Old 17-11-2012, 08:27 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,822
Optimum pixel size


Have been working on a simple spreadsheet for calculating the optimum pixel size for imaging, based on maximising both resolution and signal-to-noise ratio and accounting for optics resolution, atmospheric seeing and tracking wander. The attached table shows the results for a few combinations of scope+mount and in 2 and 4 arc sec seeing (FWHM). A few simplifying assumptions went into this (eg perfect scope assumed, secondary obstruction ignored, sensor cross-talk not considered). However, the results should not be too far from reality.

Interesting how dominant the atmosphere is, how much effect the mount has in good seeing and how similar the final resolution is for the chosen scopes -even the AAT resolution is not dramatically better than a C11 in 2 arc sec seeing.

The standard rule of thumb of 2 arc sec per pixel might be a bit coarse - probably better to use something around 1 - 1.5 if you expect reasonably good seeing.

Most of the Sony sensors, the Kodak 8300 and older DSLRs would appear to be well matched to the shorter focal length scopes at around 5 microns pixel size. If you have a longer fl scope (SCT or RC), you will be best off with one of the bigger pixel Kodak sensors, or use binning. The optimum for a very fast scope (eg hyperstar) would be one of the new generation chips with small pixels.

regards Ray
Attached Thumbnails
Click for full-size image (scopematch.jpg)
131.0 KB60 views

Last edited by Shiraz; 17-11-2012 at 08:57 PM.
Reply With Quote
Old 18-11-2012, 07:20 AM
gregbradley's Avatar
Registered User

gregbradley is online now
Join Date: Feb 2006
Location: Sydney
Posts: 16,222
Nice work and your results mirror my experiences.

Although I have generally found 1 arc sec or .66 arc sec generally to be the standard used as per sampling theory.
Sampling theory states you need a minimum of 2X to obtain a decent sample. So 3X is often used then following that
theory. In 3 arc sec seeing, which is probably very common (on a good night perhaps) that gives you 1 arc sec/pixel as a guide
or if you have occassional 2 arc sec seeing then .66.

I have noticed though with my refractors versus my CDK that small pixels do make an image pop a bit more on the refractors but
seeing effects blur the images much more noticeably on the CDK to the point where I would not use the KAF8300 camera on my CDK17 with nearly 3 metres focal length. I have put 16803 camera (9 microns) and 8300 camera (5.4 microns) at imaging the same object (M104) and the 16803 captured far more detail with less noise and blur in the same conditions than the 8300. This was proven over several images.

The trouble with the 8300 chip is small field of view. Also small well depth. QE is the same. So that limits it as well. Its harder to get a pleasing image with a smaller field of view on the same scope.

Also smaller FOV CCDs will show up the tracking errors more easily as it creates a digital zoom so to speak due to the smaller field of view.

In real life though other factors are more dominant though than pixel size which is really only an issue with long focal length scopes.
QE, Noise, accessories, download times, weight, reliability, build quality, cost, field of view, autoguiding ability, focusers, etc etc
tend to become dominant. Of course cost is the main consideration usually as the large 11002/16803 chips are a lot more expensive than KAF8300 chips. Plus accessories and scope that can take them are vastly more expensive.

Another point with the larger chips is the increased sensitivity to tilt, squareness, flex and the level of correction of the optics (round stars to the corners without vignetting, only the largest and the best focusers/flatteners cope). They show these factors up more easily.

Another factor is cost of the larger filters.


Last edited by gregbradley; 18-11-2012 at 07:35 AM.
Reply With Quote
Old 18-11-2012, 08:20 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,822
Hi Greg.

I am planning a new system to image galaxies and wanted to understand the issues re sampling, so your confirmation of results to date is helpful. I was most concerned about maximising resolution and that means selecting an optimum pixel size for the scope/camera combination as the first step after choosing the scope.

the spreadsheet is nothing particularly special, but it does include the main variables and gives some idea of what is important. I wanted to look at system issues and used the Nyquist sampling requirement for a Gaussian - 2.35 pixel per FWHM - but of course, seeing cannot be predicted at anything like that precision, so 2 or 3 are probably just as useful. Conclusion to date is that, for seeing limited operation, I need at least 10 inches of aperture and a precision tracking mount - the spreadsheet is being used to evaluate various possible scope/camera combinations that fit those initial guides. However, I posted some results for other fairly widely used combinations of scopes and sensors for interest.

the 2 arcsec per pixel rule of thumb is pervasive on the web - a very quick search resulted in
My point was that it is too coarse and we agree on that.

Added the CDK17 to the spreadsheet - the 16803 is a much better choice than the 8300. With that focal length and aperture, you will always be seeing-limited and the 8300 would be down by a factor of about 2.5x in sensitivity compared with the 16803 - for no additional resolution, bloated looking stars and a much smaller fofv. The 8300 should definitely only be used with smaller scopes.

You raised a number of other system issues that should also be considered. Couldn't agree more. My current planning process is an attempt to find a cost/capability sweet spot that will provide seeing limited resolution with manageable imaging times and acceptable image quality - quite a challenge and fun as well.

anyway, thanks for your comprehensive post.
regards Ray
Attached Thumbnails
Click for full-size image (scopematch2.jpg)
138.4 KB27 views

Last edited by Shiraz; 20-11-2012 at 04:04 PM.
Reply With Quote


Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

All times are GMT +10. The time is now 07:53 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.
Lunatico Astronomical
Astronomy and Electronics Centre