Hi Mohsen,
Try this:
*Create a Default PFlow
*Invert the Speed, enter 30.0 degrees divergence
*Increase the particle count to 1000
*Disable the Shape operator
*Frost it.
Here is frame 15 of my scene:
*In V-Ray, open the Camera rollout, enable Motion Blur and change Duration to 0.999 and Interval Center to 0.0
*Render
You should get motion blur on your Frost:
If you leave the Duration at 1.0, V-Ray will try to evaluate half a frame before the current frame and half a frame after, thus getting two different topologies and will issue a WARNING that topology is changing and not blur at all. Setting it to anything less than 1.0 (we use 0.5 here anyway) solves this.
If you leave the Interval Center at 0.5, the same will happen because the center of the interval will be between two frames and will sample both the previous and the next frame.
Obviously these are some significant restrictions, but that’s how VRay works.
Also keep in mind that the “Frame Velocity Offset” option does not produce “correct” velocities. What it does is take ONE frame (the closest to the sample time) and for each vertex, assign a velocity based on the influence of the particles contributing to that vertex. So if you have 100 slow moving particles and one very fast moving close to the same vertex, the result of the motion blur might not be exactly what you expect.
The correct motion blur is produced in “Subframe Particle Position”. In this mode, the position of each particle will be evaluated on each sub-frame, and a NEW MESH is built for that sample which can (and usually will) change topology vs. other sub-samples. So this mode produces unique meshes on each sub-sample and is best suited for Multi-Pass Motion Blur which is of course slow, but it looks right. The “Frame Velocity Offset” mode will looks wrong if used with Multi-Pass camera effect.
Alternatively, you can force VRay to evaluate a large number of Geometry Samples. The following example shows Shutter of 1.0, Interval Center 0.5 and Geometry Samples of 16. Obviously this produces 16 unique copies of the mesh (similar to MPass), and there will be a warning that the topology is changing, so the 6 sub-steps won’t be performed:
A word of WARNING: If your particle source is a PRT Loader and it is set to RETIME the sequence, this could break the Motion Blur. Let’s say you have a PRT Loader with Playback Graph set to play 2 times faster than the original sequence. If you pick this in Frost and try to use “Frame Velocity Offset”, the PRT Loader will produce two different sets of particles on the before and after sample because internally it will interpolate the existing frames and skip one, thus causing Frost to mesh two different topologies again. If you have to retime a sequence, always bake to a new PRT sequence first, then pick that in Frost and it will work as shown in the beginning…
I hope this gives you a basic idea of what is going on internally in V-Ray and Frost and you will be able to combine this knowledge to render in other renderers.