Curtis L. Olson writes:
>
> Norman Vine writes:
> > 
> > This method should also be *considerably* faster then the current
> > Panel code in that the actual updating of the instruments can be
> > done on a round robin basis reducing the work done per pass as
> > all but the instrument texture being reconstructed on this pass 
> > will just be a single texture layer probably rendered as a quad !
> > 
> > An individual instruments updating could perhaps be delayed a
> > bit more further improving performance if the actual instrument 
> > being modeled had a 'real' sampling rate we wanted to simulate
> > 
> > lazy-evaluation-yields-fast-programs'ly yr's
> 
> One of the knocks against programs like Fly and MSFS is their horribly
> slow instrument update rate.  That's one thing poeple really like
> about programs like X-Plane and FlightGear.  You can actually fly off
> the instruments because they move very smoothly ... because they are
> updated every frame.  And we would also need to be careful of texture
> space since this sort of scheme is going to increase texture ram
> usage.

Good point  .....but .....
If we can half the frame rate hit of the Panel using this kind of scheme, 
and I expect even a better boost, then we could render half the instruments 
one frame and the other half the next and still have the ~same instrument 
update rate as we currently have and an overall ~25% boost in framerate
when the Panel is displayed

(1) This is assuming that the Panel rendering hit is approx half of the overall
cost of a complete frame

Norman


_______________________________________________
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel

Reply via email to