Here’s an image with a sphere with a vray light material animated along a path over 2 frames with frame 1 being rendered. I tried using a vray light instead but the sampling was a bit shitty. Also I tried animating the colour of the vray light material to see if the colour would change over the streak but it seems vray won’t do this.
Very similar to what I was thinking of testing. Thanks for sharing.
Now that is very interesting… U guys use Vrayphys camera or standard camera?
Also I have to go up to 100 subd samples ? It kinda sky rocked by wheel when I pas 20 on geometry…
I used a standard camera; you certainly don’t need 100 geometry samples unless your wheel spins insanely fast. For car wheels I would think 6 to 8 would be enough.
Best regards,
Vlado
I have been playing with this same effect for another gig, and actually gave up on the vraycamera blur. It just doesn’t seem to behave properly when the shutter speed gets really low, and I needed that for the streakiness.
Using the vray camera option in the render dialog is working better for the movement, but I’m getting missing buckets - almost like a DR problem, but the DR is turned off. They move around from render to render.
It’s also super grainy, even with sub-d set to 100 and geometry samples at 100
Any tips?
b
If you are using the DMC image sampler, try increasing the min. subdivs value to 2 or more.
Best regards,
Vlado
That helped with the missing buckets - thanks. Still very grainy though. I think I will probably stick with the animated emitter/facing particle approach I was using before though. It doesn’t depend on motion blur at all and is a bit faster to render clean. Lacks some of the look, but I’m still playing with the textures and approach.
Lovely - I stuck with the mesh since I was getting better sub sampling results right away but didn’t get around to aa or mb subdivs. Aren’t you supposed to be on holidays though ![]()
What is your duration setting? Try geometry samples of 10 and play with duration. Set interval center to 0.5.
jbeau3d, this may be a result of an incorrect rotating motion. For example if your wheel is spinning too fast, then even 100 geo samples will not be enough to approximate positions of geometry. In your last test, you can see the sample stepping that seems to reflect this theory. Also, do make sure your wheel is actually spinning in an circular motion, in 3ds max, in some cases if your wheel is animated on multiple axes it may be flipping.
Not sure if anyone is still following this thread, but the issue should now be fixed. You can read more about the problem and fix in this thread.
-Eric
Awesome thanks mate.



