Good job! Works fine. Some suggestions from my side:
file naming convention - putting mean ADU before exp. time would be great
better calculation of exp. time when changing filters - in the setup add coefficient to each filter to divide/multiply last exp. time to estimate first exp. time for changed filter, let say L has coefficient 1, R has 3, then when changing from L to R the first exp. time for R would be last exp. time / 1 * 3
add bias mean value to the setup, to better estimate linear behavior of the chip and better estimation of the exp. time
As for point 1, sure that is possible. Any particular reason for this change in order ?
Regarding points 2 and 3, initially I had similar ideas. But in the end, for my setup, the exp time calibration takes only 10-15 seconds per filter, with about 3 iterations, although the first cal frame might indeed have either very low ADU or very high ADU. It goes quickly because I use only the center 1/3 of the frame for the calibration, so download times are negligible. Do you see something different that merits some changes ? Dont hesitate to PM me a log file from Skyflats to show your calibration sequence.
On a general note, I was planning to advertise the application on other forums (eg CN), as not all SGP users follow this one, but before that I was expecting to deal with bug reports posted here. Since that has not been the case, I wonder if people have had the chance to test it (successfully), or if people have had problems with it? A message in any case is appreciated!
Actual file name has format 2021-10-22_SII_1x1_1.9076sec_frame1.fit, but for flats the mean/average ADU is the most important value that matters. I would like to sort the flats according to the mean/average ADU.
Atik 16200 download time is around 12-15s. If you have CMOS kamera, it does not matter, but for CCD cameras it does.
@drvo I see from the log that you use full frame for the calibration step. I suggest you go down to 1/3 frame for the calibration by using this option:
It will reduce the download/processing by a factor of 10.
I also have a CCD, an Atik 460ex with 6Mpixels. Download time is about 10 s for a full frame 1x1 bin, but only about 1s for a 1/3 frame.
I will try it. Does it mean that the final calibration flames will be binned 3x3? Or is this only for computing exp. time from center of the image (ROI)?
This is a ROI, so 1/3 frame in the centre of the image, still binned 1x1. If you have a lot of vignetting it might bias a bit the ADU of the flatframes. In my case it works well and it really speeds up the calibration, to a point that it doesnt matter that the first frames are really off in ADU.
I’m setting up SkyFlatsSGP and have run into a problem. I have everything setup correctly and have double checked the latitute/longitude coordinates for location and time, but SkyFlats is saying that sunset is at 04:16, which I believe is 4AM! Log attached. Any idea what I’m doing wrong? skyflatssgp-log-2021-12-30 08-07-12.txt (1.0 KB)
First test entering the latitude and longitude as integers instead of decimals, maybe there is a parsing issue.
Second, in the calculations the program must pull the local time zone and eventual daylight saving time shifts from the regional settings of your windows account, so make sure they are properly set.
What was the expected sunset time at your location ?
The negative longitude value was the solution of course. I should have caught that myself. Hoping to try SkyFlats soon. Thanks for providing it for us!
joel
