Hi, maybe this is offtopic, but I hope that someone here uses video4linux for opengl rendering too. I got a video stream into host memory without any problems. Now I want to display it as a video texture. So first I make a glTexImage2D call with a 1024x1024 sized texture. The for every video update I use a glTexSubImage2D call. But this is very slow on my GeForce3 board and I wonder why. It takes over 10ms without mipmapping. Only for the transfer. Is this normal. I hoped that AGP 2x is much faster. The texture format ist RGB with 24bit for each pixel. The format for my video data in host memory is 24bit BGR, but the conversation from BGR to RGB doesn't matter. Have someone made the same experiences with the performance when getting video data into the texture memory of my nvidia board? thanx, Arnfried
