Jim Wilson writes: > Setting all the view offsets to 0 I was able to prove that the > position/rotation matrices generated on the model and the camera are > numerically identical. Here's a sample from the dump: > > POS Aircraft > 0.823776 -0.150286 -0.546632 0.000000 > 0.468893 0.722572 0.507965 0.000000 > 0.318641 -0.674762 0.665706 0.000000 > 5064.624023 590.030945 -1211.297729 1.000000 > VIEW From > 0.468893 0.722572 0.507965 0.000000 > -0.823776 0.150286 0.546632 0.000000 > 0.318641 -0.674762 0.665706 0.000000 > 5064.624023 590.030945 -1211.297729 1.000000 > > POS Aircraft > 0.823778 -0.150284 -0.546631 0.000000 > 0.468911 0.722528 0.508011 0.000000 > 0.318610 -0.674809 0.665673 0.000000 > 5064.621582 590.031433 -1211.296509 1.000000 > VIEW From > 0.468911 0.722528 0.508011 0.000000 > -0.823778 0.150284 0.546631 0.000000 > 0.318610 -0.674809 0.665673 0.000000 > 5064.621582 590.031433 -1211.296509 1.000000 > > Note that one is applied to the object and the other the camera. If you look > closely you'll see that the numbers in sync. This is only a couple of > iterations, but the values don't diverge over time either. > > With no offsets what so ever to either the model or the camera the jitters are > still there (I modified a model in order to observe the jitters at 0,0,0 > offset). The jitters are also there in "Chase View" so it has nothing to do > with rendering the cockpit in a seperate ssg root. > > This tells me the problem is most likely in plib, or at least plib > is where we send these numbers.
Plib for the most part is a very thin layer on top of opengl so these matricies are probably getting sent directly to opengl so most likely we are achieving some sort of numerical precision problem that is affecting opengl. This sort of thing probably isn't a bug with plib. Regards, Curt. -- Curtis Olson IVLab / HumanFIRST Program FlightGear Project Twin Cities [EMAIL PROTECTED] [EMAIL PROTECTED] Minnesota http://www.menet.umn.edu/~curt http://www.flightgear.org _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
