1.51
#...
·
 |
|---|
|
I just started a new project NGC 2170. Skywatcher Esprit 100ed, QHY 268c, 300s per image. Every image showed lines. I first thought that it is my camera or something in my system. So shot a dark at 300s, no lines. Then in 2s pictures I could see small moving points. I quickly stacked about 1h15min of data to see the effects. As you can see from the image, PI cannot correct these moving objects. I understand that the project will need much more data, but all data collected up to now are waste. Am I correct that these are satellites? Did I miss here something about a launch of satellites? Not stopping, 2h later! PI - Reject  Starless - just 1h15min data...  CS Robert |
0.00
#...
·
|
|---|
|
Most likely satellites (thank you Starlink and others...), because they seem to follow a parallel path. But normally those artifacts should disappear after stacking the frames. Philippe |
|
7.90
#...
·
|
|---|
| Take shorter frames. I mean 300s on an OSC is overkill from where I stand. |
|
1.51
#...
·
|
|---|
Philippe Barraud: Insane... If this is the future, then good night. I am just surprised that they are still present in the middle of the night, where I thought that they are in earth shadow. |
|
1.51
#...
·
·
1
like
|
|---|
andrea tasselli: Good idea. Will give it a try. I had good success using 300s with OSC, especially on dark nebula. But the amount (and effects) of these satellites really caught me on surprise. Rob |
11.41
#...
·
|
|---|
| This is commonplace with any target in Orion. As others have said, these artifacts should go away in stacking. You might need to tighten your sigma clipping parameters a bit. And more, shorter subs (perhaps around 180s) will help with that. I’m shooting this same target now. I have about 20 hours on it so far and there are satellite trails in every single one of the 400 or so subs. I don’t want to ramp up sigma clipping too much, though, because that could affect some of the fine details in this target. So integration time is important. My goal is 35+ hours. The more data they have, the better statistical clipping algorithms are going to work. |
9.82
#...
·
·
1
like
|
|---|
| The data is NOT a waste, collect more. Sigma will do it's thing. An hour is little to no time for a target like this. Keep going, everything will be fine. |
|
7.90
#...
·
|
|---|
| With stright thin lines one option is to extended the frame signficantly in PI, rotate the frame as need to have the trail vertical (or horizontal if that is more convenient) and use the CosmeticCorrection<Defect List> capability to remove the affected line. Make sure you remove the stars beforehand though. |
|
1.51
#...
·
|
|---|
Timothy Martin: Will change to 180s. I was not aware that the Orion region has these issues... I only used WBPP on default settings. So, I need some advice: I assume this is in Image Integration and change the Rejection Algorithm to Winsorised Sigma Clipping? What values would you recommend? I assume higher values will clip more? Could play around, but if you already have some recommended settings, that would be great! +35h... where I live, this would then be a 2-3 year target. I was lucky to get three (3!!!) nights in a row now in December plus favourable moon! I know that NGC 2170 is a challenging target, and a marathon project. I was planning for 10h and then evaluate... |
|
1.51
#...
·
|
|---|
andrea tasselli: Interesting approach! Definitely going to test this... |
|
1.51
#...
·
|
|---|
Timothy Martin: Not to worry - found some references in PI: Sigma low, Sigma highThe  and  parameters, respectively, of the sigma clipping, Winsorized sigma clipping and averaged sigma clipping rejection algorithms, in units of the dispersion (sigma) of each pixel stack. Efforts have been made to compatibilize the response of these three algorithms to both sigma parameters, but you may notice slight differences in performance for the same clipping points. The lower the value of one of these parameters, the more pixels will be rejected. The default values are 4 and 2, respectively.
|
|
11.41
#...
·
·
1
like
|
|---|
Rob Kiefer:Timothy Martin: I can't help you with WBPP. I use APP because of its more advanced local normalization correction. If it's any help, in APP, I use sigma rejection with a kappa low of 6.0 and a kappa high of 2.4 or 2.5 (2.4 on this target). 10 hours could be enough to get a decent shot of this target. FWIW, I would try to avoid after-the-fact techniques to remove these streaks in PixInsight. More subs and good sigma clipping will produce a better result, IMHO. |
0.00
#...
·
|
|---|
|
NGC2170 is aligned with the most heavily used geosynchronous satellite orbits. Every imaging session I have on it has been full of those parallel lines. Like Timothy, I used sigma clipping in Astropixelprocessor, and changed the kappa to 2.5 (default is 3.0). Astropixelprocessor has an option to create a removal map, which shows the pixels that were removed (second image below). I have captured 28 hours on this project so far. I'm still not happy with the noise and lack of detail in the background dust.   |
|
11.41
#...
·
·
1
like
|
|---|
David Fercho: I have the same issue with this region. Have you tried creating a super-luminance master from all the original masters? This was a revelation to me a few weeks ago in my lesson with Ron Brecher. I only have 17 hours on this target so far, but the super-luminance technique is helping with the background. Hopefully, by the time I get to 28 hours, the deed will be done. I don't mean to be pedantic, but what Ron taught me was to use the ImageIntegration process using all the masters, with the original luminance master set as the reference image, to create a new "super luminance" image. On this particular image, it resulted in an increase in SNR by about 7%, but I've seen it go as high as 10% on some images.  Here are all the settings, in case that might be useful to you:     And finally, I did a quick GraXpert, SXT, and NXT on both the original and super luminance images so you can see the difference:   The difference is subtle, but significant. Let me know how this works for you if you give it a try--or if you were already doing this and didn't need my input. |
|
0.00
#...
·
|
|---|
|
Thanks, I really appreciate the help! I’m just learning the basics in PI. For no particular reason I decided to shoot only RGB on this project. So I don’t have any Luminance master to use to follow your workflow. Maybe a big mistake! The last time out I did get a couple hours of Ha, which I see you are using also. Clouds have now settled in and the moon is waxing…so it’ll probably be a while before I have an opportunity to add some luminance. |
|
11.41
#...
·
·
1
like
|
|---|
David Fercho: You can still do this trick with just R, G, and B. Just set the reference to R. That will generate a synthetic luminance for you that might help with those dark, dusty areas. But yeah, I hear you about the moon. Looks like I have to wait a couple of weeks to finish mine, as well. Maybe I can get in some Ha in the meantime. Just FYI, I'm not sure about using Ha on this one. If I do use it, I'll keep it very subdued. I'm finding that it tends to wreck the color contrasts in the angel's gown area because it overwhelms the blue signal. |
