Exposure Time vs ISO

[Skuber]

Apologies for the beginner question, but:

In the books on photography I've read, where all other factors are equal (same F ratio, perfect guiding and all that), they seem to suggest that you would get the same results if you doubled the exposure time, or doubled the ISO rating.

So, for example, ISO 800 at 60 seconds will give the same result as ISO 1600 at 30 seconds.

Is this correct?

Do both of these settings yield the same amount of noise, or is one likely to be more noisy than the other?

I see alot of people here that shoot at ISO 800 or ISO 400. All of my pics have been shot at ISO 1600 so far - am I introducing noise into the shots that I could avoid by using a lower ISO setting?

Any thoughts would be appreciated

[brianb]

ISO 800 at 60 seconds will give the same result as ISO 1600 at 30 seconds.

Is this correct?

For normal photgraphy, yes.

In practice you have to remember that changing the ISO setting simply changes the amplification of the signal. It does not change the sensitivity of the sensor! With a high ISO you get a clearer image on the preview screen but the information is exactly the same in the raw image if you take say 30 secs at ISO 200 or 30 secs at ISO 3200 ...

The best strategy is to find the native sensitivity of your sensor (it's usually ISO 200 in modern cameras), take subexposures as long as your kit can cope with and take sufficient subexposures to build up the image density you need to show the object properly.

[Sam]

Hi Skuber,

There's a bunch of threads regarding noise but one of the best decsriptions I've seen in ages linked to this webiste:

Noise, Dynamic Range and Bit Depth in Digital SLRs

It covers all of the different sorts of noise that gets introduced into a digital image. A lot of figuring out what the best results for an individual setup comes from understanding the source of noise for your own camera. Also a lot of experimenting with different settings helps determine what works for you.

For my camera, a Canon 1000D modded, I've found the best results have come from ISO 400 but I think a lot of this depends of the sky conditions you have. Also a high ISO is good if there's a chance that aircraft trails could ruin your exposure etc.

Sam

[eddie3000]

I have a nikon d80. When I increase the gain on the sensor (ISO equivalent), I get photos with less dynamic range. There is also a factory default noise reduction algorythm (on the d80) that the user cannot turn off, thus reducing detail after about ISO800 (even in raw). This noise reduction system is independent from the other noise reduction options in the menu.

The CCD response isn't linear either. This means that the curves I need to apply in raw images for different ISO settings are different yielding slightly different overall appearance.

In general, I prefer the results with lower ISO and longer exposures.

[dph1nm]

Note that with modern Canon cameras, higher ISO gives lower read noise, so if read noise is a significant contribution to your exposure you will do better at higher ISO (which I guess many people find counter-intuitive). I have not seen any stats for other makes.

NigelM

[Sparrow]

This one really is a perennial chestnut to which I have never seen a definitive answer. I normally image at ISO800 (unmodded Canon EOS1000D) - In the cold weather I tried imaging the Californai Neb at ISO1600 for a change and wasn;t dissappointed with the results. However my guru friend is firmly of the opinion that I'd have got better results and more of the fainter detail by going for longer subs at a lower ISO. As I have now drastically improved my guiding I'm going to try subs at ISO400 on the same target for a comparison.

[James4]

Skuber, for Deep Sky Objects I crank the camera (Canon 300D, Nikon D5000) up all the way (maximum ISO). Because I shoot unguided, this is the only way I can capture some of the faint stuff with short exposures. The results are noisey and there's no doubt shooting at low ISO for longer gives much finer results. Try it both ways and compare the results. Good Luck!

[MartinB]

Lot's of different suggestions here Skuber but Sam's link is superb. You need to read this to get a proper understanding. One problem with astroimages is the high dynamic range needed to produce pleasing images. Our targets are typically very low signal but stars are easily saturated becoming bloated and bleached out.

[dph1nm]

In fact it can all be summed up in this quote from Sam's link:

Bottom line: Read noise at high ISO is much smaller than read noise at low ISO, in terms of the error in photon counting that it represents. Thus, better image quality is obtained for using the highest ISO for which the signal is not clipped.

NigelM