Jump to content

sgl_imaging_challenge_6_banner_jupiter_2021.jpg.eacb9f0c2f90fdaafda890646b3fc199.jpg

 

 

ISO selection... unity gain vs exposure time?


pipnina
 Share

Recommended Posts

So over all the time I've been imaging, the general advice is to use 800iso (as a rule of thumb) unless the exact unity gain of your camera is known to be different. However, since if the exposure time is doubled, while the ISO is halved, you end up with the same image but less noise, but the sky background could well push even the dimmest pixels on your sensor to 30% or higher. To me this sounds like a win-win as lower ISOs on DSLRs often result in higher dynamic range, and less interference, at the same time no data is clipped from the sensor.

Am I missing something? Does ISO below unity gain negatively affect faint objects or are there potential gains to be made for people who can expose enough to compensate for the ISO reduction?

Interested to hear thoughts.

Thanks!

Link to comment
Share on other sites

I always thought that its sky glow/LP dependent and that when exposing for the fainter material, such as extended galactic arms. Lower ISO and longer exposure time will capture more of the fainter stuff since you're exposing the sub for a longer time and capturing the photos from it, of course over exposing the core which will be fixed later with shorter exposed subs, stacked and processed... but if you tried to do the same exposure time at a higher ISO, the sky glow would overexpose the sub and nothing would be gained... yes the lower ISO subs are and need to be exposed for longer to have the same (overall) brightness, but the fainer stuff is there that otherwise wouldn't be captured during a shorter but higher ISO sub and that is why long exposure is needed.

As far as using lower or higher ISO for dynamic range, I think that if you stack enough subs captured at a higher ISO, you will regain most of, if not all, of the dynamic range lost due to higher gain... with in reason of course, for example ISO 800 vs ISO 1600 or 3200 as opposed to say ISO 800 vs ISO12400.

That's my understanding.

Edited by MarsG76
Link to comment
Share on other sites

6 minutes ago, MarsG76 said:

I always thought that its sky glow/LP dependent and that when exposing for the fainter material, such as extended galactic arms. Lower ISO and longer exposure time will capture more of the fainter stuff since you're exposing the sub for a longer time and capturing the photos from it, of course over exposing the core which will be fixed later with shorter exposed subs, stacked and processed... but if you tried to do the same exposure time at a higher ISO, the sky glow would overexpose the sub and nothing would be gained... yes the lower ISO subs are and need to be exposed for longer to have the same (overall) brightness, but the fainer stuff is there that otherwise wouldn't be captured during a shorter but higher ISO sub and that is why long exposure is needed.

As far as using lower or higher ISO for dynamic range, I think that if you stack enough subs captured at a higher ISO, you will regain most of, if not all, of the dynamic range lost due to higher gain... with in reason of course, for example ISO 800 vs ISO 1600 or 3200 as opposed to say ISO 800 vs ISO12400.

That's my understanding.

I suppose maybe it depends on how the ISO works, if low ISO simply sets a brightness floor (i.e. 32'000 depth camera only registers first unit of brightness once it reaches 8'000 instead of at 0) or if it divides (so every 4 photons captured translates to 1 extra brightness increase in the final image...

As for dynamic range, what you say does sound true, but obviously having more dynamic range is always good as it improves the maximum exposure you can perform and keep highlights detail.

Maybe I'll have to do some tests... 200/400/800 ISO at 6m, 3m and 1m:30s exposure times, stacking 400 twice and 800 four times.

Link to comment
Share on other sites

I find this info useful..

http://dslr-astrophotography.com/iso-values-canon-cameras/ 

oops thats the Canon page, this is for Nikon.

http://dslr-astrophotography.com/iso-values-nikon-cameras/

so for the Nikon D3200 listed in the OP signature it would be ISO 200 which funnily enough is the same as my Canon.

 

Alan

Edited by Alien 13
  • Like 2
Link to comment
Share on other sites

12 minutes ago, pipnina said:

I suppose maybe it depends on how the ISO works, if low ISO simply sets a brightness floor (i.e. 32'000 depth camera only registers first unit of brightness once it reaches 8'000 instead of at 0) or if it divides (so every 4 photons captured translates to 1 extra brightness increase in the final image...

As for dynamic range, what you say does sound true, but obviously having more dynamic range is always good as it improves the maximum exposure you can perform and keep highlights detail.

Maybe I'll have to do some tests... 200/400/800 ISO at 6m, 3m and 1m:30s exposure times, stacking 400 twice and 800 four times.

Experimenting is the best solution.. I guess that it would be camera dependent, although you might find that the end results between ISOs are similar on most DSLRs...

  • Like 1
Link to comment
Share on other sites

The choice of ISO is camera dependent and depends if your model is ISO invariant (ISO less), a lot of Nikon cameras and one Canon camera are effectively ISOless so can shoot at low ISO and still drag dark info back the same as if they had shot at a higher ISO with no added noise.

Alan

  • Like 1
Link to comment
Share on other sites

6 minutes ago, SteveNickolls said:

Hi,

This is an interesting read-https://www.cloudynights.com/topic/500991-is-dslr-unity-gain-useful/

If I imaged at ISO 800 I couldn't image for long with the light pollution around here.

Cheers,
Steve

Looking at the list in the attachment... how can QE of a sensor be 234% or 278% as in the Nikon D70 or the Sony A900??? Isn't QE a measurememnt of how many photons as a percentage a pixel reacts to, ie captures?

Link to comment
Share on other sites

6 minutes ago, SteveNickolls said:

Hi,

This is an interesting read-https://www.cloudynights.com/topic/500991-is-dslr-unity-gain-useful/

If I imaged at ISO 800 I couldn't image for long with the light pollution around here.

Cheers,
Steve

That's a pretty good link. Some of the maths doesn't make sense to me but it seems like nothing some bias or crude dark frames couldn't fix (read noise).

Looking at my camera, 800iso has bright dark current with a 6min exposure, slightly darker at 400iso but massively darker at 200iso & 100iso. So i might see benefit there, but only at 100iso when I stretched the images to be similar brightness did I notice any read noise become noticable.

I might try imaging at 200iso instead of 800 for a while and see how it works out.

Link to comment
Share on other sites

Just now, MarsG76 said:

Looking at the list in the attachment... how can QE of a sensor be 234% or 278% as in the Nikon D70 or the Sony A900??? Isn't QE a measurememnt of how many photons as a percentage a pixel reacts to, ie captures?

Maybe it's that, on average, each photon produces 2.78 electrons of charge? Which probably means it's 2/3 electrons with a bias towards 3. I wouldn't know, in all honesty. Your guess is as good as mine... :dontknow:

  • Like 1
Link to comment
Share on other sites

Hi pipnina,

11 hours ago, pipnina said:

I might try imaging at 200iso instead of 800 for a while and see how it works out

That's a decent way to see what you get in practice and under your local conditions. I would like to do some experiments to look at the practical results but right now I'm not sure when the next clear night will be to try that.

Good luck with any trials you do.

Cheers,
Steve

  • Like 1
Link to comment
Share on other sites

On 07/02/2018 at 19:10, pipnina said:

if the exposure time is doubled, while the ISO is halved, you end up with the same image but less noise,

Well strictly the noise is higher in the longer exposure - it is the signal-to-noise which is better. But that is nothing to do with the change in ISO, it is the increase in exposure which is improving matters.

NIgelM

Link to comment
Share on other sites

22 minutes ago, dph1nm said:

Well strictly the noise is higher in the longer exposure - it is the signal-to-noise which is better. But that is nothing to do with the change in ISO, it is the increase in exposure which is improving matters.

NIgelM

Noise is higher? Is that like dark current or noise in some other sense?

Link to comment
Share on other sites

Isn't iso just a measure of the gain applied to the signal / photons captured by the pixels of the sensor chip?

Exposure time / the number of photos captured will determine the signal to noise ratio.

Changing iso will just stretch of compress this signal...... Won't it?

Could be wrong.....

Link to comment
Share on other sites

49 minutes ago, Kropster said:

Isn't iso just a measure of the gain applied to the signal / photons captured by the pixels of the sensor chip?

Exposure time / the number of photos captured will determine the signal to noise ratio.

Changing iso will just stretch of compress this signal...... Won't it?

Could be wrong.....

Yes the number of photons is the key to SnR... higher ISO does help a lot in speeding up the capture of brighter objects or object parts, but it limits the time exposure per sub since it will amplify the sky glow also, this is where ISO needs to be dropped, to allow longer sub exposures to capture the fainter parts of a object.

Link to comment
Share on other sites

6 hours ago, sheeprug said:

Interesting, but unfortunately no use to me since the Canon 750D is not included.  Anybody know how it might be possible to fine data for it?  All I know is it's 12 bits.

SR. 

These are the values for your camera and it's 14bit not 12bit

QE=49%
Unity ISO=1400ish
Full well=24k
Read noise=1.9e- @ 1600 ISO

Link to comment
Share on other sites

22 hours ago, wxsatuser said:

These are the values for your camera and it's 14bit not 12bit

QE=49%
Unity ISO=1400ish
Full well=24k
Read noise=1.9e- @ 1600 ISO

Thanks for that - interesting.  If you get a moment, can you let me know your source please?

Thamks,

SR  

Link to comment
Share on other sites

On 09/02/2018 at 14:09, pipnina said:

Noise is higher? Is that like dark current or noise in some other sense?

Photon shot noise will be higher in the longer exposure, as it is proportional to the sqrt of the number of photons collected. I guess the dark noise will be higher too. The read noise only depends on the ISO though (and it is often lower at higher ISO).

NigelM

Edited by dph1nm
Link to comment
Share on other sites

16 minutes ago, dph1nm said:

Photon shot noise will be higher in the longer exposure, as it is proportional to the sqrt of the number of photons collected. I guess the dark noise will be higher too. The read noise only depends on the ISO though (and it is often lower at higher ISO).

NigelM

So I get more noise from collecting more photons? I'm not sure I follow :/ I would have thought it to be the other way around.

Link to comment
Share on other sites

1 hour ago, pipnina said:

So I get more noise from collecting more photons? I'm not sure I follow :/ I would have thought it to be the other way around.

You get more noise, but it only goes up as the sqrt of the number of photons, whereas the signal itself is the number of photons. So the ratio of the signal to the amount of noise goes down (which is what you want). So if you get 4x the exposure you only get twice the noise (sqrt(4)), and the ratio of signal/noise is 2x higher.

NIgelM

  • Thanks 1
Link to comment
Share on other sites

3 minutes ago, dph1nm said:

You get more noise, but it only goes up as the sqrt of the number of photons, whereas the signal itself is the number of photons. So the ratio of the signal to the amount of noise goes down (which is what you want). So if you get 4x the exposure you only get twice the noise (sqrt(4)), and the ratio of signal/noise is 2x higher.

NIgelM

Does this apply only to single subs or also to stacked images? If I had 4x30s shots and 1x2m shots, would the signal/noise be the same (only accounting for photons, not read noise or dark current or other factors.)

Link to comment
Share on other sites

23 hours ago, pipnina said:

So I get more noise from collecting more photons? I'm not sure I follow :/ I would have thought it to be the other way around.

Collect 100 photons and the noise will be +/-10, collect 10000 and the noise will be +/-100, so strictly speaking you have more noise (and much more signal) - but people tend to scale their images so they are about the same brightness, in which case the longer exposure will look less noisy to the eye (which is what matters after all).

NigelM

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

  • 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.