9.48
#... |
|---|
|
I thought the following image was possibly my best from 2020, but after several months, it has not received as many "likes" from my followers as a few others from the same time period. Cave nebula (Sh2-155) I have the following specific questions: 1) Are the star sizes/shapes/halos a major shortcoming? 2) Is the image just too.... red? 3) Should I get a new noise reduction routine? With nearly 15 hours of integration, noise should not be a big issue. 4) I stretched it more than some people would. Is the contrast too high? 5) How is the framing/rotation? Does it bother you? Thanks for any comments you may have! |
10.10
#...
·
1
like
|
|---|
|
5,4,3 are all good in my opinion. As for 2; it is red, but your white balance is good... its just red. I think it is very dramatic. My main point in commenting is I don't have an answer for 1, but I get them sometimes in my images. I like to think the halos are diffusion through the gas clouds, but it might be moisture in the atmosphere (fog?), some dew?. It is actually in my data, so I don't think it is processing. Cheers |
9.15
#...
·
2
likes
|
|---|
|
I like the color, the red is rich and fades into purple where more oxygen emissions mix with hydrogen. The Cave Nebula is a tricky beast . There are some bright portions for which you certainly have enough integration time but then the SNR drops quickly in the transition regions and the image can lack depth. To address your questions specifically: 1) Stars are a bit on the fuzzy side. Not necessarily big but they lack definition and you have lost a lot of color in the star field. Also, many cores are blown out. If you think of the star intensity profile, I attempt as hard as I get not to get a flat top in the stars. If a star is bright and blows up, I want it to be bright everywhere until it disappears in the white. This is a downside of Masked Stretch. It leads to that star shape. I often stretch both with Histogram Transformation and Masked Stretch and then blend the two until I get a pleasing star profile. 2) As I said, I think you nailed the nebula color. There is a reflection component on the lower right side that couldmaybe stand out a little more but those are aesthetic choices. Star color can to be improved. There is a great tutorial by Barry Wilson that I have found very helpful for star color: https://barrywilson.smugmug.com/PixInsight-Tutorials/Boosting-star-colour-Repaired-HSV-Separation 3) The image is a bit on the noisy side. When you look at it at native resolution you see that in the shadows your SNR drops fast and the image loses depth. I think you need more data but a bit of well executed noise reduction can certainly help as well. 4) The nebula is well stretched and the image is not necessarily too contrasty. The downside of how you stretched the image is in the star shapes and color. 5) Framing is good. The composition is balanced with strong diagonal lines across the whole image. I need to put this target on my list for next year as it is a beautiful corner of the sky. Incidentally, I am also in the Capital District - I am in Niskayuna and envious of your darker skies. I mountain bike and am very familiar with the trails in Pittstown State Forest. Cheers, Luca |
|
9.48
#...
·
1
like
|
|---|
|
Thanks very much, David and Luca! I was hoping to do a revision on this one, and you've given me some ideas. It seems that star shapes and halos could use some improvement, and maybe some additional noise reduction is in order. I included some very long (10 minute) luminance subs that produced the stars with flat tops. Since I also have short luminance, RGB, and Ha, it should be possible to replace the brightest stars with star images constructed from the less overexposed subs. The noise in the darker regions is due to clipping the background to reduce light pollution effects. You know how it goes: you can either have an image with low noise and poor contrast, or you can go for strong contrast and worse noise. I will try using Topaz AI only on the dark regions of the image to remove the noise, without affecting stars or details in the bright parts. I know this software is somewhat controversial, but I don't see any harm in using it to smooth out the dimmest parts of an image, where there is little to no useful information in the data anyway. |
