NikonUser wrote:rjlittlefield wrote: I still can't make the numbers work out to explain such a large halo just from perspective effects, but if you're confident there's no shift, perspective must be the problem.
--Rik
Does the following help to confirm perspective effects rather than an alignment problem?
(I take images in this sequence, 1st is the front of the subject, I then move the subject towards the lens for subsequent shots.)
I opened the first and last images in Photoshop, pasted the last image into the 1st and made this last image 50% opacity and then flattened the image. Appears to me that they line up perfectly, note that the halo around the front wing is much less in this image.
Great test, and fascinating results!
I pulled your earlier HF image (from 9 posts above here) and your Photoshop'd image back into Photoshop so I could see them together and click between them. (To make viewing easier, I also did a very minor rescale and shift to get the two results registered against each other, presumably due to slightly different cropping in preparing them for posting.) Then I sat for a while, clicking back and forth and trying to make sense out of what I'm seeing.
This is a challenging problem.
I agree, the front and rear images appear to line up perfectly with the HF result, and there's a lot less halo in this image.
What this tells me is that none of the previous explanations are correct. Despite what I wrote, the halo is not coming from either shift or perspective.
I had been assuming, and I think Charlie had as well, that the haloed-out portions of rear wing were completely obscured by OOF front wing.
Obviously that's not true, since in your Photoshop merge we can see rear-wing detail clear to the edge of the front wing. No doubt the contrast of the rear wing detail is reduced because of the OOF front wing, but there's still plenty left to work with.
So the question becomes, why did the software choose not to show that rear-wing detail, in the vicinity of the front wing?
With Helicon Focus, big halo will sometimes happen if the "Radius" or "Smoothing" parameters are increased a lot beyond their default values of 8 and 4. But then I would expect to see halo in other places too, like where the base of the front wing occludes those bristles on top of the thorax.
So I'm confused. I can't recall seeing this combination of effects before, and I don't understand what's going on. That makes it a learning opportunity for me!
One other thing I notice is that the Photoshop merge seems to show a lot more fine detail than the HF result does. Take a look at the tips of the wings. In the Photoshop image, the fine serrations on the leading edges are sharply resolved, and there's a lot of fine-scale texture in the membrane, both in the clear regions and in the darker spots. In the HF image, the serrations are blurred and there's much less texture visible in the membrane. These differences may be hard to see in separated images, but they become obvious if you do the layer-and-click thing in Photoshop.
Maybe the difference in visible detail is due to JPEG compression. I notice that both files are about the same size. That would allow a higher JPEG quality in the Photoshop merge, since most of that image is OOF. But conceivably it's due instead to whatever process is generating the halo, and it would be nice to know that's not the case.
So, more questions...
What method and what parameter settings are you using?
And if you look at the actual pixels on your end, is the HF result losing detail compared to the Photoshop merge, or more importantly, compared to the original images?
--Rik
) subjects. But when I look at the output 1:1 the results from HF seem to look more "naturally photographic", and it doesn't mess around with the contrast as much. 
