Ive stumbled upon something I think might be a bug, its easily avoidable but still.
When rendering with motionblur and an interval center of 0 first frame of the timeslider only gets a portion of the blur justified (using linear out-of-range types so theres data on the movement before this frame). Im unsure if this is something that is correct in render theory but to me it doesnt seem that way. Setting the first frame of the timeslider to one frame before I want to render and rendering the same frames using “range” in time output resolves the problem.
Its really easy to reproduce but let me know if you want an example.
Is it correct tho? Its almost a philosophical question To me its wierd that changing the animation range changes anything but the amount of frames. Its also wierd to me that theres different results depending on if its the camera or the object thats moving, shouldn’t it be just the relation between the two that matters?
For its own internal reasons, V-Ray pre-caches the camera motion for the entire animation range (this is used e.g. when doing fly-through calculations for the light cache).
It is not so odd for a plugin to consider only the portion of the scene within the animation range - for example, the Havok dynamics simulation, the Cloth modifier etc. also do that. Along these lines, you may consider V-Ray as physical simulation of a sort…
To me its wierd that changing the animation range changes anything but the amount of frames. Its also wierd to me that theres different results depending on if its the camera or the object thats moving, shouldn’t it be just the relation between the two that matters?