New Feature : Autofocus Settings editor

[Jean-Francois]

Hello Robin,

I tested the new feature this morning. It simplify a little bit the parameters selection, but ...

- If the "Overshoot" is selected ... it is not remembered ... need to change the setting each time (and the number of steps)

- the scan direction "Positive" or "Negative" is not saved ... it comes each time in the opposite selection (see first point)

- several times, the Number of measurement points and focuser steps were not remembered.
I changed it to 9 and 200, but the scan was performed with 11/100.

- for the FWHM value (from other message) ... I remark that depending on the size of the red rectangle selection, the calculated FWHM value changed a lot. With the tool "FWHM Measurement", if the red rectangle is only 20x20 pixel around the star, then the FWHM value is correct. If the red rectangle is larger, then the FWHM value is larger (much larger).

Other variation of the calculated FWHM comes form the threshold value.
Here the same 30 sec image (from folder) with the same star, but with the threshold at 20, 25 and 30 % (Yes, I know ... lot of moon background illumination and a little bit too much push the gain and exposure

FWHM_threshold_20.jpg (216.67 KiB) Viewed 120 times

FWHM_threshold_25.jpg (217.07 KiB) Viewed 120 times

FWHM_threshold_30.jpg (216.14 KiB) Viewed 120 times

Regards,
Jean-Francois

[admin]

Hi Jean-Francois,

thanks for the test report - always useful

For the overshoot settings, did you press the 'Apply' button? That is what saves those settings and applies them to the focuser. That is probably anti-intuitive with the way the main buttons at the bottom of the window work now that I think about it, so I will revisit that.

I will check out the other issues regarding saving or not of the settings. I would expect the settings set in that window to be applied

* temporarily during a 'Test' run from that window (but the old ones are put back when the test finishes, so that 'Cancel' can still exit with no change applied'
* When pressing the save changes button - then applied permanently

The single star FWHM measurement tool is fairly basic (also old code) and is nowhere near as sophisticated as the multi-star FWHM tool. It doesn't do any automatic background estimation like the multi star tool does, relying on your choice of the 'black level' setting to cut off the measured area near the visible edge of the star. Failure to do this, particularly with a large box will give dodgy readings (at least in terms of absolute values).

I think if I wanted to measure the real FWHM of a single star I would be tempted to use the multi-star FWHM tool and a small selection box around the star of interest these days - I wonder if the single star option should go or be redone as a 'one star only' version of the multi-star...

cheers,

Robin

[Jean-Francois]

Hello Robin,

Yes, I press "Apply" and "Save Changed".
When I test with the Test camera and Focuser Simulator ... the "overshoot" settings are not remembered.

OK, for the FWHM tools. You updated the multi-stars, but not yet the single star.

With the multi-stars FWHM tool ...

If I select only 1 or 2 stars, it does not work ("Too Few Stars").
With larger rectangle and more stars and it can work, but not every times.

FWHM_threshold_multi-star.jpg (238.98 KiB) Viewed 77 times

FWHM_threshold_multi-star_6.jpg (250.77 KiB) Viewed 77 times

FWHM_threshold_multi-star_6_success.jpg (246.36 KiB) Viewed 77 times

Why I need the single star FWHM tool ? ... I started a new script for the quick and precise adjustment of the camera tilt on a RASA telescope.
The user select one star and move it at 3 defined positions ("near" the adjustment screws), the script moves the telescope at 6 positions for the precise focus calculation. It work fine manually, but I wanted the next level with full automation (means with the autofocus). I need to check if the multi-stars on a small rectangle will work too.

Regards,
Jean-Francois

[Jean-Francois]

Hello,

and here the resulting image ... M33 in RGB with R = 900s, G = 600s and B = 1080s (with crop / resize and JPEG higher compression).

M33_RASA_QHY268_RGB_small.jpg (896.91 KiB) Viewed 77 times

Jean-Francois

[Jean-Francois]

Hello Robin,

I'm now doing some tests with the GSS, the ALPACA Sky Simulator and the ASCOM Focuser Simulator.
With the FWHM Multi-Stars tool ... the button "Refocus" is all the time deactivated. It was not the case yesterday with real hardware.
I can perform an automatic focussing with the "Test" button from the "Edit All Settings".

I remark a small calculation error.
If you can select 2 measurements points with 100 steps in between, you will have 100 steps total scan distance.
Now with 9 measurements points and 100 steps, SharpCap shows "900" as total scan distance. My calculation gives me "800"

Jean-Francois

[admin]

Hi,

yes, you are right on the miscalculation of the total distance - I spotted that one too and it is fixed in yesterday's update.

On the 'Refocus' button, I think the problem is simply that nothing is triggering it to recalculate its availability after using the profile editor - I will fix that too.

cheers,

Robin

[Jean-Francois]

Hello Robin,

Concerning the auto-focus and the FWHM Multi-stars tool ...

I did yesterday evening some tests with telescope/camera and sky.

First the question of the FWHM value difference. Here the same sky region with same camera parameters.
One time with the "Live Stack" FWHM History (FWHM around 7) and one time with the FWHM Focus tool (FWHM = 2.2).

FWHM_focus_tool_1.jpg (296.07 KiB) Viewed 11 times

FWHM_focus_tool_2.jpg (319.46 KiB) Viewed 11 times

For the Focus tool, the size of the measurement region (red rectangle) has a different effect as expected.
If the rectangle is larger, it has less stars ! Half size rectangle has more than double stars.
It was possible to reduce a little bit more the rectangle with more than 350 stars, compared to the rectangle with the full image width with only 114 stars.

FWHM_focus_tool_5.jpg (268.1 KiB) Viewed 11 times

Regards,
Jean-Francois