On Wed, Nov 24, 2004 at 06:47:57PM +0100, Timo Sch?ler wrote:
> >>Now I have more significant question: I've noticed when I burn a CD,
> >>the audio quality changes when I move from the sound file on the 
> >>computer
> >>(.wav, say) to playing back the audio CD.
> >
> >Depending on how you've got the CD-ROM drive connected, you might be 
> >using the DAC in the CD-ROM drive in one case versus the DAC in your 
> >sound sound when playing a .wav file.

Yes.  Except that some of the CD players in question are not CD-ROM
drives.  When testing a CD at home, I plug my headphones or speakers
directly into the CD-ROM drive's face plate.


> >>In particular, music seems to get ramped up.
> >
> >If "ramped up" means a change in pitch, that may indicate a problem 

"Ramped up" means louder volume.

It is like it went through an equalizer that selectively adjusted
the volume of the music (or perhaps just certain frequency bands
which, as a rule, tends to separate voice from non-voice).  For
all I know, this is a design feature of CD players.  I am curious
why it happens, but I am more interested in finding a way to either
accurately preview the effect or else to precompensate for it before
burning the CD.

I do not think that playing the single .wav file (as a whole) at a
different speed would by itself selectively raise the volume of
the music.


To be sure that the picture is clear: We have a stack of .wav files
that get edited and mixed.  A final .wav is produced, which is then
supplied to cdrecord for burning a CD.  If I play the final .wav
via audioplay (or audacity), it is quite distinguishable from
playing the CD.  In spots where a music track was laid behind a
segment of speech, the music is louder ("ramped up") on the CD player,
relative to the speech.




> >with the sampling rate your card was using on the .wav file.  It's not 
> >uncommon to see .wav files with 48KHz sampling rate rather than that 
> >44.1KHz sampling used by CD's...

All audio files are at 44.1KHz in order to use with cdrecord.


> cdrecord should complain on 48kHz .wav files... (it did for me, as i 
> transitioned audio from DAT [48kHz] to cd-r...)

Yes, it complains for me as well if I use anything but stereo
44.1KHz little-endian 16 bits per sample, I believe.  And unless
the file is just the right size-multiple, I need to supply "-pad"
to have cdrecord pad the audio file to meet audio CD blocking
requirements.

(I get the impression that music CDs can handle mono, but cdrecord
has resisted my attempts to supply a mono .wav.  I once tried to
burn a 90-minute mono track to CD...  This isn't really a major
issue for me at the moment, as I'm currently using 28.5 minute
tracks.)


-- 
  "I probably don't know what I'm talking about."  http://www.olib.org/~rkr/