0.00
#...
·
 |
|---|
|
hello everybody, I wanted to post here to see what I could improve on my workflow and in general in my images. right here we have my most recent shot of the iris and ghost nebula,(https://astrob.in/0bebw0/0/) which I captured with the redact 51, optolong UV/IR cut filter and the asi 183mc pro. workflow: WBPP removing light pollution in APP StarXTerminator Curves photoshop boosting dark dust (using a screen layer) photoshop camera raw filter( play around withexposure, highlights, contrast etc.) a bit of noise reduction in photoshop BlurXTerminator Boost colors of the Stars Add stars back into the image using pixelmath final touches in camera raw filter add signature constructive criticism greatly appreciated  |
1.51
#...
·
·
1
like
|
|---|
|
Hello, this is a really nice picture. Here my bullet points: I would give you a top pick nomination If i zoom in maybe the noise is not perfect but absolutely not bad The colors are nice A defect could be the stars color, but this is not really you, it's the gear i guess The framing is nice, maybe i would cut some area on the right to have a bit more balance between the iris and the ghost It's really a nice picture and you have a good technique. Keep it up. |
4.65
#...
·
·
1
like
|
|---|
|
Overall I like it. All the dust looks really nice. There is some noise, but not enough to take away from the picture in my opinion. I would have tried to stretch it a bit differently to blow out less of the center of the iris. The main thing in my opinion that takes away from the picture is the stars. Particularly the blue ones. Some are kind of magenta, and especially noticeable on the right side of the image, the blue ones have a strong blue halo on half the star. I too think the framing could have been slightly different. I wouldn't have cropped it, but moved the entire FOV slightly left as seen, but that's a personal choice and minor detail. Then comparing it to your other photographs, I would say in my opinion it's the best one yet. It doesn't have strangely oversaturated stars with artificial diffraction spikes. The background is pleasing. It's fairly natural and nice to look at. |
8.59
#...
·
|
|---|
|
Hi Claudio. At first glance, your image is great, and covers an impressive field. Color is good. Stars are pretty good. At full scale, however, I think you could improve things still. Key point #1 is noise. You have two notably distinct noise profiles. The background sky noise (the noise in the fainter details), and a starkly different noise profile in the stronger signal areas. The SUDDEN shift from noisy to not noisy is jarring and unnatural, and I think the key area where you could improve your image quality. Secondarily, I am not sure how you are doing noise reduction, but however it is being done, it is creating a "rice" artifact. Your noise appears to have these short little streaks of dark scattered about. That created an unnatural noise profile. This is a characteristic of noise that, IMHO, DOES matter...as humans are great at identifying patterns, even small ones. This is why a normal (gaussian) or poisson distribution of noise is so pleasing...its structureless and well distributed. Even a uniform noise distribution, would be preferable (would only occur if you generated noise, which can sometimes be useful.) SO! First things first, I would try NoiseXterminator. You mentioned WBPP, so you have PixInsight. I would use PI for the entire workflow, and the workflow really shouldn't need to be all that complicated. I would do gradient extraction with PI's new gradient extraction process. Then NXT, then BXT, then stretch and process color to taste. NXT should give you a much more pleasing and normal noise profile across the entire image, without the sudden and jarring jump from noisy to near-noiseless in the strong signal areas. |
8.78
#...
·
|
|---|
| I think what stands out the most to me is what looks like horrible chromatic aberration in the stars (I know the RC51 isn't top shelf, but I didn't think this is normal), and the low level noise that is distributed throughout the image. I think you could relax the lower signal areas, you don't have enough time on them to really make it pop. Relaxing the stretch will also create more natural contrast and appeal to the image, while allowing you to decrease the overall noise. Like Jon said, while it's not my favorite, I would give NoiseXterminator a shot. DeepSNR is a far superior and free alternative, but it requires very good data and integration time to work well, I feel you might not be there yet, particularly because this is OSC. Lastly, the core of Iris is pretty blown. You might wanna try some creative masking and see if you can get that under control. |
|
0.00
#...
·
|
|---|
|
thanks for all the advice, you guys pointed out some things I didn't even realise I had in my image ill try to improve on those things especially the integration time. on this target I tried to pull dust a bit too much, that's why the noise I guess. ill be getting a new camera aswell soon, the asi 294mm pro with 3 nm shot and lrgb chroma filters, so I should be getting good data.
|
1.20
#...
·
·
1
like
|
|---|
| Looks nice. One thing I was a bit surprised at is the star quality (and I think this was mentioned stars before). My experience with BlurXterminator, which you are using, is that it usually results in nicer stars than I see in your corners. It's not clear to me if you are actually using BlurXterminator on a starless image or not, but in any case, I thought that should be used when the data are linear. Maybe try using that before StarXternminator -- that should improve the stars I think. |
|
0.00
#...
·
|
|---|
Richard Carande: I usually run BXT when im alredy past all the color boosting, because from all the noise reduction my image looks a bit too soft for my taste, so i sharpen nonstellar aswell with the stars, but i could try using BXT on linear next time. |
13.12
#...
·
·
1
like
|
|---|
|
Hi, In general a very nice image. A few chromatisme on the fat stars. Easy to correct. Noise is ok for me. I prefer this than over denoised images as I see too much. The very issue for me is the data clipped in the center of the nebula. This would prevent me from promoting it. A better control of non linear streching, or the use of masks would help. cheers. You are doing well. JF |
