Re: Pixel scale vs seeing? Re: [ap-gto] Close up of M81 without CCDT67

Benoit Schillings

check out some info on sampling:

Maybe a good starting video :

The minimum sampling also has to take into account the fact that pixel
are square, so your pixels are sqrt(2) on the diagonal.

A lot of talk over the last 20 years about this, you more or less need
to sample your image at 3x or more the FWHM caused by the seeing.

-- benoit

On Sat, Feb 15, 2020 at 3:38 PM Richard Spencer via Groups.Io
<> wrote:

Rolando et al.,

Some confusion on my part--I thought it doesn't provide any resolution improvement to image with a pixel scale below the seeing, ~1 arc sec per pixel in my neighborhood (on a good night!). Never 0.3 or 0.4; is it nevertheless useful to image at such a fine pixel scale for e.g. galaxies (tiny things)?

Rick Spencer
Elkton, MD
AP 1100 with CCA-250 with QSI 683 (8300 chip)

On Friday, February 14, 2020, 04:10:55 PM EST, uncarollo2 <> via Groups.Io <> wrote:

You are running 0.6 arc sec per pixel, which to me is not oversampling for high resolution imaging. In fact, for galaxies i prefer 0.3 to 0.4 arc sec per pixel which really brings out fine detail. My 17"F8 astrograph and the QSI 683 has such pixel scale and really does a superb job on small faint galaxies, even here in Northern Illinois. You have to have good tracking, of course, and on the best nights I can get below 0.15 arc sec RMS with the 1600 encoder mount.

I see a lot of images that are way undersampled (3 to 4 arc sec per pixel) with poor guiding that produces thick stars and very little resolution. To me these images resemble Brownie camera snapshots versus images taken with an 8x10 view camera. (shows my age, doesn't it) :^))


-----Original Message-----
From: Robert Chozick via Groups.Io <>
To: main <>
Sent: Fri, Feb 14, 2020 2:14 pm
Subject: Re: [ap-gto] Close up of M81 without CCDT67

Thanks Roland. My last dark sky outing was my first use of this scope and camera. I am really confused on the whole image scale question. I bought this camera because it has the largest pixels of any of the CMOS cameras. The image scale is .6 with this camera and 1600mm. My scale is about 1.8 with my FSQ 106 - 530mm. If the recommended guidance of a scale of 1-2 for image scale is used the 1600mm should be too small an image scale. Most CMOS cameras have only 2.5-3.5 micron pixels vs 4.63 on my ASI294. Is oversampling bad? The .6 scale in my image sure looks ok. What about .4 or .3? I intentionally did not get larger than a 1600mm focal length because of this issue (and yes, guiding issues are not as bad vs 2000 and over). Everyone asks why I got an 8 RC instead of a 10 or 12 inch RC. The above reason is why. Each new CMOS camera that comes out still has really small pixels. How good would 2.5 micron pixels look on a 2500 mm scope?


On Feb 14, 2020, at 1:18 PM, uncarollo2 via Groups.Io <> wrote:

That's really nice. Sharp and great color.

A question: do you think that 1600mm is a sweet spot for all kinds of deep sky imaging, especially for high resolution work? Especially since the newer Cmos cameras have such small pixels and would be able to take advantage of a high resolution optic for small faint galaxies.


-----Original Message-----
From: Robert Chozick via Groups.Io <>
To: main <>
Sent: Fri, Feb 14, 2020 12:06 am
Subject: [ap-gto] Close up of M81 without CCDT67

I got another shot of M 81 on the same trip as the M81-82 image, this time at f/8 1620mm.

Robert Chozick

Robert Chozick

Join to automatically receive all group messages.