output maps in subframe selector script

[lucchett]

Probably a stupid question, but I cannot figure out how to find /read that multi image fits file...

I can write a file, but I always see only the first map.

Thanks a lot,
ANdrea

[mschuster]

Hi Andrea,

Please place an input frame and the generated map file in DropBox and I will check it out. What OS are you using?

Thanks,
Mike

When writing maps, you should see something like this in the log as SubframeSelector writes each map file, which contains three images:

Code: [Select]

Writing star map image file: C:/Users/Public/Documents/Astronomy/PixInsight/FITS/light_01_c_s_m.fit
0 FITS keywords embedded
Writing FITS image: 32-bit floating point, 1 channel(s), 1663x1252 pixels: done
0 FITS keywords embedded
Writing FITS image: 32-bit floating point, 1 channel(s), 526x423 pixels: done
0 FITS keywords embedded
Writing FITS image: 32-bit floating point, 1 channel(s), 526x423 pixels: done

Then when you open a generated map file, the log should display something like this, and three image windows should appear in the workspace:

Code: [Select]

Reading 1 image(s):
C:/Users/Public/Documents/Astronomy/PixInsight/FITS/light_01_c_s_m.fit
Image 1/3:
14 FITS keywords extracted
Reading FITS: 32-bit floating point, 1 channel(s), 1663x1252 pixels: done
Image 2/3:
13 FITS keywords extracted
Reading FITS: 32-bit floating point, 1 channel(s), 526x423 pixels: done
Image 3/3:
13 FITS keywords extracted
Reading FITS: 32-bit floating point, 1 channel(s), 526x423 pixels: done


[lucchett]

Thank you Mike.
I have found one of your previous post and now I think the issue is due to the lack of stars for analysis.
The plot file does contains several images.
I can update one image but unfortunately the quality is  low.
I am trying to help a friend in tilt analysis, I will ask him for better data.

Thanks again,
Andrea

[mschuster]

Ok, Andrea. You might have your friend try a Bahtinov mask for tilt measurement, focusing on each corner and measuring the focus position difference. I find this works much better than any single frame estimation. See my blog on how I did this. My fix was to install a device in the imaging train to counter the sensor tilt. Results with my FSQ were incredibly good across the entire frame.

Mike

[lucchett]

This is very interesting!
I am also installing an Atlas on my Tec.
I will let you know after reading the blog.
Thanks!
Andrea

[lucchett]

Mike,
Thank you for having pointed me to your blog, it is full of insights.

One question on the real life tuning of the tilt.

I suppose that, as you said, you start measuring the out of focus in each corner.
Any suggestion on how to run the iterative process?
Do you take a corner as reference and adjust to it?
Do you start with the opposite on the ccd diagonal?
How do you adjust/rotate the tilt adjuster vs ccd? ( no sure if I am clear here...)

Thanks a lot!
Andrea

[mschuster]

Hi Andrea,

The tilt adjuster does not rotate. It has three cone shaped screws around its outer edge that screw radially inwards/outwards. So you tweak those. The cones separate two opposing plates held together with strong springs.

With the four corner measurements I estimated the angle of the tilt across the sensor and which direction was high and which low.

Then I flipped the direction to match the actual sensor orientation in the camera. Needed to account for sensor firmware readout direction, software download flipping, and software opening flipping (PI's FITS up/down flipping). A headache.

Then I roughed in the three screw settings using a digital micrometer measuring the outer edge of the tilting device. These measurements needed to be scaled up by the ratio of sensor size to tilt device size. I did not pick one corner as a reference, instead I wanted the difference in tilt device thickness across its diameter in the proper direction to counter the sensor tilt, and the thickness across a perpendicular diameter to be equal, so as to not shift the tilt to a new direction.

By shear luck the result ended up reducing the tilt in the proper angle (as near as I could tell), but not enough. Basically the residual tilt was now half as large.

So I tweaked the screws again using the micrometer. This ended up being nearly perfect. I tightened the locking screws, still excellent. Done.

Mike