>
> + maybe a MPEG2 vbr should get a Xing header?
> at least optional via a switch?
>
Can you get Xing to generate a MPEG2 VBR file? I tried once and
it aways produced fixed bitrate output files. The problem is the
Xing header is embedded in a 64kbs frame, but at MPEG2 samplerates, a
64kbs frame is not large enough to hold the full Xing header.
Probably should just go to a 128kbs frame for the Xing header,
but I'd like to find out what Xing does first.
> + finally, whats the dynamic range of a MPEG2 layer3
> stream at 24, 22 and 16 kHz? Is it half the
> dynamic range of a 44 kHz MPEG1 layer3 stream?
> the MDCT scaling of the ath should consider this.
>
good point, and it should be tested. But I think it shouldn't change
much: The MDCT transform takes as input 576 samples (both mpeg1 and
mpeg2, any sampling rate) and computes the coefficients without
knowing anything about the samplerate. So the min and max energies
that can be produced should be about the same, and the ATH is scaled
so that it's minimum matches the min energy from the MDCT (minus our
adjustment factor).
Another way to think about this is that the in all cases, the input
pcm data ranges from -32767..32767, so the dynamic range is always
the same?
Mark
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
