My my my Corona....

[RickS]

Apologies to the Knack :lol:

This was my first serious attempt at processing the corona using PixInsight (with some help from Photoshop). Now that I sort of know what I'm doing I'll try a deeper stack...

Edit: second image is a repro using a deeper stack. More detail but a little less subtle?

Cheers,
Rick.

[MrB]

Awww, I was gonna say something witty like "You seem to have a Knack for it ;)"

Anyway... nice image!

Edit: oh dear :rolleyes: http://www.youtube.com/watch?v=T15WCcV6Fhg

[ourkind]

Fantastic capture Rick! Very nice detail, well done! :thumbsup:

[naskies]

Very nice!

[iceman]

Beautiful! Can you share any details of what you did in PI or PS?

[Larryp]

That's really nice, Rick!

[RickS]

Thanks, Simon. Love the video :lol:







Thanks, Carlos, Dave & Laurie!

[JB80]

That is most excellent. I love it. :)

[RickS]

Thanks, Mike. I'm going to write up my workflow after I try a few more ideas but I can give a brief explanation of what I've done so far...

I collected a large number of bracketed images using a DSLR on a fixed tripod. I did take a mount with me but it wouldn't track (that'll teach me for not testing it at home!) That means my main problem is registering the images. To get good results, registration needs to be done on the corona itself, not on the moon which is moving wrt to the corona.

Roughly what I did was this:


convert images to TIFF
in small batches, I used the PI FFTRegistration script to register and sum the images (this registers against the moon, but over a short time period that shouldn't smear the corona too much)
I used the PI LarsonSeknina filter to process each of the summed images. I used a few different parameters to pull out detail at different scales and then summed these using PixelMath back into one image again (for each batch)
I used PS to manually register each of the batch images.
I went back to PI and summed all the batch images using PixelMath
Finished up with some sharpening, noise reduction, etc
Back to PS to drop an original copy of the moon into the middle of the image

The image I posted was using 4 x 9 images. I have a lot more than that, so I hope to get better results when I stack everything. I also have some ideas for more accurate stacking but it will probably require scripting.

Cheers,
Rick.

[Paul Haese]

Very nice Rick. good detail all the way out.

[Peter Ward]

Humm..... The best TSE 2012 extended corona I've seen so far.

Nicley done !

[RickS]

Thanks, Jarrod, Paul & Peter.

I'm hoping to pull some more detail out of it yet but we'll see. It's not a bad result for a DSLR (Nikon D800E) and 70-200mm zoom lens.

[Virgs]

Ladies and gentlemen we have a winner, lovely subtle detail and so well processed - no hint of onion rings. My congratulations to you.

[DavidTrap]

That is looking spectacular Rick.

I'll have to catch up with you in the new year and learn more about processing - haven't the foggiest understanding of what you mean by all of that... ;)

DT

[gregbradley]

A fabulous result.

Greg.

[gregbradley]

Which 70-200mm lens - the F2.8 VR11?

Greg

[firstlight]

Stunning Rick. Makes me almost want to take my own efforts down and go home (almost). Good to see you have some great data and I can't wait to see what the final result is.

In my own case I so happy that Anne-Louise's mate came over to our possie and managed to capture the inner corona.

[jjjnettie]

:) Perfection!!

[RickS]

Thanks, Virgs!



Thanks, David. I'd be happy to show you the few tricks I know next time we catch up.



Thanks, Greg. Yes, it's the 70-200 f/2.8 ED. Think it might be the older but optically similar VR.



Thanks, Tony, but definitely keep at it! I bet you'll be surprised what you can pull out of your data.



Thanks, JJJ, but I'm sure there's still room for improvement!

[jjjnettie]

We are our own worst critics. ;)
I understand where you're coming from, but I beg to differ. :)

[RickS]

I did a repro with additional data (after finding a cunning way to register the frames more quickly). It has a bit more detail but it's not as subtle. I can't decide if it is better or not :shrug:

[ozstronomer]

Rick

Sorry I missed this one, I'd be happy with either as both are sensational,really well done:thumbsup:

[multiweb]

Superb blend. Really cool. :thumbsup:

[RickS]

Thanks, Geoff & Marc!

[ChrisM]

Rick - that looks almost surreal - a bit like a cloth with folds. Anyway - a great result from you and your gear. I too would welcome more detail on how you did that in PI.

Chris

[Jen]

Haha rick I'm glad I wasn't the only one singing this every time I looked at eclipse pics :P nice shots by the way :thumbsup:

[RickS]

Thanks, Jen :D



Thanks, Chris.

The only major change I made to my initial processing (which is described in an earlier post) for the second version of the image was that instead of manually registering my summed images in Photoshop, I calculated the X and Y shift between the first and last image by registering them manually in Photoshop and counting pixels. Then, using the timestamp information from the EXIF data I interpolated the delta X and Y for each intermediate image and did a transformation with PixelMath. That allowed me to register the images more quickly so i was able to include all of my data.

I have a couple more ideas to try. Once I'm finished playing I will try to write up my workflow in a bit more detail.

Cheers,
Rick.

[firstlight]

You have captured quite a lot of subtle interaction in the corona. very nice smooth work.

[RickS]

Thanks, Tony.