Luminance Filter issue (?)

[AstroGS]

On Monday, I had the chance to take a few subs before the clouds appeared, with the new Antlia 3nm 36mm unmounted filters.

I am planning to shoot Markarian's Chain in LRGB but, I wanted to make sure that everything works as it should. I managed to get a couple of subs with the Red filter and Luminance.

The rig for the night was:

• Stallamira ED90 Triplet + reducer/ flattener
• ASI 1600MM Pro + EFW + Antlia 3nm 36mm unmounted filters

 Each sub was 600 secs.

The red filter seems very good but, the luminance came out as an overexposed image (completely white). What might be the issue with it?

 

Light_Markarian's Chain_600.0s_Bin1_1600MM_R_gain139_20230417-221425_-10.0C_0001.fit Light_Markarian's Chain_600.0s_Bin1_1600MM_L_gain139_20230417-215911_-10.0C_0002.fit Light_Markarian's Chain_600.0s_Bin1_1600MM_L_gain139_20230417-214857_-10.0C_0001.fit

[The Lazy Astronomer]

30 minutes ago, AstroGS said:

On Monday, I had the chance to take a few subs before the clouds appeared, with the new Antlia 3nm 36mm unmounted filters.

I am planning to shoot Markarian's Chain in LRGB but, I wanted to make sure that everything works as it should. I managed to get a couple of subs with the Red filter and Luminance.

The rig for the night was:

• Stallamira ED90 Triplet + reducer/ flattener
• ASI 1600MM Pro + EFW + Antlia 3nm 36mm unmounted filters

 Each sub was 600 secs.

The red filter seems very good but, the luminance came out as an overexposed image (completely white). What might be the issue with it?

 

Light_Markarian's Chain_600.0s_Bin1_1600MM_R_gain139_20230417-221425_-10.0C_0001.fit 31.27 MB · 2 downloads Light_Markarian's Chain_600.0s_Bin1_1600MM_L_gain139_20230417-215911_-10.0C_0002.fit 31.27 MB · 2 downloads Light_Markarian's Chain_600.0s_Bin1_1600MM_L_gain139_20230417-214857_-10.0C_0001.fit 31.27 MB · 2 downloads

It looks it might simply be, as you say, overexposed. The 16-bit mean ADU of your red image is ~14K (which I would say is quite high already), and obviously a luminance filter will let more light through, so you'd expect a higher ADU for a given exposure time compared to the red filter. It seems as though you've just exposed too long for your conditions.

Side note: 600s is quite a long exposure time for broadband with a CMOS - you've got some elongation of the stars across the frame, which may be due to tracking/guiding errors across the exposure time. That said, the stars near the bottom right corner look pretty round, so it may just be a spacing or tilt issue.

[symmetal]

Yes, your exposures are much too long for CMOS cameras using broadband filters, particularly luminance. You only need to expose for long enough where the noise from the sky background is significantly higher than the read noise of the camera, ( a factor of 5 higher is considered a good figure to use). I've made spreadsheets for various cameras detailing this and this is the one for the ASI1600.

At unity gain 139 (as you've used) and the standard offset of 50, a sky background of 1965 ADU is all that's needed to swamp the read noise by a factor of 5. Inputting your 600s exposure and 14000 sky background ADU into the calculation below the chart, your optimum exposure for Red is only around 54 seconds. You must have significant light pollution where you are, which keeps your optimum exposure durations fairly short. As luminance lets through around 3 times the light of red, (probably higher as the camera will be more sensitive to green, your optimum luminance exposure would be only around 15s.

This would mean taking loads of exposures to get your wanted integration time so to save having hundreds of subs, compromise by trying 2 min each of RGB and 1 min for luminance. That should get you better looking subs.

Alan

[AstroGS]

5 hours ago, symmetal said:

Yes, your exposures are much too long for CMOS cameras using broadband filters, particularly luminance. You only need to expose for long enough where the noise from the sky background is significantly higher than the read noise of the camera, ( a factor of 5 higher is considered a good figure to use). I've made spreadsheets for various cameras detailing this and this is the one for the ASI1600.

At unity gain 139 (as you've used) and the standard offset of 50, a sky background of 1965 ADU is all that's needed to swamp the read noise by a factor of 5. Inputting your 600s exposure and 14000 sky background ADU into the calculation below the chart, your optimum exposure for Red is only around 54 seconds. You must have significant light pollution where you are, which keeps your optimum exposure durations fairly short. As luminance lets through around 3 times the light of red, (probably higher as the camera will be more sensitive to green, your optimum luminance exposure would be only around 15s.

This would mean taking loads of exposures to get your wanted integration time so to save having hundreds of subs, compromise by trying 2 min each of RGB and 1 min for luminance. That should get you better looking subs.

Alan

That’s very interesting indeed. I never thought of this like this and it feels that I always tried long exposures 600secs or 300secs just because…..no apparent reason. I had of course some really nice results so far, either with the ASI2600MC + dual band filters or L-Pro and the ASI1600 but, with ZWO 7nm filters.

 

Tonight is supposed to be have clear skies so, I will give a go with different exp times to check 1 min, 2 and 3 mins. I’m just curious now.

@TheLazyAstronomer you are right, I did not bother that much to perfect tracking that night, most of the time the exposures were unguided as the OAG was losing the star. I wanted mainly to check the filters and camera. But, i’m onboard with your comments as well re the exp time.

 

 

 

Edited April 19, 2023 by AstroGS

[symmetal]

7 hours ago, AstroGS said:

That’s very interesting indeed. I never thought of this like this and it feels that I always tried long exposures 600secs or 300secs just because…..no apparent reason. I had of course some really nice results so far, either with the ASI2600MC + dual band filters or L-Pro and the ASI1600 but, with ZWO 7nm filters.

For your narrowband or dual band filters your 300 or 600 sec exposures are fine, as those filters let through much less light than broadband filters and long exposures are needed. There's no chance of overexposing with narrowband filters.  The L-Pro would likely be better with around 2 mins or so in your skies.  🙂

In your capture program just hover the mouse over a patch of background sky in the image preview and see what the ADU reading is. The median ADU value from your image statistics will likely have a similar value if there isn't a lot of nebulocity around. If this ADU figure is between 2000 and 3000 then your exposure is fine. Exposing for significantly longer than this will just blow out your highlights more and reduce the image dynamic range.

Alan

[AstroGS]

1 hour ago, symmetal said:

For your narrowband or dual band filters your 300 or 600 sec exposures are fine, as those filters let through much less light than broadband filters and long exposures are needed. There's no chance of overexposing with narrowband filters.  The L-Pro would likely be better with around 2 mins or so in your skies.  🙂

In your capture program just hover the mouse over a patch of background sky in the image preview and see what the ADU reading is. The median ADU value from your image statistics will likely have a similar value if there isn't a lot of nebulocity around. If this ADU figure is between 2000 and 3000 then your exposure is fine. Exposing for significantly longer than this will just blow out your highlights more and reduce the image dynamic range.

Alan

It is the 1st time that I am trying to image a galaxy via LRGB. Last time I used the ASI1600MM + ZWO filters I was using an exposure time of 300 secs but, then again I was shooting the Rosette Nebula - even under almost full moon but, then again that was in SHO. I recently also shot the Bode's galaxy via an L-Pro and the ASI2600MC at 300 secs and it came out really nice. 

But, you are right when imaging on broadband the criteria change - I will try 1, 2 and 3 mins on Lum tonight + 3 mins on RGB. Unless, you suggest that even 3 mins for RGB is too much.

[symmetal]

11 minutes ago, AstroGS said:

But, you are right when imaging on broadband the criteria change - I will try 1, 2 and 3 mins on Lum tonight + 3 mins on RGB. Unless, you suggest that even 3 mins for RGB is too much.

Here's the histogram of your 10 min Red image. The top one is the full black to white range.  The signal starts well away from the left (black) edge of the histogram so you're throwing away about 20% of the dynamic range capability of the camera. You want the histogram peak corresponding to the sky background close to the left edge with just a bit of white space below it just to ensure the black signal isn't being clipped.

The ADU measurement method in the previous post are just a more accurate way of getting the optimum exposures. See what your 3 min RGB exposures histograms look like. The gap before the histogram starts will likely be about a third the size of the image above which will be better but still larger than optimum. It's a trade off between optimum exposures and how many extra subs you're willing to deal with. Your luminance will be the most critical on exposure, so see what the ADU values and histograms look like for your 1, 2 and 3 min tests. I suspect that even the 1 min exposures will still be longer than optimum but will likely be the best compromise.

Alan

[Clarkey]

On 18/04/2023 at 22:44, AstroGS said:

The rig for the night was:

• Stallamira ED90 Triplet + reducer/ flattener
• ASI 1600MM Pro + EFW + Antlia 3nm 36mm unmounted filters

 Each sub was 600 secs.

I would agree with everything that has been said. I have the same combination (with the reducer / flattener at F4.8) and I typically use 2 or 3 minute subs in Bortle 6.

Well worth watching this video by Robin Glover if you haven't already:

 

[AstroGS]

I am happy that I followed your advice as it did make sense and I believe that the outcome was much better - except from tracking that it was still averaging above 0.7" most of the time due to wind. Thank you 🙂

I shot 120 sec luminance and 180 sec RGB.

Sharing some subs as an example.

Light_Markarian's Chain_180.0s_Bin1_1600MM_B_gain139_20230420-040356_-10.0C_0029.fit Light_Markarian's Chain_180.0s_Bin1_1600MM_G_gain139_20230420-022413_-10.0C_0030.fit Light_Markarian's Chain_180.0s_Bin1_1600MM_G_gain139_20230420-005037_-10.0C_0002.fit Light_Markarian's Chain_120.0s_Bin1_1600MM_L_gain139_20230419-225102_-10.0C_0030.fit

[symmetal]

4 hours ago, AstroGS said:

I am happy that I followed your advice as it did make sense and I believe that the outcome was much better - except from tracking that it was still averaging above 0.7" most of the time due to wind. Thank you 🙂

I shot 120 sec luminance and 180 sec RGB.

Yes, the wind was annoying last night. I had to give up in the end and discarded all the subs. 😟

Your latest postings are a little difficult to interpret as the sky brightness seems to vary quite a bit between exposures as they're pointing in different directions due to the time difference, and the light pollution is probably changing. I've posted the results after loading into SGP and on the preview I put the black point at 0 and brought the white point down to around 30000 ADU so a direct comparison can be made.

The Green image was the best as far as exposure goes. It was taken at 01:00 so the sky was likely at its darkest. The sky background is still at around 7000 ADU though.

The Blue image has a much higher sky background at around 13700, much the same as your 600s Red from the previous postings. It's taken at 04:00 so not at the darkest hour, and it's possible the scattering from the light pollution has produced a higher blue sky background like sunlight scatter gives a blue sky. 

The Luminance taken at 22:51 has a similar sky backround to the Blue at 14300 ADU and is just after astro dark has begun so not as dark a sky as the Green. It's not peak white like your previous Luminance post was though. 😊 

You posted two green rather than a red so can't assess the red this time.

In conclusion I'd recommend you use even shorter exposures, around 1 min for L, and 2 mins for RG and B. They'll still be overexposed compared to optimum exposures as calculated where 2000 ADU background is preferred, but if you can cope with having all the extra subs to work with it should be beneficial. If you use dither you can set to dither, say every 8 frames for L and every 4 frames for RG and B to avoid wasting too much time having the mount settle after a dither.

With your skies, the narrowband imaging you normally do is the best choice, but at this time of year there aren't many NB targets so you're stuck with wideband for galaxies. With your skies, a OSC camera would have likely been better for you for wideband, as you would have less subs to deal with, but as you have the mono camera and filters you may as well make use of them.

Some may suggest using 0 gain on the camera to enable longer exposures, but I'm not a fan of exposing below unity gain as all you're doing is throwing away photons. At unity gain 1 photon gives 1 ADU in the camera (at 12 bit). At gain zero, 5 photons gives 1 ADU in the camera so you're just discarding 4 photons out of 5 in reality, and you'd need to expose for five times long to get the same signal as you get at unity gain. The increase in well depth at 0 gain is not so significant in practice.

Hope this helps. 🙂

Alan

 

Edited April 20, 2023 by symmetal

[AstroGS]

Hey Alan - thank you for your advice.

You see I would usually use the asi2600mc during Galaxy season but, I wanted to try an LRGB alternative. From my skies I will go back to the OSC with the L-Pro.
 

 

[symmetal]

18 hours ago, AstroGS said:

Hey Alan - thank you for your advice.

You see I would usually use the asi2600mc during Galaxy season but, I wanted to try an LRGB alternative. From my skies I will go back to the OSC with the L-Pro.

As you have a OSC, yes, that would be the best bet for you to use along with the L-Pro. Much less subs and flats to deal with. In the past mono LRGB gave better results than OSC, but the more recent cameras have narrowed that gap considerably.

Your ASI2600MC is a newer camera compared to the ASI1600MM so is even more efficient, leading to shorter exposures as optimum, which isn't really what you need, so if 2 min exposures are still badly overexposed you can use the 2600 at 0 gain rather than the usual gain 100, and 3 min exposures or so should then give good results. You lose the lower read noise feature of HCG below gain 100 but with your light pollution that won't matter, as your exposures will still swamp the higher read noise at zero gain. Even at gain zero the ASI2600 is above unity gain, so the discarding photon issue of the ASI1600 doesn't apply. 🙂

Here's the exposure chart for the ASI2600 (MC or MM) which gives you the optimun ADU values so you can judge your results. Assuming you're using the default offset of 50.

 

Alan

[AstroGS]

And here is the final image - a bit brighter that I like but, I am happy. In general the fact that the Antlia filters provide such a good control over the stars is impressive.

 

Edited April 21, 2023 by AstroGS

[symmetal]

19 hours ago, AstroGS said:

And here is the final image

That's most impressive George. Is that with the 2 min L and 3 min RGB subs? Be interesting to see what the ASI6200MC and L-Pro does as a comparison, to see if the LRGB looks better. 

Alan

[AstroGS]

Yes indeed that is the 2 mins alum and 3 mins RGB. When I get the chance I will use the 2600mc with the L-PRo and see what I can get out of that. Thanks again for the advice - it did help indeed.