#Detailed Eyelashes - deformers not playing nicely for me

61 messages · Page 1 of 1 (latest)

hoary spruce
#

So how would I go about rigging up eyelashes like I sketched up?

So far I've tried the following:

Deformer Hierarchy as follows:

  • XY Position (warp)
    • Open/Smile (warp)
      • Physics (Deform path on the mesh)

--> problem here is that doing this doesn't work on the closing. once it leaves the XY deformer when closing it returns to pointing backwards the other way instead of being warped to twist the way I'd like it to. Increasing the XY Position size does not help as the kind of warp deformation to facilitate end and start point to look any kind of good makes the journey between points very broken


Deformer Hierarchy as follows:

  • Open/Smile (warp)
    • XY Postion (warp)
      • Physics (mesh)

--> problem here is that on closing it inverts the relative position, so once the Y rigging is involved, closing makes the eyelash disappear higher up instead of going down with the eye closing.


Deformer Hierarchy as follows:

  • XY Postion (Warp, though could be rotation as it's just position, no warping into shape.
    • Open/Smile (warp)
      • X form adjustment (warp, local form change for the X turn)
        • Physics (mesh)

--> problem here once again is the same as the first iteration, and increasing the size is not an option.

putting my eyelashes as is into the Eyes XY deformer does not do what I want because the eyelashes will be stuck pointing to the side they point in the neutral position instead of away from the head.

how do I go about this properly and/or is there a good tutorial I should've seen but haven't?

supple bridge
#

Can you post a video of how it looks with the methods above? I have some ideas but seeings the movement you have currently would help more

fast venture
#

use a blendshape to make the eyelashes curve out instead of permutation

hoary spruce
#

Will take vids and try out the blendshape way as well once I'm back to work on it for sure

hoary spruce
fast venture
hoary spruce
#

yep I'm fully aware but again if I made the parent bigger I'd have to make a > Shape with the deformer to make the end position look good and it simply does not

fast venture
hoary spruce
#

which is where the cyclic dependency seems to come in in my mind. like.. I need the close to move with the XY but also the XY to move with the close and it just kinda doesn't do that nicely

fast venture
#

my 100% honest opinion

a) you can re-rig from 0 with the correct hierarchy for this using normal parameters
b)use blendshape

hoary spruce
hoary spruce
#

like hierarchy

#

I don't mind trying either

fast venture
hoary spruce
#

sorry I should've said what the correct hierarchy is here?

fast venture
#

so far what i see is that the 3 methods you made
all have the same issue
child moving out of parent

hoary spruce
#

well they always will. unless I make XY bigger and when that happens hold on let me show the issue with the last method if I make the XY deformer bigger to always keep the child inside

fast venture
#

If we go for Option A, that is only Normal parameters
you would have to do something like
🟩 AngleXY of the eyelids [both eyes]
-- 🟩 Eyeopen-smile [individual eyelash]
-- -- 🟩 AngleX-Y curve [individual eyelash]
-- -- -- 🎨 individual eyelash physics

something like that, but you would have to make sure each warp deformer is big enought so the child is always inside of it in all keyforms-permutation they have

#

also you would have to work really well the eyeopen-smile deformer so it does not flip overitself whe close

hoary spruce
#

unless I have to do better on the X position adjust, then I have to correct the position on the XY deformer and this looks terrible.

fast venture
#

what is that long deformer?

hoary spruce
#

XY position

fast venture
#

mmm the one i mentioned as parent of eyeopen is the full both eyes

hoary spruce
#

I still probably have to fix the Open/Close then as the X position adjusting moves it out of the parent

fast venture
#

there is no reason for it to move out of parent

hoary spruce
#

I've now put the X position though into the eye XY deformer

fast venture
hoary spruce
hoary spruce
#

no like look

#

the X adjustment curve is child of Close but

#

it doing its deformations causes it to go outside the range of the Close parent deformer

fast venture
#

it shoudnt go outside

#

remember you are re-rigging from 0

#

A option said "re-rig"

#

that means delete everything and do from 0

hoary spruce
#

that's why I said I'd probably have to re-rig the open close, as you can see, parent is too smol right now

fast venture
#

there is not a "probably" with "A" option you Have to

hoary spruce
#

also if I were to do X curve as blendshape, would it be on the mesh?

fast venture
hoary spruce
#

okay no that one seems rather unintuitive. on the open close feels weird. I'd rather the curve deformer and redo open close tbf

fast venture
#

remember how blendshape works

#

Blendshape is a parameter Additive that applies a deformation to all other keyforms on the same object

hoary spruce
#

yes I do know, the issue I feel is with putting it as blendshape on open close instead of blendshape on physics I guess

fast venture
#

you want the curve to look correctly in all open-smile when angleX is turned

fast venture
hoary spruce
#

right then I guess how do I make the X curve blendshape move on Angle X?
like I'm assuming physics like how I've done the augmented physics but I want it to follow 1 to 1 and not have a bounce

fast venture
#

did you made augmented physics for angleX-Y right?

#

that you have
angle X move AngleX Aug

#

well you called it angleX physics

hoary spruce
#

should I just duplicate that physics group and replace my Angle X Phys with the X adjust?

fast venture
#

no

#

you take the same physics group
go to Outputs where angle X phys should be as pendulum 1
and you add the blendshape X on the same pendulum, same scale

hoary spruce
#

oh. add it to the same pendulum in output settings and also tie it to that 1 node?

#

yeaaa

#

okay yep got it