Jump to content

SkySurveyBanner.jpg.21855908fce40597655603b6c9af720d.jpg

Lobster Claw & Bubble Nebula


souls33k3r

Recommended Posts

Hi All,

Here's my latest image. I'm not entirely happy with the results but then again here i am posting this image for comments and criticism and looking for ways to improve.

Ha: 80 x 300s 

SII: 80 x 300s

OII: 80 x 300s

Processed in Pixinsight so kindly let me know what i can do in PI to improve this image because that's all i know :)

Thanks in advance.

Regards,

Image17.thumb.png.b29550943691d04cd359b31b5302038d.png

Edited by souls33k3r
  • Like 13
Link to comment
Share on other sites

What's the camera and scope?

This is a deceivingly difficult target, and you have done very well.  

CC:  I think you have overdone the noise reduction and smooth the nebulosity too much.  I don;t use Pixinsight, so can't help with that.

Carole 

Link to comment
Share on other sites

10 minutes ago, carastro said:

What's the camera and scope?

This is a deceivingly difficult target, and you have done very well.  

CC:  I think you have overdone the noise reduction and smooth the nebulosity too much.  I don;t use Pixinsight, so can't help with that.

Carole 

Cheers Carole, you're too kind.

So true, not an easy target at all. 

Scope was Esprit 100 with ASI1600MM using unity gain. 

I think you may be right, hence the reason why I'm posting the image to figure out what I can do to improve and what I've done wrong. 

Link to comment
Share on other sites

I agree with Carole on the NR. I do NR in Ps because I find the colour select tool makes it easy to pick out just the darker regions for NR. You can use masks in PI to do the same but life's too short!

If you have Ps then Noel's Actions (Now Pro Digital Astronomy Tools) are well worth having and the 'Make stars smaller' action would be good on this, I think.

Good image. The Lobster only works in NB really. I tried it in HaLRGB and gave up on it for lack of anything but red!

Olly

  • Like 1
Link to comment
Share on other sites

55 minutes ago, ollypenrice said:

I agree with Carole on the NR. I do NR in Ps because I find the colour select tool makes it easy to pick out just the darker regions for NR. You can use masks in PI to do the same but life's too short!

If you have Ps then Noel's Actions (Now Pro Digital Astronomy Tools) are well worth having and the 'Make stars smaller' action would be good on this, I think.

Good image. The Lobster only works in NB really. I tried it in HaLRGB and gave up on it for lack of anything but red!

Olly

Olly. Lobsters are red when they come out the pot, all joking aside would you kindly give me a quick rundown on how you colour pick and then do NR in PS.

Alan

Link to comment
Share on other sites

4 minutes ago, alan potts said:

Olly. Lobsters are red when they come out the pot, all joking aside would you kindly give me a quick rundown on how you colour pick and then do NR in PS.

Alan

Sure. Probably the best way is to make a copy layer and work on the bottom one with the top layer invisible. Give the bottom the minimum NR needed to get the worst parts looking OK.

Now activate the top layer and use the Colour Select tool* set to Point Sample**.  Click once on the noisiest part of the image, switch the cursor to the 'add more' function in Colour Select and add more points. You can increase or decrease the 'Fuzziness' value till you've got all the noisiest parts selected. CtrlH will hide the 'ants' so you can see what you're doing. Use a big eraser set to 100% and run it over the whole image in one go. If you now feel the NR is overdone just go to Edit, Fade Eraser, and back it off a bit. Deselect. Next, look at the parts of slightly stronger signal which are still a bit noisy. Select them in the same way and set the eraser to (say) 40% and run it over the image in one. If that's too much Edit-Fade it as before. If it's not enough set the eraser to 10% opacity and erase again.

When satisfied flatten and save. I like this method because you can control your selection very easily and see it in real time. You can also adjust the amount of NR in real time.

When making these decisions I think it best to view the image zoomed out to fit the screen and zoomed in to 100% to make sure it looks OK in both extremes. I will examine noise at way more than 100% to see what's going on but I would never make a processing choice based on pixel peeping.

Olly

* Select in top toolbar.

** Activate eyedropper cursor and the sample options appear in the top toolbar.

  • Like 2
  • Thanks 1
Link to comment
Share on other sites

27 minutes ago, ollypenrice said:

Sure. Probably the best way is to make a copy layer and work on the bottom one with the top layer invisible. Give the bottom the minimum NR needed to get the worst parts looking OK.

Now activate the top layer and use the Colour Select tool* set to Point Sample**.  Click once on the noisiest part of the image, switch the cursor to the 'add more' function in Colour Select and add more points. You can increase or decrease the 'Fuzziness' value till you've got all the noisiest parts selected. CtrlH will hide the 'ants' so you can see what you're doing. Use a big eraser set to 100% and run it over the whole image in one go. If you now feel the NR is overdone just go to Edit, Fade Eraser, and back it off a bit. Deselect. Next, look at the parts of slightly stronger signal which are still a bit noisy. Select them in the same way and set the eraser to (say) 40% and run it over the image in one. If that's too much Edit-Fade it as before. If it's not enough set the eraser to 10% opacity and erase again.

When satisfied flatten and save. I like this method because you can control your selection very easily and see it in real time. You can also adjust the amount of NR in real time.

When making these decisions I think it best to view the image zoomed out to fit the screen and zoomed in to 100% to make sure it looks OK in both extremes. I will examine noise at way more than 100% to see what's going on but I would never make a processing choice based on pixel peeping.

Olly

* Select in top toolbar.

** Activate eyedropper cursor and the sample options appear in the top toolbar.

Thanks Olly I will write this down later as I will lose it otherwise, I understand it and it sounds a good idea, not seen it before. May well be a good one for the OP to try on his fine image of the bubble. Personally I need to get my images a bit better first or they will all disappear. Hopefully weather will be OK tonight for another Cal attempt on PHD, I am beginning to think getting blood from a stone may be easier for me. Thanks again.

Alan

  • Like 1
Link to comment
Share on other sites

2 hours ago, ollypenrice said:

I agree with Carole on the NR. I do NR in Ps because I find the colour select tool makes it easy to pick out just the darker regions for NR. You can use masks in PI to do the same but life's too short!

If you have Ps then Noel's Actions (Now Pro Digital Astronomy Tools) are well worth having and the 'Make stars smaller' action would be good on this, I think.

Good image. The Lobster only works in NB really. I tried it in HaLRGB and gave up on it for lack of anything but red!

Olly

Cheers Olly. Unfortunately my knowledge of PS is super limited so doing anything in PS will take ages to even figure it out. 

  • Like 1
Link to comment
Share on other sites

2 hours ago, ollypenrice said:

Sure. Probably the best way is to make a copy layer and work on the bottom one with the top layer invisible. Give the bottom the minimum NR needed to get the worst parts looking OK.

Now activate the top layer and use the Colour Select tool* set to Point Sample**.  Click once on the noisiest part of the image, switch the cursor to the 'add more' function in Colour Select and add more points. You can increase or decrease the 'Fuzziness' value till you've got all the noisiest parts selected. CtrlH will hide the 'ants' so you can see what you're doing. Use a big eraser set to 100% and run it over the whole image in one go. If you now feel the NR is overdone just go to Edit, Fade Eraser, and back it off a bit. Deselect. Next, look at the parts of slightly stronger signal which are still a bit noisy. Select them in the same way and set the eraser to (say) 40% and run it over the image in one. If that's too much Edit-Fade it as before. If it's not enough set the eraser to 10% opacity and erase again.

When satisfied flatten and save. I like this method because you can control your selection very easily and see it in real time. You can also adjust the amount of NR in real time.

When making these decisions I think it best to view the image zoomed out to fit the screen and zoomed in to 100% to make sure it looks OK in both extremes. I will examine noise at way more than 100% to see what's going on but I would never make a processing choice based on pixel peeping.

Olly

* Select in top toolbar.

** Activate eyedropper cursor and the sample options appear in the top toolbar.

Another fantastic tip - thanks Olly. 

Link to comment
Share on other sites

1 hour ago, souls33k3r said:

Cheers Olly. Unfortunately my knowledge of PS is super limited so doing anything in PS will take ages to even figure it out. 

But Ps is designed to be figure-outable. This is why it has achieved its dominant position amongst graphics programs. In my opinion PI has never made the connection between its algorithms and its users.

Olly

  • Like 2
Link to comment
Share on other sites

4 minutes ago, souls33k3r said:

Olly, I'm sure you're absolutely right but I'm seeking assistance to better my processing skills with this image and using Pixinsight because whatever little I know about processing is all done in Pixinsight :)

Sorry, I thought you'd also mentioned Ps but, on looking back, I see that you hadn't. My bad.

Olly

  • Like 1
Link to comment
Share on other sites

3 hours ago, carastro said:

Much better result Souls, well done.

Carole 

Cheers Carole :) took your advice and ran with it :)

3 hours ago, ollypenrice said:

Far more natural. The higher black point brings a big improvement.

Olly

Cheers Olly. You're right. I just wasn't happy with the first version so started to process the image from the scratch. If it wasn't for your valuable encouragement, I couldn't have done this and for that I'll always be grateful :)

I'm still looking for ways to improve myself. :)

5 hours ago, knobby said:

I like both but the second one looks more subtle and natural looking . No mention of darks / flats ?

Cheers Gary. Yeah I settled on the second version as well. I think I've got 50 flats and equally darks and bias. Yes I've used bias and I don't know why lol

Link to comment
Share on other sites

11 minutes ago, Jkulin said:

I like the second one much better Ahmed, the bubble is just right

Cheers mate. Yeah I like the second one more. I was never happy with the first one so went back and processed the image from the very start. 

Learned a few new things on the way and I'm glad I found the challenge exciting :)

Link to comment
Share on other sites

Noise reduction in PI:

Extract luminance and apply stf as permanent stretch, if you're working on the unstretched image

Make a copy of this luminance, and with curves bring the background up to 50% intensity. Apply this copy inverted to the original image as a mask. This way, noise will be reduced rather than obliterated. And it will only be applied where needed; the dark areas of your image.

Select a preview of the image background and use Image | statistics. Note the standard deviation number.

Open tgvdenoise. Use the extracted luminance as reference, and the measured standard deviation as edge protection value. Lower the exponent of edge protection one unit. Since you're doing luminance noise reduction and not colour noise reduction, set mode to Lab and uncheck chrominance. Reduce L strength to 1.

Increase number of iterations to taste.

Apply tgvdenoise to image.

My preferred method to battle chrominance noise (colour noise and colour mottle) is Multiscale Median Transform applied to Chrominance, because as the name says, it can reduce colour noise over multiple image scales (from pixels to blobs).

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.