Matching a Telescope to a camera - Image Scale/Sampling (side by side tests)

5 replies232 views
Rouz Astro avatar
I have been testing image scale and focal length variations for some time now and complied some interesting results.

 With the long focal length CDK14, I noticed images at native are a bit soft.Finally around to a side by side test.

Picked a target and imaged 2 hours at the native 2565mm and then 2 hours at a shorter focal length of 1700mm - keeping other factors constant.

Yes, the image at the reduced focal length was smaller - but it was up sampled to 150% to match the native one. 

Here you could say its not "real" information and its the software interpolating the data - that is true.However, with an excessively oversampled image, its the same case - at my native 7x sampling rate, there is no "real" data there either.

That's also just blowing up the image scale for the sake of making it larger without any real data. Perhaps other can chime in, but I don't notice any real resolution gain past 4x the seeing.

There are some here that feel 7x is possible, my hats off to them! The disadvantages are obvious - far dimmer image and much lower SNR, much smaller FOV.   

Yes you can bin, but it would be nice to have a 62Mpixel image with a much larger FOV. 

Wrote an article on the topic as well -  
https://rouzastro.com/guides/
Helpful Insightful Respectful Engaging
Rouz Astro avatar
Here is the difference in ADUs

Rouz Astro avatar
FOV gain

Rouz Astro avatar
Long focal length vs shorter+resample This was on the same night - so back to back. The target had moved to a lower altitude:

Rouz Astro avatar
I found it interesting that that there was no loss in resolution at all. It was clear the next night so I imaged the same target exactly the same and this time with reduced focal length at the same altitude as the previous night with the longer focal length. Here everything is the same including the altitude.

Rouz Astro avatar
With the increased brightness, I then compared the results with integration time: f/7.2 with 10 x 600sf/4.7 with 6 x 600s Both had similar SNR (pixinsgiht) and looked similar visually. Note we are using about half the time at f/4.7