reply to post by THE_PROFESSIONAL
It is not as good as the one from 2005 without stacking.
I've done the experiment, both from Tikaboo and other scenes. Stacking does not help. This about this for a second. The imagery is wiggling due to
thermal distortion. You stack some wigging shots, you end up with blur. Less noise mind you, but a blurry image.
When you do astrophotography, you are contending with about 2.5 miles of atmosphere. It is done at night (DOH!), so there is substantially less
thermal interference. Shooting Tikaboo is done in the daylight over 26 miles of air. It is a totally different ballgame.
Note also that your results depending mostly on atmospheric conditions. Some days are clear, and some days you can't even see the base. So it is very
hard to compare image techniques from one session to another. Rather, you need to try your techniques at the same session to remove atmospheric
conditions out of the equation.
This is the first google hit from image stacking, and it follows pretty much what I know from studying signal processing and real life results:
If you check out this page, you can see shots at full scanner resolution (film) showing how much the image changes from frame to frame:
all images are not created equally
This is an image of the hew hangar I shot:
Because I used film, you need to deal with grain, but otherwise it is not too shabby.
Here is the whole webpage:
I never bothered to link it since I have a digital body now. Unfortunately, a lightning storm prevented my last photo session.
Getting back to image stacking, the problem is that term is a poor description of the process. Rather than stacking, you are averaging. It is a
completely linear operation done in the time domain.
There is one stacking technique I've played with that might improve image quality that is done in the frequency domain. A commercial program
"helicon" does this, as well as a free program called imagej.
extended depth of field
This nonlinear scheme sort of improves the stack by using the sharpest parts of the image to make the final composite. However, there is a problem in
that sometimes the very thermally distorted images in themselves are pretty sharp if the shutter speed is fast enough, so it doesn't always do it's
I think the best bet out of all these techniques is simply to shoot the same scene multiple times and pick the sharpest frame. There may be some other
technique I am not aware of that might be better, so I am open to suggestion. But I have spent considerable time experimenting with image summing and
don't believe it is useful for daylight telephotography.