Arcsecond per pixel ‘rule’...When is it important?

[AntHart]

Hey  

I’ve been watching a few vids on YouTube tonight about matching cameras to lenses/scopes.

Kept hearing arcsecond per pixel so googled that and found Atik cameras website with a table.

pixel size of the cameras chip divided by focal length times 206 to get the ‘arcsecond per pixel number’

From what I gather in plain English if the number you get is between 1-2 your golden and will have a beautiful detailed image...Awesome!...except...

I did the calcs for a ‘well known/well advised setup for beginners like myself’

Canon 6d - Samyang 135mm ...it came out almost 10!!!

Ive seen stunning images with this setup so how is this calculation right?

........

Is it because it’s such a wide field view when you zoom right the way in/pixel peep it would look ‘bad’ anyway regardless of the camera used?...not sure I’ve described that the best but sure you get my meaning!

Is this more a ‘rule’ for very long focal lengths getting right in on the action and something not to worry about with the upto 200mm focal lengths I’ll be intending on using? 
 

It’s got me puzzled!

Cheers for any advice 

Ant 

[ollypenrice]

The sampling rate, as you're already suspecting in your post, needs to be considered in terms of what kind of target you're imaging and how you want to look at the picture when it's finished.

Most galaxy images projected onto chips by amateur focal length telescopes are small. For the final image to give a decent screen size and contain pleasing details the pixels of the chip must, therefore, also be small. A sampling rate of about 1"PP will offer this. Much below that and you'll be very lucky not to see the theoretical level of detail blurred out by seeing and guiding effects.

Once you have this image you'll want to show it at 100% (AKA full size), meaning 1 camera pixel is awarded 1 screen pixel. To dispense with the need for your viewers to zoom in to this scale you'll probably want to crop out some of the surrounding background sky. This is an example of a small, delicate target to be shown at full size. About 1"PP is necessary. (This will also require you to shoot a lot of data because full size presentation is very, very intolerant of noise.)

Most of the larger nebulae are quite diffuse and contain few tiny details, certainly fewer than galaxies. Their projected image covers most of the chip and many may need a mosaic. The idea is to present the entire nebula, which means you won't want to crop the image. It might be nice to present it full size so that viewers can zoom in and see the smaller details but that is a luxury, not a necessity. Although the tiniest details, where there are any, will not be resolved at coarser pixel scales an image shot at up to 3.5"PP will still look smooth, un-pixelated and and have round, rather than 'blocky' stars. For presentation at full size I would not, personally, go above 3.5"PP.

Widefield images are not generally intended to be zoomed in on unless they are made at high-ish resolution and constructed as mosaics. If not zoomed in on they will look fine, as you point out, at far coarser pixel scales.

Finally, if there are images you like taken with Equipment X then Equipment X will work for you once you've mastered it.

Olly

 

 

[Endolf]

There is some detail on this tool about it, it also shows that the "ideal" pixel scale changes with the sky conditions. Worth having a play. Given my camera (and a couple of the ones I'm looking at as future upgrades) I used to to judge the approximate maximum useful focal length for my skies when I was looking for a new scope for galaxies.