Trouble with getting colour into my OSC ASI2600mc

[simmo39]

This may be a long and rambling post, sorry.

I have just got hold of a ASI 2600mc. I have for a few years been using an ASI 1600mm pro for narrow without any real problems other than over use of the stretch button in PI. I went for the ASI 2600 as I didnt get much success with RGB filters on my ASI 1600 and thought I would cut out the middleman and go for OSC. I did a quick and dirty test the other night on M 31 30 x 120s at gain of 100 ( it was a quick test as Im suffering with a back problem and sciatica  and couldnt guarantee that I would be able to get my kit back in after a long session ). I have already sorted a darks library and on the night I did flats and dark flats but due to my back I dont think I had the settings right for the lights ( more on that in a bit ).  I did a quick stir of the data in PI and the image that came out was v poor full of banding and v bad gradients. So I had a quick look at my flats and found they looked very dark. I had a quick think and realised when using flats aid in APT i had left the ADU setting at 20000 witch is fine for my mono ASI 1600 but seems no good for the new camera, also the exposure time was only 1s and from my 1600 that always worked better at 2s and above.

I redid the image again in PI without the flats and i got a better result but the image stars lacked the colour I was expecting. Know first question, Have i done a to long exposure or am I missing something in PI?  I have put the image below ignore the blown out core that I think I can sort Im more concerned about the lack of colour in the stars.

 

Next question is about my flats. Is my ADU setting to low? and when I did the flats as per the APT ccd flats aid my LED lightbox was nearly at the off setting to get it to do the 1s exposure normally for 1600 its a lot brighter ( image below of flat )

Sorry for the ramble but thanks in advance for any help and replies.

 

 

[mackiedlm]

With my ASI 2600MC  and APT flats aid I set target ADU to 35000. So your 20000 is a bit low but I don't think it would  make them unusable.

Depending on what I'm using for flats, (sky or my led tracing panel with various layers of white t shirt) I've had flats exposures from 0.5s up to 5s. Everything I've read about the 2600 says actual flats exposure time is not that important.

I think you have plenty colour in there and if you got rid of the green cast it would start to show well.

 

If you wanted to share an untouched stack I'm sure someone would have a look at it for you.

Good luck. It is an amazing camera and I'm sure you'll get it working for you.

[simmo39]

6 minutes ago, mackiedlm said:

With my ASI 2600MC  and APT flats aid I set target ADU to 35000. So your 20000 is a bit low but I don't think it would  make them unusable.

Depending on what I'm using for flats, (sky or my led tracing panel with various layers of white t shirt) I've had flats exposures from 0.5s up to 5s. Everything I've read about the 2600 says actual flats exposure time is not that important.

I think you have plenty colour in there and if you got rid of the green cast it would start to show well.

 

If you wanted to share an untouched stack I'm sure someone would have a look at it for you.

Good luck. It is an amazing camera and I'm sure you'll get it working for you.

Hi and thanks for the reply. I will up the ADU next time. I have had a play with the image and removed the green cast, still looks a bit washed out but there is a little colour there. I dont think this was the best target to do a test on but it was forced on me by the Moon and Xmas lights! lol

[mackiedlm]

23 minutes ago, simmo39 said:

Hi and thanks for the reply. I will up the ADU next time. I have had a play with the image and removed the green cast, still looks a bit washed out but there is a little colour there. I dont think this was the best target to do a test on but it was forced on me by the Moon and Xmas lights! lol

I recon if you do a saturation boost in CurvesTransfomation youll see the colour coming out. How did you rmove the green?

At least you had some clear sky to do a test!

[simmo39]

6 hours ago, mackiedlm said:

I recon if you do a saturation boost in CurvesTransfomation youll see the colour coming out. How did you rmove the green?

At least you had some clear sky to do a test!

I used the SNCR function in PI to get rid of the green, I will have another play with the data later on in the week to see if I can get some colour. I need a good back and some more clear skies to do some more tests. Fingers crossed. 

[tooth_dr]

I did a similar shot on M31, 30 x 120s gain 100, the colour was good so I don’t see anything wrong with your exposure times. For my flats I am for 27000 ADUs, and as said above, actual flats exposure time doesn’t seem to matter. Mine vary from 0.8s to 20s.

Maybe your sky conditions and the moon haven’t helped. 
 

Would you mind posting the stack?

[simmo39]

48 minutes ago, tooth_dr said:

I did a similar shot on M31, 30 x 120s gain 100, the colour was good so I don’t see anything wrong with your exposure times. For my flats I am for 27000 ADUs, and as said above, actual flats exposure time doesn’t seem to matter. Mine vary from 0.8s to 20s.

Maybe your sky conditions and the moon haven’t helped. 
 

Would you mind posting the stack?

Hi, thanks for the reply. When I get chance l will post the stack and you can have ago. 

[simmo39]

Well I think I managed to get some colour out of the data, Still got a blown core and the rest of the image is a bit rough and ready but it shows there is colour in them there hills!

[Ouroboros]

@simmo39 I thought there would be colour there. What did you do differently? 

[simmo39]

12 minutes ago, Ouroboros said:

@simmo39 I thought there would be colour there. What did you do differently? 

Well I tried what this guy did , not to sure on his tecknique but it was quick and easy to follow. I think his data is a lot better than what i have but it did show colour. just need to learn how to control the core better.

 

Edited December 13, 2021 by simmo39

[Ouroboros]

OK. After I’ve got my initial stacked (integrated) image in PI I apply the Screen Transfer Function to stretch the image just to see what I’ve got. It usually has a horrible colour cast, like yours, so I reapply the STF with the chain link icon deselected. This de-couples and stretches each colour channel separately. I can then see what colour I’ve got before I go through my workflow of background normalisation and colour calibration etc. 

Edited December 13, 2021 by Ouroboros

[simmo39]

11 minutes ago, Ouroboros said:

OK. After I’ve got my initial stacked (integrated) image in PI I apply the Screen Transfer Function to stretch the image just to see what I’ve got. It usually has a horrible colour cast, like yours, so I reapply the STF with the chain link icon deselected. This de-couples and stretches each colour channel separately. I can then see what colour I’ve got before I go through my workflow of background normalisation and colour calibration etc. 

I did try something similar before but I still lacked a lot of the colour I have got now. I will do a test on a different target to see what I get. Just need some clear night. 

[Ouroboros]

@simmo39 Oh, yes, same here. I have found photometric colour calibration seems to give pleasing (to my eye) colours. But then I prefer subtle. 😀

[Scooot]

I’ve found the colours are one of the best things about the camera and you shouldn’t have to try hard to bring out the colours. 
EG this is a single uncalibrated 3 minute sub at zero gain I took a couple of nights ago. All I’ve done to it is debayer, copied an STF to histogram transformation and resized it. I haven’t even colour calibrated it. To see it without a colour cast I just unlinked the STF. As you can see there’s plenty of colour in it and I’d guess you have better skies than me in Norfolk.

Maybe try doing a quick process on an uncalibrated sub to see what that’s like.

[Scooot]

To check your exposure, try @vlaivs suggestion, I’ve found with my light polluted skies 3 minutes at zero gain is pretty good, any longer doesn’t get me a better SNR. I screen shot it so I don’t forget how to do it  

[Budgie1]

Just a thought here, so bear with me.

When you stacked the image, I assume you were using the Weighted Batch Processing in PI?

Did you have CFA Image ticked under the Calibration tab and, if so , what settings did you have selected?

I'm just wondering if the image wasn't debayered correctly during stacking.

[simmo39]

11 hours ago, Budgie1 said:

Just a thought here, so bear with me.

When you stacked the image, I assume you were using the Weighted Batch Processing in PI?

Did you have CFA Image ticked under the Calibration tab and, if so , what settings did you have selected?

I'm just wondering if the image wasn't debayered correctly during stacking.

Hi, I had CFA ticked and left to auto. I normally leave all calibration settings to default.

[simmo39]

12 hours ago, Scooot said:

To check your exposure, try @vlaivs suggestion, I’ve found with my light polluted skies 3 minutes at zero gain is pretty good, any longer doesn’t get me a better SNR. I screen shot it so I don’t forget how to do it  

Thanks for that although I cant claim to understand it all, the last bit about short exposures is worth a try but I will have to find a lesson on how to do the replacement of pixel values. Not a great person with pooters! lol

[Scooot]

3 hours ago, simmo39 said:

Thanks for that although I cant claim to understand it all, the last bit about short exposures is worth a try but I will have to find a lesson on how to do the replacement of pixel values. Not a great person with pooters! lol

Calculating the exposure looks worse than it is, you can get the mean values of a sub from the statistics process. I’ll go through my workings later as an example. 

Replacing the pixels from a short sub was actually a bit more complicated than you might think.

[simmo39]

3 minutes ago, Scooot said:

Calculating the exposure looks worse than it is, you can get the mean values of a sub from the statistics process. I’ll go through my workings later as an example. 

Replacing the pixels from a short sub was actually a bit more complicated than you might think.

Thanks that would helpfull. I was thinking if I did M31 again I would go down from 120s to 90s and hopefully have less effect on the core, but I will have a play on another target next time.

i

[Scooot]

3 hours ago, simmo39 said:

Thanks that would helpfull. I was thinking if I did M31 again I would go down from 120s to 90s and hopefully have less effect on the core, but I will have a play on another target next time.

i

For the Asi2600 Mc pro 100 gain is 0.25e and zero gain is 0.75e roughly.

Read noise is approximately 1.5e at 100 gain and 3.6e at zero gain.

I can see using a calibrated debayered sub that the mean background level of the blue channel is 388 ADU. (See below)

388ADU x 0.75gain = 291 electrons

The square route of 291e is 17.05e which is the background noise.

17.05 / 3.6 (read noise at zero gain) = 4.74

4.74 is the ratio of read noise to background noise which is slightly less than 5 so I could afford to increase my sub exposure a little.

To work out by how much.

5 / 4.74 = 1.06

1.06 squared is 1.12.

So I can afford to increase my exposure from 3 minutes by 12% to 3 minutes 20 seconds.

 It does vary by the quality of each night and where in the sky you’re  imaging, when I took this sub it was very high up at around midnight so the sky would have been at its best from that perspective.

I think I’ve done that right  

Edited December 14, 2021 by Scooot

[simmo39]

1 hour ago, Scooot said:

For the Asi2600 Mc pro 100 gain is 0.25e and zero gain is 0.75e roughly.

Read noise is approximately 1.5e at 100 gain and 3.6e at zero gain.

I can see using a calibrated debayered sub that the mean background level of the blue channel is 388 ADU. (See below)

388ADU x 0.75gain = 291 electrons

The square route of 291e is 17.05e which is the background noise.

17.05 / 3.6 (read noise at zero gain) = 4.74

4.74 is the ratio of read noise to background noise which is slightly less than 5 so I could afford to increase my sub exposure a little.

To work out by how much.

5 / 4.74 = 1.06

1.06 squared is 1.12.

So I can afford to increase my exposure from 3 minutes by 12% to 3 minutes 20 seconds.

 It does vary by the quality of each night and where in the sky you’re  imaging, when I took this sub it was very high up at around midnight so the sky would have been at its best from that perspective.

I think I’ve done that right  

Thank you for guide. I will give it a study and try it out when i get some good subs. Its getting all scientific all of a sudden. lol