Calculating sky brightness/bortle rating with images?

[ONIKKINEN]

I came across this tool a while ago: http://tools.sharpcap.co.uk/

This tool was part of the ideal exposure time calculations presented by Dr Robin Glover and with it you can figure out a sky electron rate per given level of light pollution and telescope.

It occured to me that i could use it backwards by figuring out how many electrons per second do i get in an image.

Last night i was in a "bortle 6" area according to light pollution map and clear outside but the sky is significantly worse than this for many directions. I noticed that i got at worst 1700 ADUs as a median value over offset whereas at best i have gotten down to 150 (measured by NINA). I know the gain i used has a rate of 0.25 electrons per ADU and i know the exposure time which was 30s.

So 1700/30 is 56.6 ADUs per second * 0.25 gets me 14,15 electrons per second. What i dont know is my effective QE since i am imaging with a OSC camera. ZWO quotes a peak QE of over 80 for their version of the IMX571 colour camera and i will assume it is around the green spectrum as it is with the mono version. Most of light pollution is also around the green spectrum and a bit towards the red. 

With this in mind would it be reasonable to assume a QE of around 70% for this specific case? I think its not far off at least. I also dont quite understand how the tool changes when you select mono and OSC. With OSC it seems to cut electron rate to a third.

Plotting everything in which in my case is F4.2 and 3.75 micron pixels, 70% QE and a colour camera i get close to 14,15 electrons per second at a staggering Bortle 8.2 or sky brightness of 17.8!

Did i get this correct? Someone smarter than me might be able to make more sense of this but i think its at least in the ballpark. Could be useful to gauge conditions at a glance without a sky quality meter.

[vlaiv]

Here is simple guide on how to best to do it:

1. calibrate your sub properly (remove dark current and apply master flat)

2. since it is OSC - do bayer split rather than debayering and take one green sub image to work on (this will reduce your pixel scale by factor of two - but it does not matter if you plate solve)

3. plate solve to get pixel scale

4. convert ADU values to electrons

5. select "empty" part of the image and measure median - convert to electrons per arc second squared with pixel scale (don't divide with exposure - no need)

6. find star of known magnitude that is not clipping and measure its flux

7. find sky brightness in magnitudes by flux ratio between star and background

 

[ONIKKINEN]

2 hours ago, vlaiv said:

Here is simple guide on how to best to do it:

1. calibrate your sub properly (remove dark current and apply master flat)

2. since it is OSC - do bayer split rather than debayering and take one green sub image to work on (this will reduce your pixel scale by factor of two - but it does not matter if you plate solve)

3. plate solve to get pixel scale

4. convert ADU values to electrons

5. select "empty" part of the image and measure median - convert to electrons per arc second squared with pixel scale (don't divide with exposure - no need)

6. find star of known magnitude that is not clipping and measure its flux

7. find sky brightness in magnitudes by flux ratio between star and background

 

Looks like we have very different definitions of simple 😅.

I know how to do 3-4 and but the others? No clue. Well i mean obviously i can do 1 or 2, but not both? When stacking master calibration frames on a single sub DSS and SIRIL debayers the frame and the split CFA channels option in Siril is no longer usable (since its debayered i think?). But lets try anyway, i have an uncalibrated green sub, but if i just use a part near the middle where flats are not necessary i think its not that important. Dark current is effectively 0 since there is almost no dark current at -20c and offset is 769 ADUs so this is easy to just subtract from the calculations. How do i do 5-6-7 then?

Was there something obviously wrong with what i concluded? Lets say im happy with an error margin of a few electrons here and there. The point is just to have some idea better than a wild guess as to what the sky conditions are if i am in a new location or transparency is bad etc.

[vlaiv]

1 minute ago, ONIKKINEN said:

Looks like we have very different definitions of simple 😅.

I know how to do 3-4 and but the others? No clue.

Well, if you post single frame and flat, flat dark and dark master - I'd be happy to do it for you and make walk thru for doing it in ImageJ / AstroImageJ

 

[7170]

@ONIKKINEN check out ASTAP. It will work out the SQM for you automatically based on a solved mage. The website has a good guide for the software as a whole and all the various functions.

[ONIKKINEN]

9 minutes ago, vlaiv said:

Well, if you post single frame and flat, flat dark and dark master - I'd be happy to do it for you and make walk thru for doing it in ImageJ / AstroImageJ

 

I have calibration masters as TIF files from DSS and the sub as a FITS file, would this be an issue?

Also, i dont have flat darks matching the flats and cannot take them at the moment as the camera cannot reach the temperature needed from ambient (it was -17, camera set to -20). Darks are also for -10 instead of the sub temperature of -20. But i guess better than nothing? Here goes:

2021-12-07_06-13-42_30.00s_0058-1.00-Target-3.17-58.fits

MasterOffset_Gain100.tif

MasterFlat_Gain100.tif

MasterDark_Gain100_30s.tif'

For plate solving: 3.75 micron pixels and a focal length of 840. The FITS header has a wrong value in it. The bright star next to the comet is HD 129132.

[ONIKKINEN]

22 minutes ago, 7170 said:

@ONIKKINEN check out ASTAP. It will work out the SQM for you automatically based on a solved mage. The website has a good guide for the software as a whole and all the various functions.

Thanks, i was actually already using ASTAP for many other things and didnt know there was this kind of module. Running my subs through it report a wrong altitude though, any idea how to fix that? It says one of my subs is under the horizon and one where it should be higher than 30 degrees at 12.

[vlaiv]

I got SQM of 16.97, so quite close to your original estimate of SQM 17.8

As far as I can tell - that is normal, even if your site is not that bright at zenith.

Target is at alt 30° as far as I can tell from the data and in direction of major LP source - Helsinki.

You are on border of red and white zone at SQM 19.5 judging by lightpollutionmap.info so above SQM17 is to be expected under given circumstances.

 

 

[ONIKKINEN]

1 hour ago, vlaiv said:

I got SQM of 16.97, so quite close to your original estimate of SQM 17.8

As far as I can tell - that is normal, even if your site is not that bright at zenith.

Target is at alt 30° as far as I can tell from the data and in direction of major LP source - Helsinki.

You are on border of red and white zone at SQM 19.5 judging by lightpollutionmap.info so above SQM17 is to be expected under given circumstances.

 

 

Ran many of my subs from different nights in the same location through ASTAPs SQM measurement tool and got values ranging from SQM 17.7 to the east down to 20.2 to the north on an excellent transparency night. The worst sub i found was mag 16.3 but from another location. Surprising range of values, and especially since the zenith is not the darkest spot. Although that could also be due to rarely imaging towards zenith and so small samplesize of subs and conditions. I would normally never image towards east because of the Helsinki nebula being bright enough to cast shadows to around 50 degrees, but Leonard was there so kind of had to.

Wondered for a while how you figured out the location but then realized the FITS header also writes the coordinates reported by NINA to the header.

[vlaiv]

1 minute ago, ONIKKINEN said:

Wondered for a while how you figured out the location but then realized the FITS header also writes the coordinates reported by NINA to the header.

Yes, it gives time, target coordinates and approximate shooting location

With such early target I wondered about astronomical darkness and twilight, but it was still deep darkness.

3 minutes ago, ONIKKINEN said:

Ran many of my subs from different nights in the same location through ASTAPs SQM measurement tool and got values ranging from SQM 17.7 to the east down to 20.2 to the north on an excellent transparency night. The worst sub i found was mag 16.3 but from another location. Surprising range of values, and especially since the zenith is not the darkest spot. Although that could also be due to rarely imaging towards zenith and so small samplesize of subs and conditions. I would normally never image towards east because of the Helsinki nebula being bright enough to cast shadows to around 50 degrees, but Leonard was there so kind of had to.

Indeed, there can be quite a bit of variation.

There is handy tool called SQC - sky quality camera software - that I can't find anywhere online - except results. You can see those on lightpollutionmap.info when you turn on SQC measurement - it looks like this:

With simple guide camera and all sky lens - we could do something similar - create map of LP for our locations (above image is actually fairly close to my current location - about few km away).

 

[Fegato]

17 hours ago, 7170 said:

@ONIKKINEN check out ASTAP. It will work out the SQM for you automatically based on a solved mage. The website has a good guide for the software as a whole and all the various functions.

Thanks - hadn't spotted that in ASTAP, I'll have a go!  PS I used to live in Surbiton many years ago...  I've moved to a cloudier but darker location more recently!