hello, the right pix_image reset the "new image" flag that is set by the left pix_invert. (you can see that thanks to a pix_info)
i thinks it's a bug that should be corrected : the left pix_image should not be affected by the right one, at least, a pix_separator should prevent that. currently, toggling only the left pix_invert does not only have no visual effect, but it do also trigger computation of the one on the right. here is a workaround this problem : setting the "new image" flag only after the right part of the chain is rendered. Cheers c
Le 20/09/2012 17:19, Marian Weger a écrit :
Hello list, I just stumbled across some strange behavior of Gem. I attached a small patch to demonstrate this problem. If multiple [pix_image] connect to the same gemhead (with separator), changes in "pix_..."-effects that are connected below get only effective for the first one. The resulting texture of the other branches gets only updated if the image is reloaded. Of course, everything works fine, if i use individual gemheads for each branch. Is it a bug or a feature? How can I solve it? Tried it in Pd-0.43-2 with Gem 0.93.3 on OSX and Pd-0.42.5-extended on Windows. thanks! marian _______________________________________________ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
pix_image_test.pd
Description: application/puredata
_______________________________________________ Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list