Re: Sony's new prosumer: Canon going to cringe?

"Doug McDonald" <mcdonald@xxxxxxxxxxxxxxxxxxx> wrote:
RichA wrote:
Curious, but how is Canon going to squeeze out more DR if the pixel
count climbs?

Better amplifiers. That's the only way.

I'd phrase this as "lower noise circuits". But you also need better A/D
converters.

You have to increase the
maximum electron count in the well, so you have to get the amplifier
noise down (and the digitized bit up of course) to get the
better possible range. This is not easy especially with smaller pixels.

The well depth is proportional to the area of the pixel, so it isn't going
up. The big-pixel Canon cameras have some amount of room for improvement in
the circuit noise, amplifiers, and A/D converter areas (the 1DIII claims to
have 14-bit A/D converters, but I haven't seen any tests yet that determine
if that is doing any good) as the reference in my other note shows.

By the way, this is one reason that 3 colors per pixel systems are a bust:
you lose a factor of square root of three in sensitivity/noise/dynamic range
due to each pixel requiring three wells (charge storage devices). Oops.

David J. Littleboy
Tokyo, Japan


.

Relevant Pages

  • Re: 21MP 1Ds Mark III Announced
    ... this electron noise, and you get noise about 20 electrons. ... is not limitations of the sensor, but limitations of the A/D ... Fast A/D 16-bit converters deliver 78-82 dB SNR, ...
    (rec.photo.digital)
  • Re: Megapixels and noise
    ... is photon noise, and we need higher QE and deeper full wells ... 16-bit A/D simply is not going to provide any improvement. ... your figures for the sensor are correct, ... the sensor could have a dynamic range ...
    (rec.photo.digital.slr-systems)
  • Re: High Resolution AD Converter Noise
    ... > contains 8 parallel switched capacitor SAR (no analog mux) from Analog ... > Now customer complains about unknown noisy results using ... > I can hardly believe the AD976 is poor becouse the noise is much above the ... due to the DC-DC converters, oversampling may not work as well. ...
    (comp.arch.embedded)
  • Re: amplify 40kHz audio signal using TL082: first two stages are fine, but high noise from the third
    ... >> I am making an ultrasound signal receiver. ... All amplifiers are in inverting input mode. ... >> is significantly corrupted by noise. ... on the transducer characteristics. ...
    (sci.electronics.design)
  • Re: For all the group
    ... Jim Thompson wrote: ... It's getting tedious checking all filtered-out posts for good guy/bad ... I'm sure it's noise related, and thus depends on the configuration... ... Monolithic 16-bit converters that sample at 100ksps have about 1-2 counts of RMS noise -- this means that your output number will routinely vary by four to eight counts from nominal. ...
    (sci.electronics.design)
Loading