C19 - IC5146 - Cocoon, with DSLR.

[MARS1960]

Lovely clear night last night so as soon as C19 came out from behind the tree i was ready.

15 x 900secs lights @ ISO800,

30 x Flats.

cropped and processed in PI and a couple of iterations on star colour using astro tools.

Forgive the added diffraction spikes but it keeps the GF happy .

Hopefully i can add more subs tonight.

Thanks for looking.

[MARS1960]

Just found the missing 9 frames so stacked again and a slight difference in processing.

This is now 24 x 900secs.

I think it's a bit of an improvement but C&C is more than welcome.

[Astrosurf]

I much prefer the second one. Lovely.

Alexxx

[wimvb]

Nice images, both. You may have stretched nr 2 a bit too much, imo. Personally I prefer the first version

[MARS1960]

Thanks guys.

Many images i see have much richer star fields with beautifully colored pinpoint stars, how can i get my stars like that? I mainly use pixinsight but am very new to it.

TIA

[alacant]

1 hour ago, MARS1960 said:

pinpoint stars

300s?

[wimvb]

 

3 hours ago, MARS1960 said:

Thanks guys.

Many images i see have much richer star fields with beautifully colored pinpoint stars, how can i get my stars like that? I mainly use pixinsight but am very new to it.

TIA

It's "just" a matter of having good optics, and getting focus, collimation, tracking, guiding and seeing right. At the same time.

[MARS1960]

26 minutes ago, wimvb said:

 

It's "just" a matter of having good optics, and getting focus, collimation, tracking, guiding and seeing right. At the same time.

Is that all :).

Tracking and guiding is fine, well i think it is, focus i used live view to get vega as small as humanly possible and i use an SW100ED, seeing last night and tonight is very good so is it just that i'm using a low priced doublet? Can i use something in PI to help? Iv'e heard of a simple star mask but are there better options?

I have now got another 4.5 hrs so 9hrs in total to stack and process tomorrow.

TIA.

[wimvb]

To get starsize down I use a combination of the following:

1. Deconvolution (in the linear stage). This one is hard to explain, and even harder to get right. You need to make a good star mask, a luminance mask (= luminance image that is stretched with STF) and a point spread function, which is a model for the average star shape.

Deconvolution is a process to counteract seeing distortion. When done correctly, it will sharpen an image and make stars smaller (tighter). There are several tutorials on the internet, but even then it may take a while to get it right.

2. MorphologyTransformation. Use a contour star mask that is slightly blurred. This mask puts a donut shape on each star that will reveal star edges and protect star cores and anything that is not a star. Then use MorphologicalTransformation with Morphological selection set to 0.2, amount to 0.5, structuring Element to 9 and iterations to 4 - 5. Tihis will dim small stars and make larger stars smaller

3. My first stretch is almost allways a masked stretch. I use default settings, but clipping = 0. Masked stretch will stretch the image in steps (iterations). In the first iteration, the image is stretched ever so slightly without a mask. In the next iterations, the image is stretched but with the previous iteration result applied as a mask (inverted to protect the brightest parts). This will increase the protection on stars, but not on faint detail and background. The result is a very flat image that has detail in faint signal, but without bloated stars.

4. HistogramTransform to increase contrast.

If this is more than you can digest, start with just masked stretch and histogram transform. Once familiar with this, do the 2nd process (Morphological Transformation) before stretching. You will see that just these processes will help tremendously to keep stars under control.

Good luck.

 

[geordie85]

39 minutes ago, MARS1960 said:

Is that all :).

Tracking and guiding is fine, well i think it is, focus i used live view to get vega as small as humanly possible and i use an SW100ED, seeing last night and tonight is very good so is it just that i'm using a low priced doublet? Can i use something in PI to help? Iv'e heard of a simple star mask but are there better options?

I have now got another 4.5 hrs so 9hrs in total to stack and process tomorrow.

TIA.

Do you own a bahtinov mask?

I never bother trying to focus without one

[MARS1960]

1 minute ago, geordie85 said:

Do you own a bahtinov mask?

I never bother trying to focus without one

I do, but personally i find live view at X10 much quicker easier.

[MARS1960]

3 minutes ago, wimvb said:

To get starsize down I use a combination of the following:

1. Deconvolution (in the linear stage). This one is hard to explain, and even harder to get right. You need to make a good star mask, a luminance mask (= luminance image that is stretched with STF) and a point spread function, which is a model for the average star shape.

Deconvolution is a process to counteract seeing distortion. When done correctly, it will sharpen an image and make stars smaller (tighter). There are several tutorials on the internet, but even then it may take a while to get it right.

2. MorphologyTransformation. Use a contour star mask that is slightly blurred. This mask puts a donut shape on each star that will reveal star edges and protect star cores and anything that is not a star. Then use MorphologicalTransformation with Morphological selection set to 0.2, amount to 0.5, structuring Element to 9 and iterations to 4 - 5. Tihis will dim small stars and make larger stars smaller

3. My first stretch is almost allways a masked stretch. I use default settings, but clipping = 0. Masked stretch will stretch the image in steps (iterations). In the first iteration, the image is stretched ever so slightly without a mask. In the next iterations, the image is stretched but with the previous iteration result applied as a mask (inverted to protect the brightest parts). This will increase the protection on stars, but not on faint detail and background. The result is a very flat image that has detail in faint signal, but without bloated stars.

4. HistogramTransform to increase contrast.

If this is more than you can digest, start with just masked stretch and histogram transform. Once familiar with this, do the 2nd process (Morphological Transformation) before stretching. You will see that just these processes will help tremendously to keep stars under control.

Good luck.

 

Thats great, thanks for taking the time Wim. 

I shall look forward to trying this tomorrow.

Many thanks.

Mark.

[geordie85]

1 minute ago, MARS1960 said:

I do, but personally i find live view at X10 much quicker easier.

When I used to use my 600d I'd zoom in X10 set it to the highest ISO and exposure time so I could see the bahtinov spikes very easily.  It made getting perfect focus such a breeze without the need to take any exposures.

[MARS1960]

1 minute ago, geordie85 said:

When I used to use my 600d I'd zoom in X10 set it to the highest ISO and exposure time so I could see the bahtinov spikes very easily.  It made getting perfect focus such a breeze without the need to take any exposures.

It probably is a better option, maybe i should be a little more patient and give it another try.

[wimvb]

5 minutes ago, MARS1960 said:

Thats great, thanks for taking the time Wim. 

I shall look forward to trying this tomorrow.

Many thanks.

Mark.

Just for kicks, when you do a masked stretch on this image, do the following variation:

clone the original before any stretch

apply masked stretch to the original with default settings as per my previous reply

apply masked stretch to the clone with 0.400 as TargetBackground, increase iterations to 200

apply HistogramTransformation to adjust the background (black point) to the same settings for both images.

There will very likely be a very distinct colour difference in the nebula.

 

Have fun!

 

[MARS1960]

1 minute ago, wimvb said:

Just for kicks, when you do a masked stretch on this image, do the following variation:

clone the original before any stretch

apply masked stretch to the original with default settings as per my previous reply

apply masked stretch to the clone with 0.400 as TargetBackground, increase iterations to 200

apply HistogramTransformation to adjust the background (black point) to the same settings for both images.

There will very likely be a very distinct colour difference in the nebula.

 

Have fun!

 

Thanks Wim, i can't wait to try all this out, excellent thank you.