13 replies to this topic

[Linwood]

So got a new mount and first night with the C11 was spending the evening working on guiding, but I had to point somewhere. Due to wind I was limited in direction, and picked the heart of the heart nebula.  Randomly, did not try to frame anything in particular (and really should have gotten more of the upper right side I think), and did narrow band because the moon was near and bright.

 

So I wasn't expecting much, and did not get a lot of data (about 11 each band, 600s subs, 2800mm @ f10, ASI6200MM bin 2x2 gain 100).

 

So with all those caveats of why this is not an impressive photo I could use some advice.  Is the level of detail about what I should expect?  To me the stars are round but a bit bloated and not a lot of detail in the nebulosity.  I'm not sure I know how to judge seeing; PHD2 said the RA osc was about 0.3 or so if I recall. 

 

I did not do anything to reduce star size -- I'm not trying to make this better, it is kind of boring to me, I am trying to understand if optically and guiding wise it is what I should expect.  

 

I integrated to a synthetic luminance and the PI tool says median eccentricity is 0.4158, so like I said, stars seem round.  Just a bit big.  But it's maybe my third NB image, so not sure what to expect.  Also I did not have a lot of experience with the C11, especially at 2800mm, since it rides kind of heavy on my old mount.

 

So... what should I expect, and how does it look in that regard? 

 

Here's a CN size: 

 

 

Here's a full size image (46 megabytes): 

 

https://www.captivep...O/i-ZkrsdTF.jpg

 

Post processing: usual calibration, took top 5 for star alignment master, then aligned all to it, drizzle integrated, blend HSO though I increased O quite a bit (square root for nebulosity), photometric calibration, integrate HSO to synthetic luminance and blend it in (I'm wondering if that contributed to the bloating, using all three channels for luminance).  Color tweaks in photoshop. 

Edited by Linwood, 04 December 2020 - 06:14 PM.

[imtl]

Linwood,

 

1. Is there a reason you are binning 2x2 in acquisition? You bin 2x2 in acquisition (and therefore MIGHT lose details) but then drizzle integrate. Not clear.

2. Stars FWHM is from seeing, tracking errors etc. What is your FWHM according to PI?

3. Your star bloat could come from several sources. Hard to tell and need to delve more into what you got and done. It might be the increased OIII that caused it. It could be your stretching methods.

4. Why are you doing photometric calibration on an SHO image? What is the meaning of that calibration? The tonemap is not real no matter what.

5. You can see in this thread from the last few days, details regarding processing of SHO images and use of synthetic lum. I learned a lot from it and I think you and other will too.

[Linwood]

1. Is there a reason you are binning 2x2 in acquisition? You bin 2x2 in acquisition (and therefore MIGHT lose details) but then drizzle integrate. Not clear.

Well, that's a good question.  I was binning because even 2x2 I'm at 0.54"/pixel, so I didn't see any reason to capture at twice that.  

 

And... well, I've just always been doing the drizzle integration. When I was doing OSC with CFA Drizzle I understood that to be good regardless of over/under sampling (with pixinsight) but I am probably operating on inertia here, and did not need to drizzle.  

 

 

2. Stars FWHM is from seeing, tracking errors etc. What is your FWHM according to PI?

So... where?   On a sub?  On the luminance?  

 

The Luminance gives me a number I don't understand (0.5586 px  *** see below, this was MAD FWHM ***).   I went back to the original subs (I've deleted the intermediate processing ones) and ran subframe selector, and there it ranges from 1.992 to 3.568 with the average at 2.6.  In retrospect of what I just wrote, I'm not sure I understand the units on either of those, I always just look for "small is good" for example when ranking subs to get a reference.

 

4. Why are you doing photometric calibration on an SHO image? What is the meaning of that calibration? The tonemap is not real no matter what.

Well, simply put, because I liked the result. The few NB images I did half the time it seems to help, half the time it makes a mess.  So I'm not thinking it is making the colors accurate, I just liked the results.  It may be more about its final steps that neutralize the background though, I should try just doing that. 

 

Thanks for the thread pointer, reading now. 

Edited by Linwood, 04 December 2020 - 06:55 PM.

[sn2006gy]

I would step back and simplify things a bit.

 

I don't see anything wrong with Binning the 6200 but I wouldn't bin it and then drizzle integrate necessarily. 

 

I'd forget with all the complex stuff and get the basics down.  Shoot 300 second subs, integrate with standard SHO processing/workflow. Report back.

 

Once you're happy with that - then try more advanced things and longer subs.

 

On the positive side, your star shapes look great. They're round Tracking doesn't appear to be your problem!

[Linwood]

Beautiful images in that other thread, and a lot of useful info.  Also going to try the bubble nebula with the C11 soon.

 

I redid my combination without a synthetic luminance. I got a LOT less noise for reasons a bit unclear, but exactly (to eye) the same size stars and detail. 

 

 

I measured the FWHM in each separately. 

 

Oiii: 10.10px  (median FWHM)

Sii: 9.197px

Ha: 9.563

 

So there's not a large difference in them.

[Linwood]

On the positive side, your star shapes look great. They're round Tracking doesn't appear to be your problem!

Does that mean there is a problem? 

 

Again, not trying to doo all the things I would if I wanted this to be pretty (like maybe remove and shrink all the stars).  Just trying to understand if this looks like what to expect, or do I have a problem? 

[imtl]

Well, that's a good question.  I was binning because even 2x2 I'm at 0.54"/pixel, so I didn't see any reason to capture at twice that.  

 

And... well, I've just always been doing the drizzle integration. When I was doing OSC with CFA Drizzle I understood that to be good regardless of over/under sampling (with pixinsight) but I am probably operating on inertia here, and did not need to drizzle.  

 

Well, with that image scale at 2x2 bin you are probably not really losing anything significant if your seeing is average i.e. 2"-3". But as a general case, there is no advantage in CMOS binning in acquisition since its doing software binning anyways. Maybe besides file size. Which you have completely negated when doing drizzle integration.

 

 

Beautiful images in that other thread, and a lot of useful info.  Also going to try the bubble nebula with the C11 soon.

 

I redid my combination without a synthetic luminance. I got a LOT less noise for reasons a bit unclear, but exactly (to eye) the same size stars and detail. 

 

 

I measured the FWHM in each separately. 

 

Oiii: 10.10px  (median FWHM)

Sii: 9.197px

Ha: 9.563

 

So there's not a large difference in them.

then with your image scale of 0.54"/pixel you get roughly 5.1" FWHM. That's a bit large. Could be below average seeing for the night but it might also be your mount cannot handle that focal length very well? Not sure. Don't worry about that for now because I think Byron is quite right with what he wrote. Go simple and basic. And then analyze that and see how to proceed.

 

Byron, The stars might look round just because of the bloat and any problems (if there are any! Don't get stressed Linwood) might be hiding underneath the bloat.

[sn2006gy]

Does that mean there is a problem? 

 

Again, not trying to doo all the things I would if I wanted this to be pretty (like maybe remove and shrink all the stars).  Just trying to understand if this looks like what to expect, or do I have a problem? 

No, no problem. 

 

Actually, I think the stars are rather fine. I would be curious if Bin2 drizzle + 600 seconds may have caused the foreground stars to be more bloated than needed so I would avoid drizzle integration  and I would start with 300 second subs. Your other stars are all nice tight points of light.

 

As for shooting nebula - i think the color scheme is what breaks any texture/resolution - you have detail in there but its almost like its all inverted so we see the stars more than anything else.

 

If you do some of the SHO or HOO integrations that Rista talks about here: https://jonrista.com...-pixelmath-hoo/  I think you will have some amazing very deep nebula shots.

Edited by sn2006gy, 04 December 2020 - 07:07 PM.

[sn2006gy]

Zoomed in on the full size image and the stars definitely have the signature of "blown out" aka over exposed... (or the processing did that - can't tell - everything feels inverted heh)

[Linwood]

then with your image scale of 0.54"/pixel you get roughly 5.1" FWHM. That's a bit large. Could be below average seeing for the night but it might also be your mount cannot handle that focal length very well? Not sure. Don't worry about that for now because I think Byron is quite right with what he wrote. Go simple and basic. And then analyze that and see how to proceed.

Is there a factor of two missing.  I binned then drizzled, so the final image was back to 0.277"/px, which puts me back around 2.7 not 5.4? 

[Linwood]

Zoomed in on the full size image and the stars definitely have the signature of "blown out" aka over exposed... (or the processing did that - can't tell - everything feels inverted heh)

Here's one uncalibrated Ha sub at 1:1 in the center.  

 

 

The brightest star there is 0.66 (by dragging the cursor over it) which is high but not awful.  But 75x the nebula level. 

 

So over exposed?  That's not a problem I usually have, I'd love if 300s worked better.  
 

Mean is 568 against background of about 505 (DN's), AvgDev 13.6.

[Linwood]

As for shooting nebula - i think the color scheme is what breaks any texture/resolution - you have detail in there but its almost like its all inverted so we see the stars more than anything else.

Yeah, I keep playing with colors. I probably should stick with some more standard schemes, but ... playing is fun.  I probably skipped that time in kindergarten with finger paints and am making up. 

[imtl]

Is there a factor of two missing.  I binned then drizzled, so the final image was back to 0.277"/px, which puts me back around 2.7 not 5.4? 

Oh, you did not mention it so I assumed you did the FWHM calculation on 3 calibrated subs. My apologies. Yes, if it is on the x2 drizzle integration then you are at 2.7".

[Bill G.]

With all that said....yes there are saturated stars...but I like it!!!

 

Bill G.