Like most websites, SGL uses cookies in order to deliver a secure, personalised service, to provide social media functions and to analyse our traffic. Continued use of SGL indicates your acceptance of our cookie policy.

# Pixel size question

## Recommended Posts

Hi all, I've been reading through the 'Choosing a CCD' sticky in this forum and wondered if someone could clarify a point for me.

I have a value using the formula resolution=206*(p(um))/(f(mm)). The guide says a value of 1”/pix will be about the highest usable resolution and a value of 1.5-2”/pix is quite good

My value is 1.236, is this better than the quite good value but still usable?

Edited by BigBeard

##### Share on other sites

the numbers are only one aspect of the choice, FoV, system F-Ratio and effects on usable chip area, prefered Targets, Budget Related filter choice. quiet a bi to weigh up outside the theoretical numbers.

##### Share on other sites

To ne honest I would not worry greatly. I initially read the SBIG site and they said 1"/pixel, and the reason I likeed it was very simply it is easy to remember. When I read other sites I found that there were values up to 4" and 6" and if I recall one at 8".

SBIG then pointed out that for planets you could go to 1/2" or 1/4" per pixel.

So in a way pick just about anything you feel like within what may be described as "reason".

Seems you "should" (according to a recent imager on a course) dertermine the Airy disk size and aim for that to be around 1.5 to 2 pixels, reason being that a star covered more then a single pixel and so was "less" square. Made sense at the time.

You seem to have the SBIG option of what is 1:1 or a bit bigger, actually I think SBIG did aim at 1" per pixel and also said 2" per pixel was good also. And that is where you are at present.

Go get an image and see how it turns out.

The other appoach is that saying "The scope and camera are operating at 1.236 arc seconds per pixel" sounds impressive.

Edited by ronin

##### Share on other sites

1"/pixel is the lowest limit that I would recommend to a budget amateur for several reasons.

You need a good mount to guide at that resolution - this means that for most part budget mounts (like HEQ5 / EQ6, etc) will even not be able to do that out of the box - they will need a bit of tweaking / modding (like belt mod, hyper tuning, etc). So there is certain level of skill needed to go even at 1"/pixel. Second is seeing - one needs good seeing to exploit 1"/pixel. Seeing needs to be less than 2" given scope sizes that amateurs usually use, and guiding precision to give ~3" or less FWHM stars. You will also like to throw the biggest aperture at that resolution and probably cooled camera to be able to record anything but the brightest of things (or spend many, many hours capturing subs).

1.23"/pixel is, depending on your skill level quite usable resolution. Advantage of having high resolution setup is that you can always bin things down if your gear / experience level is not up to the task. So by 2x2 bin you will get 2.5"/pixel - that is resolution at which most budget mounts can guide well. Only drawback is that by binning you are creating smaller images (not sure if this is important to you).

I currently have multiple setups (combinations of OTAs, cameras and focal reducers) that are in the range of: ~0.5"/pixel, ~0.72"/pixel, 1"/pixel, 1.3"/pixel, ... but when I process data I never go below 1"/pixel (0.5"/pixel I bin 2x2 to get 1"/pixel, etc ...).

• 2

##### Share on other sites

Yes, there is no point in going for a pixel scale which the seeing and guiding won't allow, and it does get increasingly difficult the lower you go in arcseconds per pixel. The trouble is that nobody can tell you what your limit is because it is defined by your mount, your guiding and the seeing on the night. With very good mounts (Mesu 200) I find that 1.8 "PP is nicely resolved, fast and easy. With half sized pixels, the same scope can take on small galaxies, which is great, at O.89"PP. But a number of things change at the same time. My filters are parfocal at 1.8"PP but not at 0.89. This is perfectly logical since critical focus is governed by pixel scale. Also the seeing becomes more critical. If the FWHM values on the night are not good, there is no point in shooting Luminance at 0.89"PP. I will, instead, shoot colour and wait for better seeing to shoot luminance. If I had to pick just one pixel scale for my (good) location it might be around 1.5 to 2.0 "PP.

(I'm only talking about deep sky here.)

Olly

##### Share on other sites

Thanks for the responses folks, I knew it wasn't that straightforward and now I have more to work with

##### Share on other sites

Funny enough I was looking at Atik on u tube yesterday.. the formula is pixel size divided by focal length multiied by 206.. you are looking at a value between 1 and 2..less than 1 and youre oversampling giving bloated stars and causing tracking errors..more than 2 and you re under sampling giving block stars..

##### Share on other sites

Pixel size ("/pixel) is a sampling size, which needs to be smaller than the smallest detail you want to capture. So, targets with a lot of small detail need smaller pixels (or longer focal length). But with smaller pixels, you catch fewer photons, simply because each pixel covers a smaller patch if the sky. You need longer exposures or a larger aperture to fill those pixels and clear the noise floor. There is always a trade off.

Amateur astrophotographers image at pixelscales (roughly) 1 - 3 "/pixel. Professional astronomers use < 1 "/pixel. E.g the Liverpool images that Göran Nilsson and I have been processing, were at 0.3"/pixel, already binned 2x2. The pixelsize was a whopping 15 micrometers (but with 20 m focal length). At 0.3 "/pixel, you need excellent seeing (i.e. a high mountain top above the clouds) and a large aperture to hoover photons (2 m). Professional astronomers don't use long exposure times, because time is precious and longer exposures means more risk for ionizing, cosmic radiation to hit the pixels. Amateurs otoh, use smaller apertures, rougher sampling and longer exposure times. Or a cartload of money.

## Create an account

Register a new account