Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

DSLR ISO


Recommended Posts

Can someone tell me were along line of 'the photon enters camera' to 'Raw image output' iso is applied?

am not sure on were and how it works, my thinking was that it just amplifies up the voltages before they are converted to counts by the ADC?

but am not sure and i would like to know if there is a way through the use of ISO to improve the dynamic range of DSLR's?

Ally

Link to comment
Share on other sites

Not entirely sure what you mean but I understand that lower iso is better for recording a wider dynamic range. So to record star colours, for example, it's better to use iso 100-400 rather than 800-1600.

Have you had a play with high dynamic imaging (HDR) yet? That is, take three or more images with exposures 2 stops apart and combine in packages like Photomatix or photoshop.

Link to comment
Share on other sites

i have, and it works wonders on certain pictures but having to have the tripod with me is a bit of a pain.

I was thinking about imaging galaxies and trying to get the core and the faint spiral arms in one exposure length. not easy because you need really long subs to get the arms well and often the core burns out. I have been doing some work on the SNR of stacked images and i looks to me that if you use the method of take long, short and medium exposures to get a better dynamic range then they all must be of the same SNR quality other wise the fainter stuff is lost ;)

so Better dynamic range is the question...

Link to comment
Share on other sites

it could be the sense node sensitivity, which is set by the capacitance of the pixel. This is the uV/e-. This capacitance turns charge to voltage, with a conversion defined by the capacitance. For the H16 it is 31uV/e

but I feel that the iso is the voltage applied to the ADC, governing how many DN your camera spits out based on the input e-

ISO is a myth that higher ISO gives better snr. It gives more DN, not e-.

I would strongly think that read noise would be lower if you used a lower iso. Which would obviously increase the DR, full well should be the same for all ISO.

there is a simpler answer,

just buy a modern back thinned large format 15x15um pixel CCD. Youll have all the full well, sensitivity and DR you could want!

Paul

ps in theory stacking should show a root N improvement in DR since the read noise for N averaged images is sqrt(N)

Link to comment
Share on other sites

I have been doing some work on the SNR of stacked images and i looks to me that if you use the method of take long, short and medium exposures to get a better dynamic range then they all must be of the same SNR quality other wise the fainter stuff is lost ;)

so Better dynamic range is the question...

To record the full range of tones present in galaxy, from bright core to faint spiral arms, I would personally take long, medium and short exposures and blend them together in Photoshop using the layers and mask tools. I think you would have much better control of the final result.

Link to comment
Share on other sites

Hey guys was researching stuff about shot noise for my Astronomy project and stumbled over this little gem.

Pretty much the same stuff i have been doing for the SXV H16 but it has an interesting bit in it about ISO and SNR. on page two.

I don't think it is too heavy on the Maths and looks very well written but the really nice thing about it is that it means i don't have to do this stuff for my own camera not knowing what i might get

Noise, Dynamic Range and Bit Depth in Digital SLRs

enjoy

Link to comment
Share on other sites

nice read, didnt read it fully, but I saw some stuff.

he keeps referring to PRNU noise dependant on signal.

wrong

FPN is dependant on signal. PRNU is a constant related to the FPN.

secondly, and this shocked me....

he states CDS cant be done in CCD sensors.

this is the biggest load of Rubbish i have heard from someone claiming to be a scientist in quite a while.

CDS cannot be done in cmos sensors. I believe they use DCDS, depending on the number of transistors in each pixel.

nowhere did i see a single PTC. Why not post some standard PTC's vs ISO, isnt that far simpler and a hell of a lot more intuitive. Is he plotting wierd stuff so we dont cotton on to his dubious tests?

anyway, apart from that, there looks like some interesting stuff, and well worth a look.

Thats my take, but im old fashioned

paul

Link to comment
Share on other sites

CDS is used to remove reset noise from CCD sensors. Full stop.

He is right when he says you have to destroy the 'state' when reading out. One must read the whole sensor, which is different to CMOS, where individual pixels can be clocked.

but i am very confident that i recently read something (a paper) stating that DCDS could be used for CMOS but only dependant on pixel arcitecture

Link to comment
Share on other sites

Might i point out this is a theoretical particle physicist not a CCD or CMOS person and that this work is just a passing thought about camera noise not a referenced journal like say janesick or others in the solid state detector world would write.

His plots are close to PTC's only that he has not logged them and from the fact that he doesn't seem to know the gain of the camera can be got from the PTC, i can only assume that he doesn't know they exist.

Other than that, i think it is a nice article because it is written by someone interested in Photography not CCD's so is more relevant to people on SGL looking for better images without the maths...

ally

Link to comment
Share on other sites

i have since read the rest of the article, and most of it very good and pertinent.

He discovered that read noise falls with lower ISO, and result I thought may have occurred. As a result, SNR is increased with lower ISO.

So I hope his work has at least debunked the myth that higher ISO yields better images.

I just was nit picking a bit, but CDS is done in CCD sensors as illustrated in James janesicks Scientific Charge Coupled Devices page 557. Here he shows a diagram of a CDS circuit used to remove reset noise.

This is my only real quam

paul

Link to comment
Share on other sites

Dennis, i had all ready emailed Prof Martinec, about using some of his stuff in a report and i might of asked him a few questionis about the stuff if i thought he would be interested in it, but as far as i can tell this is just a wee test he ran 'just for fun' not a published article.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.