Use the input resolution in the blur instead of hard-coding 1280x720. Also fix mipmap...
authorSteinar H. Gunderson <sgunderson@bigfoot.com>
Tue, 9 Oct 2012 10:16:30 +0000 (12:16 +0200)
committerSteinar H. Gunderson <sgunderson@bigfoot.com>
Tue, 9 Oct 2012 10:16:30 +0000 (12:16 +0200)
commit322c03693ae23e5a5843dee9dcb74dfb3b7d43d8
tree6c222a18471ee1ee3b2417cc831b8bef42c0e8e5
parent048e90293b5f5ed4b922d60c4cb47c4b5d24b85e
Use the input resolution in the blur instead of hard-coding 1280x720. Also fix mipmap calculation by odd sizes, so that we know we will hit exactly on the right pixels when we sample on the first pass.
blur_effect.cpp
blur_effect.h