When I use AP to process a set of .hdr images and write a .hdr output, AP seems to sometimes exaggerate the brightness of in one of the source images causing a "bright spot" in the final image. At first I thought that what AP was showing me was an anomaly caused by the problem of displaying a 32-bit image file on an 8-bit monitor, and that when I rendered the anomaly would not be there, but it is.
For example, the first screenshot shows the situation after loading and stitching 10 .hdr files. These files were made using Photomatix in batch mode from sets of bracketed images with the same exposure/exposure spacings. The preview on the right shows a bright spot and in the second screenshot you can see that this seems to be image no.8, c02-03.hdr which is also the reference image. The third screenshot shows the stitched .hdr file in Photomatix's .hdr viewer - the portion of the image corresponding to the bright spot is clearly way too bright.

Can someone tell me what is going on here please? My thinking is:

1. It could be that there is a problem with image 8 but I am sure it was made from bracketed images in the same way as all the other images and that it was made from the same number of brackets with the same exposures/exposure spacings.
2. It could be that when I took the original photos, there really was a bright spot on the wall. I don't think this is true. In any case, if you look at the original photos, the bit of the wall in image 9 is even brighter (because of the table lamp at the bottom of the picture).
3. It could be something to do with image 8 being chosen as the reference image, but I don't know how significant this is because I have never used colour correction/anchors. I understand the general concept i.e. the other images colours will be normalised relative to the reference image, but image 9 is brighter than image 8 and yet it is not brighter in the preview or the final render.
4. AP has a bug of some sort.

Comments? Opinions? Workarounds?

A

Last edited by Aeriscera (2013-03-12 19:19:54)