How does the D3 achieve such high ISO?

Started Feb 17, 2008 | Discussions thread
ejmartin Veteran Member • Posts: 6,274
Re: 1D3 full well?

ohyva wrote:

ejmartin wrote:

Dark current noise is thermal noise. In the above referenced
sources, average dark current is a small fraction of an electron per
second, compared to read noises which are several electrons. So for
exposures less than a second or so, thermal noise is totally
negligible, as I mentioned.

Nope. Dark current is mostly caused by different leakeges in the chip.
We now talk about real wordl devices, not any abstrated ideal models.

Really? Googling "dark current noise", the first listing
http://www.roperscientific.de/tdark.html

states quite clearly that it is thermal noise:

"Dark current arises from thermal energy within the silicon lattice comprising the CCD. Electrons are created over time that are independent of the light falling on the detector. Said electrons are captured by the CCD's potential wells and counted as signal. Additionally, this increase in signal also carries a statistical fluctuation known as dark current noise. "

I have no idea what you are talking about here. Please be specific,
rather than alluding to vague "unkown quantities" look at the data
and state an effect or contribution to noise which is unexplained by
the standard noise model.

Unknown quantities means something we cannot "calculate" from
simplified ideal models. Parasitic components are due to
non-idealities in the silicon process and is something that vary from
sensor to sensor. To some extend the effect of these can be studies
with simulators like Spice and statistic methods like Monte-Carlo
type stimulus.

I invite you to look at the references I gave
http://astrosurf.com/buil/20d/20dvs10d.htm
http://clarkvision.com/imagedetail/evaluation-1d2/index.html

and state why you think that the "simplified ideal model" is a poor fit to the data. Sometimes simple models are just fine, containing the essential physics and stripping away irrelevant details.

As NR is typically decision controlled median type of local filtering
I very much doubt you can see that clearly in any fourier based
analysis. But I have to admit it's quite long time since my last math
lessons.

That is easily checked, I'll do an example and post it if I have the
time. But median filtering is a linear averaging over neighboring
pixels, and therefore will erase the noise spectrum at high spatial
frequency, no question.

Averaging or "intelligent peak cutting" or something else.
As everywhere there are the simple text-book methods and then the
more advanced intelligent methods.

Please, provide us the examples.

I did if you'd care to look:
http://forums.dpreview.com/forums/read.asp?forum=1032&message=26861064

Noise reduction of this sort, be it median filtering, wavelet based, whatever, involves removing power from high spatial frequencies; that's why there is a tradeoff between noise reduction and detail preservation.

I have to say you have an inexhaustible supply of red herrings.

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