New try digital question

I hope this answer is less obvious as my last question:

As some of you might know, I'm an astronomer. Well astronomy has ment a great deal in the devellopment of CCDs. Problems we see in satellites are that the CCDs degrade. Now are these CCDs exposed to quite some different elements than your camera CCD, but still I'm curious...

How about the degradation of the CCDs and CMOSs in DSLRs. Do you see more noise in your shots now than it was 1, 2 or even 3 years ago. (With the same body!!) Do you feel the CCD is less sensitive?

Cheers

Peter

I've had my 1D for almost two years with no degradation. I would think that the space environment is much harder on them due to various forms of radiation, potential temp cycling, very long exposures, etc. Its a very interesting question though!

..

Good Question Peter!

I've been running a D30 for 2yrs and a D60 for a year and haven't seen any degradation. I think one big difference between the CCD/CMOS sensors in use in the cameras and in Astronomy may well be in the noise cancelling algorthyms employed. AFAIK the noise cancelling employed in cameras happens just before and just after the exposure with the circuitry measuring the noise levels of the "empty" sensor "pixels" and then deducting this from the image when the shot is taken. I'd guess that this would also have the effect of reducing image-visible signs of sensor degradation.

I'd guess that the signals being used in Astonomy are so much weaker that to use a strong noise cancellation circuit or algorthym would run the risk of masking the actual signal itself. And that the noise cancellation will be done totally separate from the image capture. And again, that might be why sensor degradation is actually visible.

The point made about radiation is a good one too as a space environment is a very hostile place for electronics circuitry

I guess we'll know in a few more years

Perhaps a bigger problem with space CCDs is the spectral resolution. This means that the CCDs are not able to discriminate 'every' colour.

Have you noticed any colour shifts, or less colourfull frames over these years. If not, is there a way of testing this with the colour card tests. Must is that you've did this test when you purchased the body....

Peter

Peter,

I'd guess most of us wouldn't have done any tests like that when we first bought the camera as 99% of us wouldn't have know much about the DSLR's at the time. 2yrs is a long time in the learning curve

Most of us nature photographers would just re balance/re level the photo...ie edit out any colour cast/loss, so we wouldn't notice a slight change. We also don't work with constant light levels or colour temperature ie the effects of the changes of daylight and atmospheric conditions would overwhelm any effect of age on the sensor.


In the new world of DLSR's and Digicams I doubt that it would ever be a problem as there will always be something bigger and better out there in a couple of years time...by the time a camera is getting to the age when any problem with the sensor becomes apparent, it will be technologically obsolete. Unlike the film SLR's etc that are still being used 15-20yrs or more after they were made. (Before anyone says they still work...I've a Kodak Autographic Junior that I can't get film for anyone know where there is some in Europe )

I'd guess you could probably find an answer (you'd certainly start a discussion ) on somewhere like DPreview

I was thinking about posting this on Dpreview.

But this colour resolution degradation is not a shift. Your sensor is just not capable any more to discriminate between slightly different colours and reading them as one. F.e., imagine you have 10000 greens and you can resolve all of those with your new camera. If your colour resolution degrades 2% over a certain period, you'd only see 9800 colours of green. You'd loose fine detail!

The reason why I'm asking this to myself and now you is because of all the new DSLR rumours of Canon. Now that the 1D seems to be discontinued news must be on its way. If the new one would cost $3000 it must work for me for about 5 years. That's long, but vital prizewise.

Cheers

Peter

Peter,

unlike in scientific circles where you're trying to be exactly acurrate on the colour/wavelength, the colour of the subject and the light it is in are not absolute, nor need they be. Otherwise you'd be carrying around a light meter that gave you exact colour temperature readings in Kelvin...and they're expensive

Further, if you are looking to have an image published, printed, or displayed on the web, then the colours & brightness etc are all likely to be changed by the guy doing the pre-press work...for publication in magazines etc they'll also be dropped into CMYK colour - a colour space that has even less gamut than sRGB

Basically what I'm saying is that a shift or loss of colour discrimination on the sensor is not going to make much difference compared to the messing around and loss of colour information that is going to be carried out on the image later!

I think you'd be more likely to notice an increase in noise as the sensor and support electronics age.