through and for your replies (anyone)
On Thursday, January 31, 2002, at 04:16 , Helix Tradesman wrote:
quoted 36 lines if you make a VBR file you are punching the midrange of the sound
>> if you make a VBR file you are punching the midrange of the sound
>> right in the gut.
>
> how do you figure?
>
> i still havn't heard a good argument against VBR. it's theoretically
> best because it will dynamically assign a frame size basic on content.
> in practice VBR is only as good as the encoder you use, but by the same
> token so is a constant bitrate mp3. your mp3 is only as good as your
> encoder.
>
> the real question with fixed bitrate vs VBR should be this: is the
> algorithm which determines frame size going to pick a large enough
> value for the frame? if you don't trust LAME to pick this value
> correctly, how can you trust it to correctly decide which parts of the
> music are expendable? personally i trust that the frame size picking
> algorithm is at least as good as the algorithm which decides what sound
> to eliminate.
>
> as for joint stereo i think it's preferable to true stereo, but i
> wouldn't claim to be an expert on the specifics. again, your mp3 is
> only as good as your encoder. most of the info on www.r3mix.net is
> quite good and worth reading over (although their myths section is
> bullshit).
>
> cheers
>
> _________________________________________________________________
> Join the worldís largest e-mail service with MSN Hotmail.
> http://www.hotmail.com
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: idm-unsubscribe@hyperreal.org
> For additional commands, e-mail: idm-help@hyperreal.org
>
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at
http://mail.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: idm-unsubscribe@hyperreal.org
For additional commands, e-mail: idm-help@hyperreal.org