NGC 2359 from waste basket to passable

[Rodd]

I decided to let the rig run unattended between 12:00 and 4:30 last night on Thor's Helmet.  It is really the only target in a suitable position to allow me to do this without risking a pier (tripod) collision.  The clouds broke at 12:00 as forecasted (imagine that!).  I would have much rather collected more broadband data as the Moon was not out, but I had to get up for work and could not man the ship as it were.  Too bad.  My focuser temperature compensation worked great the previous night while collected red subs for NGC 1333--but last night I think it had trouble, as FWHM crept up all night.  Originally I was going to delete the subs straight away but then decided to calibrate them and take a look at what I had.  I ended up throwing out 25 5 min subs and keeping only 31.  58 would have made all the difference.  The focus was not very good, though I managed to hide that a bit.  Noise too was high too.  I don't think I will continue with a colored image.  I was looking forward to a widefield shot of this iconic target.     Two versions--normal orientation (I shot it in the first image's orientation)--Most like the second orientation, including me.                                                                      

 

 

 

[wimvb]

I like the view of that helmet floating in space. What filter did you use on this?

[gorann]

I think it looks quite good Rodd, as long as you refrain from pixel peeping (for that you need more data). It is a striking object and I like the wide field so I think it woul be worth while to get more data if you have more nights where you can leave it running after midnight. More data would possibly also bring out more interesting structures in the sky.

[Rodd]

On 08/12/2018 at 04:21, wimvb said:

I like the view of that helmet floating in space. What filter did you use on this?

The 3um Ha.  Conditions were bad though.  See my second attempt below--much better conditions.

 

On 09/12/2018 at 00:10, gorann said:

I think it looks quite good Rodd, as long as you refrain from pixel peeping (for that you need more data). It is a striking object and I like the wide field so I think it woul be worth while to get more data if you have more nights where you can leave it running after midnight. More data would possibly also bring out more interesting structures in the sky.

Thanks, Goran--if you can't pixel peep, then what's the point in having a small pixel camera? ? .  Anyway--I was so discouraged by the results I deleted them.  I would not have used the data in an image anyway--the FWHM values were in the sevens!  Last night I was blessed with a truly decent night.  After finishing up the blue filter for NGC 1333, I switched to Ha and "let it ride" from 12:30 until 6:00 on Thor's Helmet.  I got 46 decent subs.  Here is a quick process. try as I might, however, I can't make the darn thing look decent.  I usually can with 46 Ha subs.  

Anyway--I am pleased with the data.  Maybe I just need a lot more of it.  This is a trend that I, quite frankly, am getting sick of.  I am at 1500 subs and counting for NGC 1333.  And I fear that image will not come out any better than my last attempt with a tenth of the subs.  Again, I am back to my old friend...."I am missing something".  One thing I am sure of....I have come to despise the stars.  I must switch out the spacers to see if the coma like aspect is in fact due to the 1mm spacing issue.  

Just saw it--a bubble structure much like the soap bubble nebula near the Crescent Nebula.  Look to the lower left of the Helmet--at the 8:00 position near a bright star.  A definite bubble.

 

 

[gorann]

I see the bubble! Interesting.

I think your image looks generally quite good. However, the dust looks a bit too smooth to me, almost glossy, and I cannot see any grains in it. Is that due to NR or maybe it is just downscaling to stop me from pixel peeping? The stars are better in the bottom part of the image than the top. You could try playing with spacers but maybe there is a tilt issue. Can tilt be adjusted on the ASI1600? On the ASI071 there are tilt adjustment screws but I cannot remember having seen them on the 1600.

[Rodd]

7 hours ago, gorann said:

I see the bubble! Interesting.

I think your image looks generally quite good. However, the dust looks a bit too smooth to me, almost glossy, and I cannot see any grains in it. Is that due to NR or maybe it is just downscaling to stop me from pixel peeping? The stars are better in the bottom part of the image than the top. You could try playing with spacers but maybe there is a tilt issue. Can tilt be adjusted on the ASI1600? On the ASI071 there are tilt adjustment screws but I cannot remember having seen them on the 1600.

I don't think its tilt--on the TOA without reducer the stars were perfect.  I will try the 1mm spacer first.  If that doesn't work I will try the FSQ 106 native (F5).  That was the actual recommendation from Tak America.  

The smeariness is due to noise control.  I had no choice.  For some reason, I am often plagued with black speckles (very dark pixels).  Even when I collect 800 subs.  But not always--it is target dependent.   I don't understand them, and short of one of Vlaivs treatise on the fundamentals of signal vs noise, Ino one can really explain them in a way that allows me to correct them out.  They do not usually appear in Ha data though.  Mue denoise helps.  In this case I used too much TGV denoise.  

Rodd

[celestron8g8]

Awesome image capture !! Don’t see TH often these days . Well done !! 

[Rodd]

23 minutes ago, celestron8g8 said:

Awesome image capture !! Don’t see TH often these days . Well done !! 

Thanks Ron.  The second data set is growing on me.  The first was pretty...."eh".  I am looking forward to capturing allot more Ha to see if I can improve the depth a bit.  If can then I will add some OIII.  But you are right--this target is in a tough spot.  Very often I will get low clouds near the southern horizon that make shooting this target impractical save on the best of nights.  The saving grace is I only need 2 filters.  But it could easily end up being one that is never finished due to delays and the target slipping away into the mists of spring.

Rodd

R

[wimvb]

2 hours ago, Rodd said:

In this case I used too much TGV denoise

Just as I suspected. Try MLT noise reduction. Follow Warren Keller's recipe (if you have the book).

2 hours ago, Rodd said:

For some reason, I am often plagued with black speckles (very dark pixels).  Even when I collect 800 subs.  But not always--it is target dependent.   I don't understand them, and short of one of Vlaivs treatise on the fundamentals of signal vs noise, Ino one can really explain them in a way that allows me to correct them out. 

There's a plugin called TGVinPaint, that's supposed to clean them. Works a treat in PixInsight v 1.8.4, but it isn't updated for v 1.8.5, afaIk. There's also a pixelmath formula that may work. I'll try to dig it up tonight.

[Rodd]

9 minutes ago, wimvb said:

There's a plugin called TGVinPaint, that's supposed to clean them. Works a treat in PixInsight v 1.8.4, but it isn't updated for v 1.8.5, afaIk. There's also a pixelmath formula that may work. I'll try to dig it up tonight.

I didn't know pixinsight had plug ins.  Is it a script?  I have had trouble getting new scripts to download for some reason.  I do have both versions of PI--the one with the color mask and the previous version.   I know I should update to the newest, but I have stopped getting the messages.  I think it has been over a year so I will have to pay for it. 

Vicent Peris taught us that TGV denoise was better.  At the time I was a neophyte and could do nothing but take his word for it.  The pixel math formula sounds promising.  But I am sure the solution will not be as good as eliminating the problem in the first place.  If its sky related, I am doomed.  But the nagging suspicion that I am missing something won't go away.

Rodd

[wimvb]

2 hours ago, Rodd said:

I didn't know pixinsight had plug ins. 

Sorry, bad phrasing: MODULE is the right word for this, I believe.

2 hours ago, Rodd said:

I think it has been over a year so I will have to pay for it. 

Updates are free until v 2 arrives. In PixInsight, if you go to Resources -> Updates -> Check for Updates, they should become visible.

2 hours ago, Rodd said:

Vicent Peris taught us that TGV denoise was better.

It is, in general. But if it were the single best method in all cases, then why keep the other methods? Sometimes TGVDenoise can leave these "wiggly" step-like artefacts that I see in your image (need to pixel peep). Using a 50% mask avoids this much of the time. But in some cases, I've had better results with MLT.

2 hours ago, Rodd said:

The pixel math formula sounds promising.

The PixelMath expression is something like this:

if a pixel value falls below a certain threshold, replace it with a more appropriate value, based on the median. Otherwise, use the original.

iif($T < 0.003, (median($T)-$T)/3 + $T, $T)

This expression replaces pixels that have a value lower than 0.003 with values that are the original value plus one third of the deviation from median. Be careful, this expression will work on a galaxy in an even background, but maybe not on a target with a lot of nebulosity, or a very rich star field. Also, be prepared to experiment with the values for the threshold (0.003) and the fraction of deviation from median (3 here).

Another option, if the pixels are truely black in the original. Create a defectmap, where all the black pixels of your original are indicated by black (0), and the non black pixels by white (1). Then use the process DefectMap on the original.

To create a defectmap use pixel math:

iif(($T == 0), 0, 1)

 

And finally, a method of my own creation:

• Create the defectmap as just described. Invert it. (Or use: iif(($T == 0), 1, 0) )
• Use Morphological Transformation (MT) to dilate it (Dilation, 3 x 3 structuring element, amount 0.5).
• Apply to image as a mask
• Use MT with default settings EXCEPT for  operator (use Morphological Selection), and amount set to 0.25, on the masked image.

This last method probably works better on a rich star field or an image with nebulosity, than the previous methods.

2 hours ago, Rodd said:

But I am sure the solution will not be as good as eliminating the problem in the first place.

My guess is that it can be caused during image calibration. You could add a tiny pedestal to the calibrated images (it's a setting in ImageCalibration, Output files).

[Rodd]

20 minutes ago, wimvb said:

My guess is that it can be caused during image calibration. You could add a tiny pedestal to the calibrated images (it's a setting in ImageCalibration, Output files).

I looked at my subs coming off the camera and the speckles were there.  

 

21 minutes ago, wimvb said:

Another option, if the pixels are truely black in the original. Create a defectmap

Not sure of this--they look black, but that doesn't mean anything.

 

I will play with the pixel math formula.  Not sure about pedestals.  When I get home I will upload a raw sub and a whole calibreated stack--you will see what I am talking about.

 

Quick question regarding computers.  My laptop is not displaying the contents Flashdrives when I plug them into the USBs.  The flash drives work because I can use them on my desk top.  The computer device manager says the USBs are working correctly.  Any idea what this could be?  I tried restarting....nothing.  Thye were working fine yesterday

[wimvb]

7 minutes ago, Rodd said:

I looked at my subs coming off the camera and the speckles were there. 

Too low offset for your camera, maybe? This can happen at higher gain, where the lower tail of the histogram gets clipped. But with your light pollution, that shouldn't be the case. If the speck would always appear, and at the same location, you'd expect these to be dead pixels. But you wrote that sometimes they don't appear. That rules that idea out.

 

10 minutes ago, Rodd said:

Quick question regarding computers.  My laptop is not displaying the contents Flashdrives when I plug them into the USBs.  The flash drives work because I can use them on my desk top.  The computer device manager says the USBs are working correctly.  Any idea what this could be?  I tried restarting....nothing.  Thye were working fine yesterday

Usb's are funny things. I have a faulty usb on my dslr. When I used it, windows would detect the usb (sound notification), but it wouldn't show up in the file explorer. Eventually it died completely (camera side).

The cause can be a faulty usb port (try connecting another usb device, such as a mouse), or a faulty driver (try removing and reinstalling). You can also try to connect them via a powered usb hub. It might just work. If the computer usb port is ok, it's easiest just to replace the flashdrives.

[Rodd]

23 minutes ago, wimvb said:

f the computer usb port is ok, it's easiest just to replace the flashdrives.

Here's the problem.  The flash drives are working fine on other computers.  My laptop device manager says that the USBs are working properly and that  I have the latest drivers installed.  But I can't access the contents on the flash drive.  Drive E does not appear.  If my USB ports don't work...I am 100% out of commission!  This could be bad.

[wimvb]

That's why you need to check with another device as well.

4 hours ago, Rodd said:

If my USB ports don't work...I am 100% out of commission!  This could be bad.

Let's hope for the best. The data loss you suffered a while ago should be enough trouble for this year.

(The main reason I use a Raspberry Pi based mount computer is that I can't afford to lose a computer "in the field". An RPi is cheap and replacable.)