Jump to content
  • Join the online East Midlands astronomy club today!

    With active forums, two dark sites and a knowledgeable membership, East Midlands Stargazers has something for everyone.

Optimum ISO


Clive

Recommended Posts

Since acquiring my Canon EOS200D just under a year ago, based upon what I had read on the WWW at the time, I have always used an ISO of 1600.  However, I recently came across article on the WWW regarding the best ISO for astrophotography which put some doubt in my mind.

 

When taking astro-photos with my 400mm F5.6 lens, at ISO 1600 I limit the exposures to 180s as this gives a sky background (as measured by DSS) of just under 10%.  So I thought I would measure the camera noise at various ISO settings whilst keeping the exposure such that (theoretically) I would still end up with a 10% sky background.  Then, based upon a maximum possible signal of 14336 (2^14 – 2048), I could calculate the signal to noise ratio.

 

ISO         Exposure                Noise                SNR

400         720s                      160                   39dB

800         360s                      131                   41dB

1600       180s                       93                    44dB

3200       90s                         64                    47dB

6400       45s                         85                    45db

12800     22s                         133                   41dB

25600     11s                         222                   36dB

 

Based upon this it would appear that for this camera I should be using an ISO of 3200, which would be great if it’s true as I could get twice as many subs in the same time than at present and have less noise.

 

The real test will be to try it, but before I get a clear night, if anyone can see where I’m going wrong, please let me know before I spend an age re-doing all my darks and flats!

Link to comment
Share on other sites

I was always told the best ISO for a Canon was 800.

Mind that was way back when I had a 350 

Link to comment
Share on other sites

Thanks for the info. If I remember correctly, the ISO1600 recomendation  was for photos using a 'standard' lens (I didn't have a telephoto then). With my 100mm F2.0 lens the maximum exposure is 45s to get a sky background of less than 10% so at the weekend I'll repeat the test with these shorter exposures to see what I get.

Link to comment
Share on other sites

Does this not depend on the specific sensor for any given DSLR ?

A bit of a horses for courses type of thing....?

 

For my Camera (6D) i've read that iso 800 - 1600 works best.

Link to comment
Share on other sites

Yes it is very camera dependant and their seems to be a lot a different opinions out there. I've read that the 6D is highly iso-variant and the recommended ISO is 6400 whereas the 80D is iso-invariant and the recommended ISO is 200! Unfortunately, when I bought the camera it had only been recently released so their was little about it on the web. In a way I'm hoping mines on the 80D side - I'm not sure I could cope with 24 min exposures!

Link to comment
Share on other sites

There’s some articles on cloudy nights by Michael Covington he was looking to see wether the 200d was better than the Nikon d5300 ( which I have ) the Nikon faired slightly better I believe the 200d is iso invariant so above a certain iso you gain nothing  bar more noise but better dynamic range at a certain point that I believe is 400 for yours .

Link to comment
Share on other sites

Thanks for the info. I've read in several places that the D5300 is a good choice for astrophotography. I'm back at work today but when I get the chance I'll look up those articles. 

Link to comment
Share on other sites

Having looked around the web over the weekend, it appears that the signal to noise ratio (dynamic range) of a camera is estimated by taking a single exposure at each ISO setting using a constant shutter speed.  When using my 100mm lens at F2.0 for wide angle shots of the sky, I limit the exposure to 45s at ISO 1600 in order to avoid excessive sky background from my light polluted sky (< 10% as calculated by DSS).  I therefore took a 45s ‘dark’ (i.e. with the lens cap fitted) at each ISO setting between 100 and 25600.  The raw files were then imported into IRIS and the RMS noise within the central 512 x 512 pixels (an area with no hot pixels) of each image measured using the IRIS STAT2 command.  The results show that the camera is ISO-invariant (as corroborated by several web sources):

 

large.1770613039_FIG1.png.ba74b0617fa2e0a240a763deda3e72f0.png

 

Based upon the above graph, I could increase the dynamic range of the resulting image by nearly 4 EV by exposing at ISO 100 rather than ISO 1600 (with the consequential increase in exposure time from 45s to 720s to maintain the same exposure value).  To check if this is true I took a series of ‘darks’ at each ISO setting between 100 and 25600 but this time changing the exposure time to give the same effective exposure value for each image (based upon 45s at ISO 1600).  The result clearly shows that the optimum ISO is actually 1600, any less does not improve the dynamic range whereas any more decreases the dynamic range.

 

large.1046768020_FIG2.png.dc28c345d696297aeb68202fce05dd99.png

 

The above tests were obtained with the camera in a warm room. The average camera temperature extracted from the images EXIF data using EXIFLOG was 31.1°C (constant exposure time) and 29.7°C (constant exposure value).  The second test was therefore repeated with the camera placed in a fridge resulting in an average camera temperature of 11.9°C.  The plot clearly shows the effect of temperature with the dynamic range increasing to just under 11 EV (an increase of nearly 3EV!) and the optimum ISO is now 200:

 

large.798264490_FIG3.png.8d560faa0efb68006f132e6bafb0d058.png

 

To achieve a DSS calculated sky background of less than 10% with my 400mm F5.6 lens, the maximum exposure must be limited to 180s at ISO 1600.  Using this as the ‘standard’ exposure value results in the following plot showing that the optimum ISO value may actually be 3200:

 

large.528684163_FIG4.png.4f8a738b084d573296144999b191361e.png

 

So it looks like that for an ‘average’ night where the ambient temperature is around 5°C, if I’m using my 100mm lens I should be exposing for around 360s at F2.0/ISO 200 or 90s at F5.6/ISO 3600 with my 400mm lens to achieve the best signal to noise ratio.  The next step will be to try to understand the data and then to try and create a noise model so that I can set the ISO and exposure time to the optimum value for whatever the ambient temperature happens to be - and then to actually try it out for real.

 

And despite all the information I’ve read, I now feel the need to see what the built-in long exposure noise reduction does ….. !

 

  • Like 1
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
×
×
  • 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.