Thoughts on which imaging rigs to concentrate on

[Filroden]

3 minutes ago, Gina said:

Hmm... rather a lot of failures in the StarAlignment process - lost nearly half of them   32 left.

That doesn't sound right. I've found it can be confused by the difference in star sizes between subs of different exposures so it may be easier to register them to their own sets then, once integrated, register the final images to each other?

[Filroden]

22 minutes ago, Gina said:

Screenshot of PI auto-stretch and full resolution image saved in PI as PNG.

There seems to be a dashed satellite trail running through the centre of the image, passing close to the Horsehead. What pixel rejection did you use?

It's amazing that even in the unstretched image you can just make out the Horsehead!

[Gina]

5 minutes ago, Filroden said:

That doesn't sound right. I've found it can be confused by the difference in star sizes between subs of different exposures so it may be easier to register them to their own sets then, once integrated, register the final images to each other?

That's exactly what I did (at least I think so) registered to the same exposure sub.

[Filroden]

1 minute ago, Gina said:

That's exactly what I did (at least I think so) registered to the same exposure sub.

Even odder. I know you can relax some of the settings in some of the lower drop downs but I've never tried so I don't know what to suggest. When I ran into this problem (trying to register blue subs to Lum subs) I ended up using the manual DynamicAlignment process as I only had 10 subs to align.

[Gina]

Here's the integration of 32 60s subs.

[Gina]

I think I need to separate M42/43 from the Flame and HorseHead area.  A longer focal length would be better.  Must try the 200mm f4 Super Takumar.  M42 will want the Esprit 80ED telescope.  I've imaged M42 and companions before but not with such a high resolution camera.  4656×3520 px versus 2750 x 2200 px.

Edited January 21, 2017 by Gina

[Gina]

Here's the 430 x 10s Ha subs with curves applied.  Needs DBE applied to smooth the background levels which I think would improve things considerably.  Seems to me that using a vast number of short subs really helps with this high dynamic range target.  I think I shall have another go at this with the same lens but framed to put M42 further from the edge.  I also want to capture OIII subs.  Advantage of this lens is the wide f2.5 aperture.

[Gina]

I've registered the Ha 3 and 5 nm integrated images and applied PixelMath Ha5-Ha3 directly ie. no offset or gain modifiers.

Here's a screenshot of the result auto-stretched in PI.  I haven't bothered to play with the stars.  The Rosette which is supposed to contain NII is looking quite bright with NII but whether some of the other objects contain NII I don't really know.

[Gina]

Auto stretched Ha3 as comparison.

[Gina]

Been playing with PixelMath both multipliers and offsets and can't seem to find a better match than 1:1 though Ha5-(Ha3-0.02) is pretty close I think.  Here's a screenshot from PI auto-stretched.

[Gina]

Comparing the 1:1 and with offset I still think the 1:1 is best.  The two sessions were taken on the same night one following the other and conditions seemed constant AFAICT.

[Gina]

I think the next thing may be to assign the two images to different colours and see what we get.  I have the OIII as well so could do a 3 colour image.

[Gina]

Registered OIII with Ha3, Ha5 and Ha5-3.  So the three I'll use are Ha5-3, Ha3 and OIII or maybe Ha3 Ha5-3 OIII to RGB ie. mapping NII to green.  Probably could do with a more complex PixelMath to avoid too much green.

Any suggestion for the mapping very welcome

[Filroden]

There's two alternatives suggested here http://www.almadenobservatory.net/Color_Imagery.html

 

[Gina]

Very interesting - I'll try them all - thanks Ken

[Gina]

Worked through the Light Vortex tutorials for this and ready to apply PixelMath.  Didn't bother with LinearFit - I'll correct in PixelMath.

[Gina]

One example...  Screenshot from PI auto-stretch.  Plus same image with curves applied.

Edited January 22, 2017 by Gina

[Gina]

SCNR applied then curves.

[Gina]

Full resolution image - Orion 28mm lens HON palette histogram stretched - all processing in PixInsight and saved in PNG format.

[Gina]

Now processing the other widefield area with 28mm lens.  Rosette to Eagle Nebulae.  Here's Ha3 and difference images.

 

 

 

 

 

[Filroden]

There looks to be much more structural differences between those two images, suggesting more N is actually being shown rather than just H.

[Filroden]

Just also thought. The images must be quite balanced in intensity as amp glow has been completely removed in the difference image. 

[Gina]

Colour image with much fiddling in PixelMath   Curves applied too.

[Gina]

There's NII in the Rosette and in the Eagle by the look of it.  First NII Ive seen in the Eagle only ever done Ha, OIII and SII before.

[Gina]

Full resolution image saved in PNG.