Different stack of yesterday's Wirtanen data

[SteveNickolls]

42 minutes ago, happy-kat said:

Interesting another imager using 30 second subs for a comet, I'm thinking it is the sweet spot for the comet not becoming elongated as they are going so fast

Yes a valuable observation, the streaks wouldn't be there or be a lot shorter!  ?

Good luck if the sky is clear.

Hey presto!

That's using the technique in the video.

Cheers,
Steve

[happy-kat]

That's got good definition, I'll hopefully try a process tomorrow night or Tuesday

[SteveNickolls]

Thanks happy-kat. In the whole process it's the use of the LIFE Module on the comet head adjusting the 'Mask Fuzz' value to a very high level that removes the star streaks left in the background.

Cheers,
Steve

[happy-kat]

Be interesting to see how close to the black point the background gets but the image looks nice and clear stars and comet. I might yet make something better on my data.

[Stub Mandrel]

DSS's comet mode essentially uses sigma kappa to remove the elements that move in each image.

I got MUCH better results from Wirtanen than Catalina, partly because I collected more data but mostly because the comet was moving a lot faster across the image.

The automatic comet and stars stacking in DSS is poor, because it applies no pre-processing before combining the two images.

With Wirtanen the comet-only image was excellent with the remains of the stars being little more noticeable than the faint patterns left by the movement of noise, and largely removable by usual 'background flattening' approaches.

This is the blue layer from the FITS I saved of the comet on;ly stack, the ONLY processing has been the stretch in FITS liberator. No noise removal or any attempt to smooth out the background texture, although I may have applied gradient exterminator (not sure).

This is the final comet only image, after noise removal, gently increasing saturation etc.

The star only image showed a significant 'ghost' of the comet'.

This is the stars-only stack from DSS, this time only the green layer of the FITS, initial stretch but NO gradient exterminator. As the comet is a green/cyan colour the residual 'ghost' of teh comet is more marked on this than in the blue and red layers.

Despite the apparently profound mark, I was able to remove it in two stages: (1) I used magic wand to select it, contracted the mask clear of the stars within the 'ghost', applied a feather edge, then used curves to darken it to a consistent level. (2) a 'hockey stick' curve to remove remaining tone variation in the background. This has also had the star colours enhanced. It left the stars looking a bit hard edged:

Final combination was very few steps:

Take the stars image, apply 'less crunchy, more fuzzy' to slightly soften those stars. Overlay 'stars' on the comet as a layer using 'lighten' mode. Raising the lowest level of the comet layer until a 'hard' boundary around the comet disappeared. Finally a top layer and dropping the black point to get a more natural looking background. (note all processing done on 16 bit tiffs, these are 8-bit PNGs)

 

 

[SteveNickolls]

Thanks Neil for this, very helpful and informative.

Best regards,
Steve

[Stub Mandrel]

Thanks Steve. To complete the comparison, this is a rough stretch of the 'star and comet' dual stack in DSS. The benefit of processing the star and comet images separately is obvious. Note the comet position is set by your choice of reference frame.

[SteveNickolls]

Well here's my final attempt at the recent comet-

Have learnt a lot regarding StarTools and its mask tool and LAYER module from the exercise. Good luck to all those attempting further imaging of the Comet while it is passing through Taurus and Auriga.

Cheers,
Steve

[Stub Mandrel]

Looks good from here! Well done Steve

[happy-kat]

First pass and in order to sort the background it has ended up totally clipped. If you do not succeed the first time try try again.

 

[happy-kat]

Sorry to Michael we have hijacked the thread as continuation from the comet star artifacts from DSS stacking.

I had a second go at processing this today. I think for this to succeed I will need to process the stars and background image much better. I am happy with the comet version as the comet colours sorted themselves out just from a mask on the comet and reducing the red sky glow, bingo the green was there. In front of me is Dark Art or Magic Bullet to see if I may get mileage with part processing the background externally once the file has been split using startools. The back ground here is hovering at 1 or 2 so just n the edge of total clipping. I tried one grow action on my star mask but this can now be seen as a small brown edge to all the stars. I think I also need to create an image as to how I did the comet to layer in the Pleiades as there was nebulousity to reveal.

[SteveNickolls]

On 18/12/2018 at 14:31, happy-kat said:

I think for this to succeed I will need to process the stars and background image much better.

Hi happy-kat, with the image out of DSS that had the stationary stars I made a star mask in StarTools and from the image with stationary comet and moving stars removed the comet but it still had stars visible in that image. I used 'mask' to cover the comet and in COLOR module rendered the stars white. I then added the comet with bleached stars with a 58% overlay of the original coloured stars to get the final image. Hope that may help.

Cheers,
Steve

[happy-kat]

I will do another process and crop more of the left off as that is where the hideous murk and sky glow band was as it it upsetting my colour spread. I can just fit the Pleiades in though I will probably tone the comet down on the next process of this data set. This is as a layer of three images, stars, comet and Pleiades. I have more ideas in my head to use to fix the few issues on colour and the few bright stars in my comet layer that I need to blend out before recombining.

Looks over worked now so need to find that sweet spot.