Arbitrary Thoughts on Exposure for Digital

Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
> In practical terms, does this mean that for high dynamic range scenes, you may wish to trade off shutter speed and depth of field in favor of lowest ISO setting?

Yes, in favour of letting more light into the camera.

> And technically speaking, is dynamic range controlled more by the depth of the quantum well than the noise of the first preamplifier stage?

Maybe not by depth, but by light collecting ability.
 
Joined
Aug 17, 2006
Messages
173
Location
Maryland
Thanks Iliah. I like to shoot birds in flight and have been using higher ISOs for higher shutter speeds to counteract my and the birds motions. I've noticed that in many cases, the birds' white heads have been blown. I'll now try lowest ISO and wider lens opens openings to get the best of both worlds. Hope there is enough DOF at 200mm f/2.8.
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
The higher is ISO setting, the less is usable well depth with Nikons. To use full well depth at ISO higher then base raw file needs to be non-linear, with compressed highlights. That poses huge calculation burden on Nikon Capture to decompress data using high precision math; and makes most of third-party converters completely unusable.
 
Joined
Aug 17, 2006
Messages
173
Location
Maryland
Are there more than 16 bits of dynamic range in Nikon sensors, where 1 stop = 1 bit = 6dB//1 volt? I assume converting 12 bit compressed NEF to linear integer or floating point is easy, LUT.
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
Dynamic range in stops and dynamic range in bits are not the same. More important here, compressing highlights instead of truncating them results in leaving less space for midtones, and increase in posterization. For high ISO raw file needs to be 16 bits at least.
 
J

jeremyInMT

Guest
so is it better then, to use ISO200 and EV -2.0 instead of ISO800, given the same light and aperture so that we can use, say RML, to do the push more accurately?
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
so is it better then, to use ISO200 and EV -2.0 instead of ISO800, given the same light and aperture so that we can use, say RML, to do the push more accurately?

For my D2X the answer is ISO 320 and exposure compensation with push in RML. I suggest testing - it may be that overall better results with your camera are ISO 400 and -1eV, or some setting between 200 and 400. It may well be wrong altogether.
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
The difference is that the scale can be different, for example, 1 bit can correspond to 1/2eV.
 
J

jeremyInMT

Guest
For my D2X the answer is ISO 320 and exposure compensation with push in RML. I suggest testing - it may be that overall better results with your camera are ISO 400 and -1eV, or some setting between 200 and 400. It may well be wrong altogether.

I have a D80...how do I go about testing? In other words, what am I looking for as the best setting? sorry I'm a little dense about this, but I do want to take advantage of the smart people here. :)
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
I have a D80...how do I go about testing? In other words, what am I looking for as the best setting? sorry I'm a little dense about this, but I do want to take advantage of the smart people here. :)

You can use Imatest (by Norman Korens' team) to evaluate noise and resolution. Set the camera for correct ISO 800 exposure, take shots using different ISO settings, push in postprocessing, feed the resulting files to Imatest.

http://www.imatest.com/
 
Joined
Feb 13, 2005
Messages
1,900
Location
Blackpool Lancashire UK
No it does not :)

Try it yourself, set different white balance for two different shots, and look at raw data.

from the people who designed the camera.

Please explain analog signal processing in this case more specifically.

H,H: The acuteness of the tone curve is directly related to the minuteness of signal partitioning. The more acute the gradation, the more precise the luminosity measurement. With conventional digital cameras, reproduction of gradation depends solely on the results of digital measurement, eventually causing gaps in gradation.

To minimize such gaps in gradation, the D2X introduces our unconventional approach of analog adjustment for white balance in the first stage of signal processing (believe this is AD conversion as it was explain to me). This approach helps to achieve smoother, more accurate gradation than conventional digital cameras are capable of,

for smoother display of tones from highlight portions to shadow portions of the image. With optimized distribution of analog and digital white balance gain, the new engine produces smoother gradations with consistent and smooth transitions, all with exceptionally pure color rendition."

It was explain to me part of the white balance information is used to boost the gain to a individual channels this part of the white balance information is none retrievable and uneditable the remainder of the white balance information you can edit as normal in the raw file.

Phillip
 
Last edited by a moderator:
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
Phillip,

> Please explain analog signal processing in this case more specifically

I will, but first please make the direct experiment I suggested. It is much better then discussing with a Russian native speaker the Japanese marketing text translated into English, where it is not even mentioned if it refers to raw or jpgs :)

BTW, as I mentioned I'm not an English speaker. Please educate me: is it a polite way of saying things: "I will let you go and find it"?
 
Joined
Feb 13, 2005
Messages
1,900
Location
Blackpool Lancashire UK
Phillip,

> Please explain analog signal processing in this case more specifically

I will, but first please make the direct experiment I suggested. It is much better then discussing with a Russian native speaker the Japanese marketing text translated into English, where it is not even mentioned if it refers to raw or jpgs :)

BTW, as I mentioned I'm not an English speaker. Please educate me: is it a polite way of saying things: "I will let you go and find it"?

Hi Iliah and must apologise for that as I was editing it and it should have said (I will go and find it) I use a voice recognition software package for writing my correspondents and it does make mistakes, there is some information here

http://www.nikonimaging.com/global/technology/scene/10/index_02.htm




new_pa1.jpg
Subscribe to see EXIF info for this image (if available)





"The diagram above clearly demonstrates the D2X's four channel readout and independent color processing and A/D conversion which is carried out. Although Nikon aren't revealing a huge amount of detail about what goes on at the 'pre-conditioning' stage (just before A/D) the idea is that data from the individual channels can be amplified specifically to apply some amount of WB correction and to maximize color gradations before conversion to a digital value in the A/D converter."

Also some of the information come's from the professional seminars that nikon invites me to.

Phillip.

:smile:
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
> The diagram above clearly demonstrates the D2X's four channel readout and independent color processing and A/D conversion which is carried out.

Yes.

> Although Nikon aren't revealing a huge amount of detail about what goes on at the 'pre-conditioning' stage (just before A/D) the idea is that data from the individual channels can be amplified specifically to apply some amount of WB correction and to maximize color gradations before conversion to a digital value in the A/D converter.

For raw files pre-conditioning coefficients are constant. I know it from experiment. And I know the reason why they are constant - highlight clipping, like it happened in D1 original, where white balance was applied to raw data, not used as a tag.
 
Joined
Aug 17, 2006
Messages
173
Location
Maryland
Does that mean in raw + jpg mode, two samples are taken, one with constant analog gain coeficients for raw and a second with white balance controlled coefficients for JPGs?
 
Joined
Jan 29, 2005
Messages
2,761
Location
nowhere
As far as I can see JPGs are going through ASIC while NEFs are not. I do not use JPGs at all in my workflow so I was never motivated to dig deep into it, sorry.
 
Joined
Aug 13, 2006
Messages
3,235
Location
Berlin, Old Europe
But don't they "both" go through ADC and the same preconditioning beforehand before they split up into JPG and NEF after ADC?

That's what I would imagine.

If so, JPGs would come out differently depeding on QUAL setting (RAW or RAW+JPG) because of constant preconditioning coefficients when shooting RAW.

This however, I don't imagine. :Dizzy:
 

Latest posts

Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Nikon Cafe is a fan site and not associated with Nikon Corporation.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Forum GIFs powered by GIPHY: https://giphy.com/
Copyright © Amin Forums, LLC
Top Bottom