I am doing sphere images and taking three rows of images, up, level, down, with a 30% overlap between images in each row.
In AutoPanoGiga 3.0.8 I arrange the images from each row into layers (up, level, down) and would like to know if there is a way to weight the "level" layer so these images are preferred over the up and down image layers during blending.
I've tried different blending techniques and masking without much success.
Does anyone know if a layer can be weighted so it is preferred during blending?

Thanks very much

Can you not just flag all the required images as Reference Images (you can have as many reference images as you want or none at all) and set Colour Mode=Auto

I'll try that but I think that will only influence color blending, not preference of the images for content. The problem is that my lens has better resolution in the center than on the edges, so the lower edge of the top row tends to be more distorted with less resolution than the level row with the horizon in the center.
For some reason AutoPanoGiga often selects the lower resolution top row or bottom row leaving a poor resolution blotch in the finished image. I can fix it in photoshop but it would be much better if I could somehow get AutoPanoGiga to give preference to the sharper middle layer.
When I try the diamond blending approach (which is supposed to do just that) I get more ghosts of leaves and branches.

I am not aware of any means of controlling that in APP/APG.

I believe it may be possible with PTGui.

There are a number of lens distortion parameters you can set in the APG "Settings" in the Optimization tab so it might be worth an experiment with these and also to have a look at the User Guide

It is unusual to find that much difference in the middle shots compared to the top/bottom - can we see some screenshots

APG Masking should work so long as you leave some overlap area for APG to process but it seems to me as though this is just circumventing the real problem

There are others on here who have more detailed knowledge of these sorts of distortion problem !

best
Martin

What lens are you using and with which camera body?

I am using GoPro Hero3 cameras. They do a pretty amazing job considering their size and cost. The lens is a 15mm equivalent. The corners are softer than the center and the fall off is worse underwater because the housing uses a flat port. So I stagger the images from the three rows putting the soft corners in the central, sharp area of the overlapping image. Here is an example of an underwater image - in general I find the results are equal to what I get with my Canon 7D:
http://www.360cities.net/image/triton-attack-crown-of-thorns-starfish-new-caledonia#358.80,5.70,60.0

I am uploading some above water full resolution crops from the top, level and bottom rows.




AutoPano can do deep focus images from a stack of images taken at different focal distances. But it does not seem to be working too well and sometimes will select the soft areas from the top or bottom row instead of the center row (where I usually have the most important part of the pano).

I have tried changing the optimization parameters but the default settings still do a better job. If anyone has any suggestions for settings that would be a big help.

By the way, I do have PTGui but AutoPano is so much easier to use, takes less time, and has always produced the better image for me. I like AutoPano