> From: Tim Hughes <timhughes@xxxxxxxx>
>
> There is what is called quantization noise... In a good
> design the sensor noise is a lot bigger than quantization noise, so
> A/D resolution then adds very little quantization noise , exactly
> as you suggest.
>
> Generally random noise adds as RMS quantities, so quantization
> noise is a small effect... ideally you want the A/D to have... a
> convertor noise level itself of less than 0.5LSB...
To grossly paraphrase this excellent summary, dynamic range *is*
related to bit depth in a complicated way.
In any optimal design, the bit depth should give a good indication of
dynamic range, using the "1/2 LSB" rule-of-thumb Tim alluded to.
Claude Shannon 101.
However, good A/D converters are cheap compared with good sensors,
and manufacturers will sometimes inflate their bit depth past the
point that makes sense, digitizing 5-6 bits of noise, just to have
one up in the specmanship war.
:::: Hydrogen power will dramatically reduce greenhouse gas
admissions. -- George W. Bush ::::
:::: Jan Steinman <http://www.VeggieVanGogh.com> ::::
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|