Is this a gimmick? 22-bit converter from Pentax

Started Sep 14, 2006 | Discussions thread
MitchAlsup
Veteran MemberPosts: 4,617
Like?
Re: Is this a gimmick? 22-bit converter from Pentax
In reply to sjprg, Sep 15, 2006

sjprg wrote:

Here again the more accuractely that you can digitize the analog
signal the more easily you can distingish between noise and signal
with DSP techniques.

But if the analog signal is quantized (such as, say, into individual electrons) being able to count them more precisely than 1/2 an electron, you do not get any more useful information. That is, all the bits between bit 17 and bit 22 ARE noise!

Most modern A/D chips can convert well above
the 2 Ghz range.

Perhaps I should remind you that much of the noise performance of the 20D over the 10D was obtained by SLOWING the A/D converters down! A 20D runs channels of A/D at 1 MHz while the 10D ran 2 channels at 24 MHz. Running the A/D at fast frequencies adds NOISE to the conversion process!

This implies that the scanning rates could be be
increased.

This demonstrated a profound lack of understanding as to what noise is and how to be "low noise".

The camera manufactures have kept us in the dark as to
processing frequencies, scan rates, etc.

No, they are not keeping us in the dark, it is quite easy to find this information. That you are in the dark is a function of not looking for that information before opening your mouth.

My personel beleif is that
they are milking the technology with incremental improvements well
below the technical limits.

Best joke I've herad all day.

I don't know if the manufacturers use sample and hold between the
sensor and the A/D converters, they don't say.

Google up Double Correlated Sampling.
{snipped for brevity}
--
Mitch

Reply   Reply with quote   Complain
Post (hide subjects)Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark post MMy threads
Color scheme? Blue / Yellow