Mind reader. I was dealing with this problem today at work, for maybe 3 hours and no success? And then, out of no-where, here you release a video with the answer. Law of attraction in effect? :-D Super thanks!
Well, your strand would need to be moving and maintaining its point order for motion blur to be generated. When people go to make strands from particles, it's often made by generating it from the point's trajectory. And, because of this, the topology changes at every frame. For something like hair/fur this isn't an issue, but I would generally say that it's probably better to either go the route of particle motion blur or generating strands if you're doing this for moving fx elements.
@@cgforge Not really sure about that, isnt deformation blur calculated based on point number? Steven Knipping has a workaround where he creates the final point amount and then makes the point number into id and transfers the simulated points position to these points but thats not really practical.
@@linusrosenqvist9145 I've found that redshift internally recognizes the id attribute and will automatically switch from point number to id when present
@@cgforge i have simple popsim with changing point number and id attribute, but redshift doesnt feel like to render it with blur it just says my point topology is changed so can only use motion vectors which only keeps giving me linear motiunblur
What if I want different pscale motion blur on particles, per frame? So they give out a bit of a un-uniform look? Can get this to work in Mantra, but not RS.
Tyler Bay using a modulus per 1 frame in a VOP to calculate a different pscale per subframe. In Mantra, this works to make non-perfect motionblur for a rain effect. Cannot get this to work in Redshift, maybe it’s the subframes?
@@Fakepilot Hmmm I haven't tried that one yet. My guess is that redshift is looking at velocity instead of your subframes. Be sure to double check your geo spreadsheet for that changing pscale value first though. If you step through subframes, you'll want to confirm that those values are changing with each subframe.
@@cgforge I checked Integer frames off, velocity deleted, I see a different pscale in the spreadsheet for every subframe, but still nothing different in the RenderView.
@@Fakepilot Hmmm yeah I'm not sure what the fix is off the top of my head to be honest. You could try using a trail sop, connecting the previous subframe with a polyline, and then applying motion blur to the resulting strands. That might be an interesting route to go, but at this point, I'd try posting on the redshift forums for the devs since they understand the inner mechanics better than anyone else.
Mind reader. I was dealing with this problem today at work, for maybe 3 hours and no success? And then, out of no-where, here you release a video with the answer. Law of attraction in effect? :-D Super thanks!
Thanks man, this always trips me up.
amazing! my question will be how do you get motion blur if you render strands?
Well, your strand would need to be moving and maintaining its point order for motion blur to be generated. When people go to make strands from particles, it's often made by generating it from the point's trajectory. And, because of this, the topology changes at every frame. For something like hair/fur this isn't an issue, but I would generally say that it's probably better to either go the route of particle motion blur or generating strands if you're doing this for moving fx elements.
Saved the day for me :)
Very useful tips in there, thanks a bunch. Cheers.
So this would only work with a constant point count right? I find that thats very seldom the case with pop sims
So long as you have an id attribute, you should be good to go
@@cgforge Not really sure about that, isnt deformation blur calculated based on point number? Steven Knipping has a workaround where he creates the final point amount and then makes the point number into id and transfers the simulated points position to these points but thats not really practical.
@@linusrosenqvist9145 I've found that redshift internally recognizes the id attribute and will automatically switch from point number to id when present
@@cgforge oh, thats really great, strange that it didnt seem to work for me, guess there most be something else i overlooked
@@cgforge i have simple popsim with changing point number and id attribute, but redshift doesnt feel like to render it with blur it just says my point topology is changed so can only use motion vectors which only keeps giving me linear motiunblur
What if I want different pscale motion blur on particles, per frame? So they give out a bit of a un-uniform look? Can get this to work in Mantra, but not RS.
So you're saying change the pscle of each particle so that they are different sizes?
Tyler Bay using a modulus per 1 frame in a VOP to calculate a different pscale per subframe. In Mantra, this works to make non-perfect motionblur for a rain effect. Cannot get this to work in Redshift, maybe it’s the subframes?
@@Fakepilot Hmmm I haven't tried that one yet. My guess is that redshift is looking at velocity instead of your subframes. Be sure to double check your geo spreadsheet for that changing pscale value first though. If you step through subframes, you'll want to confirm that those values are changing with each subframe.
@@cgforge I checked Integer frames off, velocity deleted, I see a different pscale in the spreadsheet for every subframe, but still nothing different in the RenderView.
@@Fakepilot Hmmm yeah I'm not sure what the fix is off the top of my head to be honest. You could try using a trail sop, connecting the previous subframe with a polyline, and then applying motion blur to the resulting strands. That might be an interesting route to go, but at this point, I'd try posting on the redshift forums for the devs since they understand the inner mechanics better than anyone else.
incredible ! thank a lot !
this... is very helpful right now 😂