Jpeg Sun stack, Better than RAW! Jan 29th

[ArmyAirForce]

I did a little experiment this morning shooting my usual 100 RAW images of the Sun, then did 100 and 200 Jpegs, with the Nikon D50 set to "Jpeg Fine" and sharp detail. They shot and downloaded to the card much faster. Perhaps the seeing improved a few minutes later when I shot the Jpegs, but even stacking a similar amount as the RAW, I got cleaner results. I'm going to have to experiment some more and see if the results are repeatable.

Jpeg 26/121 stacked

RAW 23/109 stacked

[michael.h.f.wilkinson]

Nice shots, but difficult to call whether the jpeg is better due to seeing. The camera might be applying some sharpening to the jpeg output which isn't done in the raw output

[ArmyAirForce]

I'm still using wavelet sharpening on both after stacking, and the Jpeg came out better. More easy to see the difference in the original, non-uploaded files.

[Montana]

Lovely image, such super detail :)

Alexandra

[Davey-T]

Good stuff, to be honest the small amount of W'Light I have done has all been JPEGs, nice to see a bit more activity since the last look I managed.

Dave

[Pete Presland]

Both images are very nice And another active region on its way as well for the weekend! The JPEG does look a little sharper though but as Micheal said the camera might be applying some sharpening to the jpeg output.

[ArmyAirForce]

....The JPEG does look a little sharper though but as Micheal said the camera might be applying some sharpening to the jpeg output.

It is, but I still can't sharpen the RAW and get as good or better results, either via wavelets or normal sharpening tools.

[rwg]

It may well all be down to noise - RAW images are considerably noisier than jpeg (because the camera applies noise removal to the jpeg images). Extra noise will obviously lead to problems when you start sharpening.

Try doing some noise removal on your raw stack before sharpening and see if that helps.

cheers,

Robin

[ArmyAirForce]

I think I've worked it out. It seems to be down to field rotation. Most of my recent solar shots have been handheld with the 80ED balanced on the window sill. I can shoot 100 Jpegs in 1 min 14 sec, but it takes 3 min 8 sec to shoot 100 RAW images. In that time, the Sun has moved and rotated noticably.

When it came to processing, I used PIPP to crop to 1500 pixels square and output as TIFF, then a single align point on a Sun spot, in Registax prior to stacking. I've just tried stacking the 100 RAW files with multiple align points on several spots, and the result was much better. Today's set of Jpegs were still better than the RAW, but after stacking with multiple align points, the results were much closer. Jpeg was still fractionally better, no matter how the RAW was processed, but only just ( which could have been down to a change in seeing ). Next chance I get, I'll do the same tests, but with the scope on an EQ mount, and see how they compare.

[DrRobin]

Hi,

I am surprised that field rotation is noticeable over such a short period, I wouldn't have thought it would be. However, you are always better off getting as many pics in a smaller time as possible, less things change, like sunspots groups and seeing. When I take 12 panes for a mosaic it comes out much better if I get them all in about 10 minutes rather than an hour, if there is cloud.

I would always use multipoint stacking, the sun is so big that a single point is never going to track and align it correctly.

Both are really good shots, but the jpeg does look better, so I would shoot on jpeg from now on, it will make the processing faster as well, less file to load and handle.

Robin

[ArmyAirForce]

I had a go at the Moon on the 30th, but with the scope on the NEQ6 mount. Both Jpeg and RAW came out well, though it is difficult to make a direct comparison, as the Jpeg was against an evening blue sky, and light levels had dropped a bit for the RAW, so that came out a black sky. With both giving good results, the small amount of field rotation is most likely the factor affecting the images, though I'll continue to experiment.