Canon CMOS Sensors

[avandonk]

Here is a good read if you want to know more about Canon CMOS sensors as it explains things in plain english.

http://www.the-digital-picture.com/R...hite-Paper.pdf

Bert

[Dennis]

Thanks Bert, had a quick scan and it looks very readable - now filed away for a rainy day's reading in more depth.

Cheers

Dennis

[h0ughy (David)]

Originally Posted by Dennis

Thanks Bert, had a quick scan and it looks very readable - now filed away for a rainy day's reading in more depth.

Cheers

Dennis

have done the same thing - though Dennis may get to read it well before me due to the weather up there in the promise land

[tornado33]

Thanks, Bert I will read through that tomorrow.
With their CMOS chips performing so well with just a single stage peltier cooler installed in a camera that wasnt designed to be cooled, it is remarkable how well it performs.

[avandonk]

Yes Scott there are several reasons why Canons CMOS sensors perform so well compared to "cheap" CMOS chips. First is the extra on chip circuitry then the ULTRA pure silicon they use to suppress random noise. The article explains all the other reasons. They are even cheeky enough to say why CCD's have shortcomings compared to their CMOS chips.

I would really be interested in seeing how a Canon 5D chip would perform with a two stage Peltier cooler to -40C and no Bayer filters and a sixteen bit A to D converter or better. The QE would be as good as any CCD and the signal to noise far better.

Makes you think does it not?

Here is a thread where I collected data at an ISO of 1600 with the 5DH camera cooled to a few degrees above zero.

http://www.iceinspace.com.au/forum/s...ad.php?t=27502

Bert

[Peter Ward]

Bert, I've just skimmed through the piece to which you refer. An interesting read!

But I lifted this bit from the Canon blurb....

"The brighter the light that hits the photodiode, the greater the electrical charge that will accumulate within it"

I have a problem with this. For a more complete explanation see here
http://en.wikipedia.org/wiki/Photoelectric_effect

Hence I suspect my understanding of photodiodes is not up to par...or perhaps Canon are simplifying things a tad..

[avandonk]

It is a white paper Peter not a scientific treatise.

Bert

[Peter Ward]

Bert,

Understood, I was simply highlighting the energy of the photoelectrons increases with increasing frequency of incident light....not the amount of light implied by the Canon white paper. This aspect of frequency is pivotal to the understanding of the photoelectric effect, which all CMOS and CCD devices take advantage of.

True, once you cross the threshold frequency, more incoming photons means more liberated electrons, which is perhaps what they meant to say?

An interesting read just the same

[sejanus (Gavin)]

the cameras newer than the 5d look very good as well. I have done some long exposure tests with a 1d mk3 - though it's designed for sports the noise (or lack) in long exposures uncooled is remarkable.

I have 2 x 1Ds mk3's coming soon and will test them out as well though I'm not sure if they'll be as good given how many pixels they have stuffed in them.