Generating stereo series w Zerene

[iconoclastica]

I am doing a series of work (the fern spores) for which I now always do want to have stereo pairs and a central pMax. I used to do this sequential, first a conventional pMax stack, followed by changing the prefs to generate a 2-image +/- 1% stereo pair. Now I know I always want all three of them, I try to set the prefs to 3 images, still +/- 1%, expecting to be image_001 to be central and straight. Half of the time this is what I get, but in the other cases image_000 and image_001 are identical, and both of them identical to the separeate, non-stereo'd stack.
What could possibly cause this difference?

[rjlittlefield]

I am doing a series of work (the fern spores) for which I now always do want to have stereo pairs and a central pMax. I used to do this sequential, first a conventional pMax stack, followed by changing the prefs to generate a 2-image +/- 1% stereo pair. Now I know I always want all three of them, I try to set the prefs to 3 images, still +/- 1%, expecting to be image_001 to be central and straight. Half of the time this is what I get, but in the other cases image_000 and image_001 are identical, and both of them identical to the separeate, non-stereo'd stack.
What could possibly cause this difference?

This is quite bizarre! What you describe is a standard workflow: 3 images at +-X%, giving _001 straight down the middle, _000 and _002 making a stereo pair. I have never seen the behavior that you describe, and I cannot think what could be causing it.

I suggest this: the next time it happens, immediately do an Options > View Console Log, click once in the log window, then right-click Select All, right-click Copy, and paste the resulting log into an email addressed to support@zerenesystems.com . Attach also three screengrabs, showing each of the three output images selected and showing in the right-hand image panel. That should give me enough information to get started on figuring out what's going on.

--Rik

[iconoclastica]

I'll come back to this as soon as I get it again. It happened quite a few times last week, but right now I haven't seen it again. Doing fewer (longer) stacks at the moment.

[rjlittlefield]

Thanks for the update. I look forward to more information.

--Rik

[iconoclastica]

Thanks for the update. I look forward to more information.

--Rik

mail sent.

[rjlittlefield]

mail sent.

Received, solved, replied, and confirmed. Recording the solution for other people and future references...

Thanks for the project. I see what's going on now.

The problem is that Stack New Images cannot generate stereo on its first pass, because it does not know the number of images and therefore cannot know how much to shift each image.

In that situation it just sets the shifts to zero and forges ahead, unfortunately with no warning about what it's doing.

So the first stack output is zero shift.

But then the number of images is known, so subsequent stacks get the expected shift.

As a result, if you specify 3 outputs expecting to get -X, 0, and +X, and Stack New Images, what you actually get is 0, 0, and +X.

The only solution is to avoid stereo shift when running Stack New Images.

--Rik