Tim Nielsen wrote:
>
> Hate to break it to you, but recording a quiet -40 dB level signal in
> nature is no different than recording a -40 dB in a music studio. The
> program material itself has no relevance to the discussion at hand.
> We're talking level of sound and bit-depth and resulting noise and
> imaging and they're completely independent of program material.
No, this is completely wrong!!!
It actually depends on the absolute sound levels that are present at
the microphone location! In a mucsic studio, you are recording
relatively loud sound sources at very close distances. For instance,
there might occur absolute sound levels of more than 80 dB. If you
record such a lound sound at an extremely low gain that produces only
a level of -40dBFS in the resulting sound file, it is no surprise that
the much softer room acoustics is lost. With this particular gain
setting, the usable range of a 16 bit recording system would
theoretically range from 120 dB down to about 30 dB. However, the
inherent noise floor of a good microphone can be as low as 6dB(A).
Note that it is this faint noise floor than can be heard in the silent
periods of a recording when it is done in a quiet studio.
However, when you record softly vocalizing animals at much longer
distances, a -40 dBFS level can usually only be achieved by
dramatically increasing the gain setting (compared to the above studio
setting). As a result, the faint microphone noise floor will also be
amplified proportionally. Consequently, you will get a 16 bit
recording in which the least significant bits are already occupied by
the microphone noise floor.
So, what can be learnt from that? It is VERY important to understand
the limitations of ALL components in the recording chain. Once you are
familiar with those limitations, you will be able to adjust the gain
in such a way that the microphone noise just masks the quantization
noise of the 16 bit file format. Then there will be still plenty
headroom for most louder sound events that usually occur in nature.
If one does not understand these things, even the 24 bit recording
format cannot compensate the mistakes made by the subject behind the
recorder...
Lou, I'm sorry for continuing to post this kind of "misinformation" ;-)
Regards,
Raimund
|