NGC 7662, The Blue Snowball

[iansmith]

This is my first deep sky image taken of NGC 7662 over several nights last October and November. It was taken using a C11 and Atik 414EX, guided using ONAG and an ASI178MM all on a Mesu 200. Image capture details:

Ha: 15x1200s and 12x300s both at 1x1 binning.

O3: 15x1200s and 12x300s both at 1x1 binning.

Red/Green/Blue: 10x150s at 1x1 binning.

For a first attempt I am quite pleased with the results, but there is room for improvement. The jet at the side of the outer halo is not very clear and my star colours are not great, still having difficulty with that one. Any advice would be much appreciated.

Cheers, Ian.

[John]

I'm not an imager but that really is a stunning image to my eyes Ian

[Davey-T]

Nice close up Ian, do you think you should try binning 2X2 ? I use SCT with Atik 314L, basically the same camera and as I recall unbinned gets you an arc / secs / pixel of less than .5

Best check my maths as it's only from memory 

Are you using a focal reducer ?

Dave

[niallk]

Beautiful! 

[iansmith]

7 hours ago, John said:

I'm not an imager but that really is a stunning image to my eyes Ian

Thanks John.

 

[iansmith]

7 hours ago, Davey-T said:

Nice close up Ian, do you think you should try binning 2X2 ? I use SCT with Atik 314L, basically the same camera and as I recall unbinned gets you an arc / secs / pixel of less than .5

Best check my maths as it's only from memory 

Are you using a focal reducer ?

Dave

Hello Dave,

I don't use a focal reducer so the image is at the C11s prime focus. Your memory is pretty much spot on, the resolution is 0.48" which I know is oversampled for deep sky imaging. I haven't yet tried binning, but I probably should as it would bring down those integration times.

Cheers, Ian

[iansmith]

6 hours ago, niallk said:

Beautiful! 

Thanks niallk.

[ollypenrice]

Very well done. We don't see many of these!

For me the background is a bit dark and will probably have clipped some of the outer parts of the stars, making their colour harder to work on. (Star colour is mostly found outside the stellar cores, round the edges of the star. You can then pull it inwards, the easiest way being to ask Noel Carboni to do it via his 'Increase Star Colour' action. )

Olly

PS As Davey says, I doubt that you'd lose any real resolution by binning. You might even gain by reducing sub length and exploiting the best patches of seeing from more but shorter subs. Only experimentation can tell.

[iansmith]

2 hours ago, ollypenrice said:

Very well done. We don't see many of these!

Thanks ollypenrice. I hope you like them 'cause I have built my imaging rig to image planetary nebula so there will be plenty more to come

2 hours ago, ollypenrice said:

For me the background is a bit dark and will probably have clipped some of the outer parts of the stars, making their colour harder to work on. (Star colour is mostly found outside the stellar cores, round the edges of the star. You can then pull it inwards, the easiest way being to ask Noel Carboni to do it via his 'Increase Star Colour' action. )

I had quite a few problems with this part of the processing. On one monitor the image looked great, until I displayed it on another monitor. Then all the noise showed up horribly well! So I eventually set the black point to reduce the noise on the brightest monitor. I don't know if there is an easy way around this problem.

2 hours ago, ollypenrice said:

PS As Davey says, I doubt that you'd lose any real resolution by binning. You might even gain by reducing sub length and exploiting the best patches of seeing from more but shorter subs. Only experimentation can tell.

I think you're probably right, but it will be a while before I get a chance to experiment. I have another planetary nebula ready for processing and yet a third one I'm collecting data for. Am I right in thinking that if I bin to 2x2 then (all other things being equal) my integration time could be reduced by a 1/4?

Cheers, Ian

[ollypenrice]

34 minutes ago, iansmith said:

Thanks ollypenrice. 

I had quite a few problems with this part of the processing. On one monitor the image looked great, until I displayed it on another monitor. Then all the noise showed up horribly well! So I eventually set the black point to reduce the noise on the brightest monitor. I don't know if there is an easy way around this problem.

 

I think there is an easier way and it's almost the opposite of the one you used. I would strongly resist the temptation to use black clipping as a way of removing noise or gradients.

When stretching I carry on until my background reaches about 25/25/25 in RGB, as measured by the colour sampler tool in 3X3 average. (This removes any subjective impression given by a particular screen. I will later clip them back to 23/23/23 though this often 'just happens' as processing nibbles into the black point, as it does.) If the brightet parts of the background sky are around 25 then 'noise' consists of pixels darker than 25. Instead of clipping them out, lighten them!

Here is a hard stretched luminance stack zoomed in to pixel scale on the background. The lighter parts of the background are at about 25, where I would like them to be at this stage. The unwanted noise takes the form of all those nasty dark pixels.

So now we pin the curve above the desired 25 in many places, to be sure it won't change above that. Then we put a point on the curve well below 25 (5 in this case) and lift it to output at 11 rather than 25. This just makes all the dark pixel noise a little lighter and, therefore, less noisy.

If we attack the problem by bringing in the black point we will initially make it worse  by making the dark pixels darker still and finally only get an artificially smooth sky by clipping it out altogether.

(If there are dark dusty patches in an image they may genuinely be below the background sky value and this approach won't work, or would need to be used as a layer with the real dark regions excluded.)

The numbers 5 and 11 are just examples, not gospel. If the effect is too much when zoomed back out to 100% then we can use Edit - Fade Curves.

I know that stretching noise  is counter intuitive  but I started doing this a while back and think it gives excellent NR on occasion.

Olly

[iansmith]

Hello ollypenrice,

Thanks for the tip, I've tried out your suggestion, with the results below shown. On a brighter monitor the small jet at the 3 o'clock position is more obvious, as is a small bit of nebulosity at 12 o'clock. Of course now I have to master PixInsight's noise reduction processes to get a smoother finish

Cheers, Ian

 

[PatrickGilliland]

Nice to see close up, great work.  A hard one to process and would benefit from more data (despite the fair qty already acquired). In all though its great work to see.

Paddy

[iansmith]

14 hours ago, PatrickGilliland said:

Nice to see close up, great work.  A hard one to process and would benefit from more data (despite the fair qty already acquired). In all though its great work to see.

Paddy

Thanks for the kind words PatrickGilliland. This would definitely benefit from more data, but I was impatient to see how it would turn out, this being my first DSO. I'm also inexperienced with DSO imaging and processing. A lesson learned for project number 4 (number 2 is being processed, number 3 is being captured whenever there's a free clear night). As others have said 2x2 binning would have been a better option.

Cheers, Ian.

[PatrickGilliland]

1 hour ago, iansmith said:

Thanks for the kind words PatrickGilliland. This would definitely benefit from more data, but I was inpatient to see how it would turn out, this being my first DSO. I'm also inexperienced with DSO imaging and processing. A lesson learned for project number 4 (number 2 is being processed, number 3 is being captured whenever there's a free clear night). As others have said 2x2 binning would have been a better option.

Cheers, Ian.

 

HI Ian - best advice I can give is just to learn with each iteration.  100 smaller steps will be more productive and adaptable than 5 big ones.  Data is always key, but you should be happy with the result so far, very promising.

Paddy

[Barry-Wilson]

You've imaged this small PN target really well especially to reveal the inner shell structure.

Excellent work.

[iansmith]

2 hours ago, Barry-Wilson said:

You've imaged this small PN target really well especially to reveal the inner shell structure.

Excellent work.

Thanks Barry-Wilson. I used the 300s exposures for the inner part and the 1200s exposures for the outer halo, then combined them in PixInsight. I think some 1 minute and 10 minute exposures would have made it easier as I could probably have used PixInsight's HDMultiscaleTransform function rather than combine manually.

Cheers, Ian.