rischa Theissen demonstrates a must-know technique for adding motion blur to your 3D renders in After Effects using ReelSmart Motion Blur.
ReelSmart Motion Blur
Leave a Comment
MILG11
MovieType For Cinema 4D
rischa Theissen demonstrates a must-know technique for adding motion blur to your 3D renders in After Effects using ReelSmart Motion Blur.
Thanks Grischa and John for sharing this cool tutorial, very useful indeed.
One of my most used plugins! I love it!
@Joel You’re welcome.
@Benjamin Indeed! I love it too.
I use most of the times RSMB but vector motion blur is not my choice, but NOW will be 🙂
Thanks very useful!
Cool that I could provide some new informations!
I was waiting this tutorial, cool stuff, very thanks Grischa
You’re welcome! 😉
great tip Grishcha! now we HAVE to get it for our studio!
Thanks Brett!
Maybe I get commission from revisionfx 😉
RSMB and particular are the must have plug-ins for AE i think. defiantly worth the purchase.
That’s right! The whole Trapcode Suite is worth it.
Thanks very much Grischa & John. Definitely one of those ‘A-ha!’ tutorials. Very very useful indeed.
Cool! Probably a nice time saver!? 😉
You’re welcome Mark.
Great tutorial. Looks awesome!
you save my millions hours of render in cinema!!!
i love you!
Hope you can spend them in the sun! 😉
hey Grischa! very cool tutorial. it is really essential, so thank u and thanks to John as well for hosting it!
little question. maybe im mistaken, but i think you could just pre-comp your C4D render, and add RSMB to it [not the vector one], and youre all set. Though, that whole Multi-pass rendering, and Alpha channeling those Green boxes- why is it so necessary to do? is it a lot faster than just adding RSMB to the comp?
Hi Danny!
Thank you!
Sure, you can just add RSMB without the motion vectors but if you do that it’s guessing where the objects are going. It’s not as accurate as getting the information from your 3D package and sometimes you get artifacts. Especially in my chase where a lot of objects are overlapping each other.
And it shouldn’t take to long to render out an extra motion vector in your multi pass.
But if you don’t have a motion vector you can always try RSMB without it.
You’re welcome.
You’re welcome ahmed!
Hope you saved some render time and get more accurate motion blur!
Really great and useful information – tuts like these make you the rock stars you are. Thank you Grischa and thank you John.
Could something like XMult be used on the vector motion pass so the edges do not need to be choked back?
Thank you Eric!
I am not sure about XMult. Never heard of that.
If it generates the alpha and can shrink it a bit it will probably work. I guess the more accurate way is to use the existing alpha of the RGBA image.
But give it a try and report back 😉
XMult is a free plug-in that does the same as Knoll Unmult, that is, unmultiplies a layer from it’s background. Good for creating mattes for flares, fire, smoke etc. Give it a try Eric, and you may also want to try the built-in “remove color matting” effect.
Thanks guys, I was thinking out loud when I typed that, and realized after that there would be some issues unmultiplying but thought I could compensate by adding some levels to the RGB and alpha. The result was just OK since the motion vector values were close BUT not accurate to the original.
However a second and simpler method that worked great was simply making a copy of the motion vector pass, desaturating it 100%, and clamping the the white levels so that everything was either completely white or black and using that as the matte. This method provided a 100% accurate matte for the motion vector pass with no choking needed, no black fringing, and with out effecting the RGB values of the pass. It looks great!
Excellent Eric, thanks for the tip. Best wishes, John.
Hiya! Thanks for going through this and a high five to Eric Litwin regarding his solution to the matte of the motion vector!!
One thing I’ve been thinking about though, is how would you combine Motion Blur with a Depth of Field blur.I haven’t had much chance to experiment with that, and was wondering how you’d stack the effects etc. Should you apply a motion blur to a pre-comped depth layer, before using it?
Thanks again!
Hi Anders!
To be honest: Never did it before.
I would try to apply the DOF first and on top of that the motion blur.
Thanks so much Chris!
hey i dont have a “PNG” out put in my save tab. wtf?! im in r11 if that makes a dif
Sorry al,
can’t help you out with that problem.
Good to hear that! 😉
btw AWESOME reel! Saw your work before and that are some great C4D skills there.
it may be a nice thing that you are doin some tutorials, bad thing is that you are so not familar with rsmb vectors. check out google first entry for rsmb vector: http://www.revisionfx.com/support/faqs/motion_vector_FAQs/motion_vector_alpha/ to see how NOT to use the motion vector. vector passes are not supposed to have an alpha chanel!
Thanks for the information Tobias. Shame you wrapped it in such an aggressive comment.
Hey Tobias:
You might want to reread that ReelSmart FAQ.
That link you’ve provided appears to say the exact OPPOSITE of your assertion that the motion vector channel should not have an alpha. The way I’m reading it, areas of a motion vector image that have an alpha below 100% will be ignored by RSMB, and the plugin will properly blur the moving elements. If you don’t precompose with an alpha RSMB thinks the black or whatever color is behind the colored vector images contains valid motion data and creates a corresponding blur, which is wrong.
So you’re wrong. And Grischa is right.
I’ve been learning C4D, and in my initial experiments with RSMB Vectors I was using the basic vector pass without an alpha and got weird results; all of the objects had a glow-like blur around all of their edges.
Precomposing the vector pass with the alpha from the main render fixed that weird blur, though I’m not happy with how fast-moving foreground objects look over objects farther back. The blur around the alpha looks good, but I’m getting a sharp edge where the 3D objects overlap.
ok, sorry for the agressive style of words, to much work, less spare time. keep on
Nice quicktip Grisha. There is one thing I can add and that is with big scenes you can use the single pass render output to create your 8-bit sequence in stead of rendering twice the multi pass. I have one question and that is if you don’t have an alpha in your 3d scene because the scene is full frame do you get artifacts because of the lack of AA?
Thanks Jasper for the addition! That’s a clever way to deal with it.
I haven’t tried it yet but I would say it still works without an alpha. It should at least in theory.
Have you tried it?
Hey Grischa.
Thanks for this tutorial, I was using 8bit motion vectors in the past. Now I know 16bit is better.
Can I also add something I do sometimes. If I’m using more than one image pass, ie diffusion, spec, ambient OCC etc…. I place the plugin on an Adjustment layer above all the passes.
This blurs all the passes at once.
Thanks Michael for the tip!
Adjustment Layers or Precomping should be the way to go with multipass renders.
Hey folks, this whole exchange is oh so useful, thank you to all for your expertise and encouragement-and Grisha, I have bookmarked your tutorial, without this community, who knows what I’d be doing, Best, Craig
BTW, has anyone here compared the results with RSMB, vs Z-blur?. I have never used Z-blur, but it is popular. Any advice/info is very much appreciated, thank you, Craig
z-blur is not a motion blur plugin, it’s a depth channel plugin, used to create depth-of-field effects.
Hi Shawn, thanks, I stand corrected!-what I meant to say was the frischluft plugin- curious, can anyone here please explain pros/cons of each? z-blur vs frischluft? if this is too off topic, I understand, thanks, Craig