SharpCap Forums

Captured 6th April 2025 at 19:20 after setting up and waiting for dark. The Sun was still up. All the wrong gear - Vixen 81s @ F7.7, Red Filter and Atik 314L+. I missed the JMI hands off focuser on my Celestron C8. Captured in 16-bit MONO, SER file, at 0.5fps. Processed in AS!3, best 50% of 90 frames, a light touch in Registax.



Dave

Nice and simple. Looks like a steadier version of what you'd see visually.

(I took your advice to try the 12 inch on the moon -- the first night the 6'th was all awful images -- but the 7 'th was really sharp. Amazing how steadier air not only inherently improves the image but also allows you to find good focus which is nigh on impossible under unsteady skies)

Tim

Steadiest atmosphere tends to be early morning I have found. Also a Wratten #25 (or red) filter can help steady the atmosphere.

Dave

I also captured this at the same time - 32 FITS frames with the R filter. Best 25% of 32 frames stacked in Autostakkert!3, finished off in GIMP (Unsharp Mask, Brightness, Contrast, Curves, Crop).



The values that matter in CameraSettings:

[Atik 314L]
Exposure=6.5000ms
Capture Area=1391x1039
Colour Space=MONO16
ZWO EFW=R
Duration=62.116s
FrameCount=32
ActualFrameRate=0.5152fps

A single FITS frame:


Dave

A reprocess using PixInsight, using deconvolution to display sharper craters across more of the surface.

• PIPP to convert the SER file to FITS

• FFTRegistration to stack 90 files, first file used for registration

• Deconvolution: StDev 0.7, Shape 2, Aspect Ratio 1, Richardson-Lucy, Iterations 15

• MultiscaleLinearTransform to sharpen: layer 2 0.1, layer 4 0.05

• RangeMask: inverted to protect bright areas

• LocalHistogramEqualisation with mask: Kernel radius 50, Contrast 1.5, Amount 0.5

• HistogramTransformation with mask to brighten image

• CurvesTransformation with mask for contrast

Comparison of image in post #1 v image in this post.


Dave

Hello Dave,

Nice image, but the processing ... I prefer a shorter and more direct processing.

Concerning the details in your image. I downloaded the image and I remark something strange in the pixels.
What sort of processing gives 4x4 "intern" pixels ... here a strong zoom at the lower part of the moon:
Regards,
Jean-Francois

Jean-Francois

Thanks for the comment. The workflow was inspired by https://astrodoc.ca/moon/ , the site of Ron Brecher.

Ron Brecher is one of the PixInsight Ambassadors https://pixinsight.com/ambassadors/. He keeps company with such as Adam Block (95 APODs), https://pixinsight.com/ambassadors/AdamBlock/index.html and Warren Keller (author of Inside Pixinsight), https://pixinsight.com/ambassadors/Warr ... index.html .

I believe that trying new things is how I progress - nothing ventured, nothing gained.

I had a setup ready for a deep sky imaging session and whilst waiting for astronomical darkness, I captured 3 minutes of lunar data with a red filter. The camera, an Atik 314L+, is not the best camera for lunar work (16-bit MONO only and 0.5fps) and the scope at f/7 not ideal. I was surprised at how well the image turned out. I will certainly pursue the deconvolution technique for lunar data.

Here is an image processed in a more conventional manner - Autostakkert!3 best 25% of 90 frames, waveSharp 2 for sharpening, derind, gamma, contrast.



Dave

Hello Dave,

Tanks for the additional information.

It exists more than one "way" for an image from the camera to the monitor.
Some are simple and basic, other ... all is possible.

Is the FFTRegistration a global or local stacking process ?

Concerning the strange "4x4" pixel ... do you know where they come from ?
I ask while you do a lot of work ... but the final image seems to be 4x less resolution as it could be.

Regards,
Jean-Francois

Jean-Francois

Is the FFTRegistration a global or local stacking process ?

No idea. The script will register and integrate images & correct for rotation.
FFT registration appears in MATLAB, https://www.mathworks.com/matlabcentral ... ion-method

final image seems to be 4x less resolution as it could be.

Looking over the images I have, I think the one I posted was run through PixInsight's IntegerResample with an Upsample factor of 4 - trying to squeeze more out the original image which was captured with an 81mm refractor.

Dave