Klas Strandberg <> wrote:
>Thanks Raimund. What I would like to understand a bit is how good
the simple
>AD, DA converters in modern walkman type DAT's and MD's are. When
using a
>Portadisc compared to a cheap Sharp MD, what role does the AD, DA
converters
>play?
I would say, that there are only minor differences between the AD
and DA converters chips between the various devices. Today, these
chips are very cheap (5 ... 20 $) and are generally of a high
quality. However, the overall quality of a recorder depends also on
other parts that will make the difference. Among these things are
the pre-amplifier and the specific design of the recorder (e.g.
shielding against internal digital noise). I still have not compared
the Portadisk to the cheaper MD's. Besides my mini-laptop equipped
with an external 24 bit audio interface, I'm still using a consumer
DAT recorder SONY TCD-D3 recorder (purchased at about $600 in 1991).
This old device still works satisfying for my purposes and I can not
find any artifacts that could be attributed to its AD converter.
>And the converters on cheap sound cards?
The converter chips on cheap sound cards might be the same models
than those in consumer MD recorders. Therefore, here applies the
same I mentioned above. More important than the A/D converter itself
is the shielding against digital noise and the noise performance of
the pre-amplifier (even that of LINE-IN if you want to have the
maximum possible dynamic range). Especially the microphone inputs
of cheap internal soundcards are often of a poor quality (certainly
in laptop computers).
>Will a typical cheap converter on a standard sound card make
visible changes
>on a sonogram?=A0
No. However, it depends on the dynamic range that you wish to
display on the spectrogram. Usually (e.g. for species identification
or other sound parameter measurements), it is not required to
display any signals below -60 dBFS. This level is far above the
higher inherent noise levels of cheaper soundcards. Additionally,
the potentially lower dynamic range of cheap soundcards will be
masked by the environmental noise and the inherent noise of the
microphone that is already present in a recording. If you want to
display the full dynamic range of the recording (including the
softest background noise) on a spectrogram, the higher inherent
noise of cheap soundcards will become visible (but only if that
noise is stronger than that of the original recording).
>Would you say that cheap AD, DA converters cause more trouble than
ATRAC and
>MP3??
No! Cheaper AD converters may add a very small amount of additional
noise only (leading to a dynamic range of e.g. 85 dB instead of 92
dB). There are no other significant distortions. The artifacts that
occur in ATRAC or MP3 may lead to extremely degraded dynamic ranges
of less than 20 dB at worst cases (which will of course never occur
so extremely in common nature recordings).
To my mind, the problem in ATRAC is, that the potential artifacts
are highly unpredictable. Sometimes, the sounds will be recorded
properly, but in some special cases you will get unreliable results.
This is the reason why I always prefer DAT or other recording
techniques without compression. I'm considering to buy the new
PMD670, even if it has only a resolution of 16 bits. As we have
discussed earlier, the limited dynamic range of a 16 bit system
would be sufficient for most nature recordings...=20=20
Regards,
Raimund
________________________________________________________________________
________________________________________________________________________
|