POV-Ray : Newsgroups : povray.binaries.animations : Request: Rune's foot placement algorithm ( 0kbbu!) Server Time
2 Nov 2024 13:19:35 EDT (-0400)
  Request: Rune's foot placement algorithm ( 0kbbu!) (Message 1 to 3 of 3)  
From: Greg M  Johnson
Subject: Request: Rune's foot placement algorithm ( 0kbbu!)
Date: 30 Oct 2001 22:19:24
Message: <3bdf6dbc$1@news.povray.org>
[As I understand the server, it would be a discourtesy to post big anims
in p.a., but no real harm done to have text discussion in p.b.a.,
right?]

Rune, you were kind enough to share with us your actual code for that
spiral walk you had.  If you have time sometime soon, could you
pseudo-code this (verbal description) for us over-35 types who can't
follow your code?

As an example, here is how I do my feet:
1) An include file is called which has the position of the
"toeballpoint" of the left and right foot in it. This is stored as a
linear spline of <x(=0), y, and z> values as the "clock" goes 0 to 1.0.
2) The position of the knee is calculated by IK.
3) The rotation (FK) values required to get knee and toeballpoint at the
desired location are calculated.
4) An include file is called which is the blob body, purely FK in
construction.  The joints rotate according to the defined variables.
5) The body is moved at rate x, while the include file in 1) is moving
the foot backwards at rate x.

With this setup, I can imagine it being really hard, really easy, to
implement a walk cycle along a winding path.
Q's:
i)  Do you absolutely require the foot to stay put, with no rotation
slippage, no matter how sharp the curve??
ii)  How do you do yours?  The huge favor I'm requesting is just the
algo, probably no more detailed than above.  Apologies again for being
so dense as to not understand your code.


Post a reply to this message

From: Ken
Subject: Re: Request: Rune's foot placement algorithm ( 0kbbu!)
Date: 30 Oct 2001 22:52:14
Message: <3BDF7573.CCBC4FED@pacbell.net>
"Greg M. Johnson" wrote:
> 
> [As I understand the server, it would be a discourtesy to post big anims
> in p.a., but no real harm done to have text discussion in p.b.a.,
> right?]

Actually the purpose of povray.animations is text based discussions on
the techniques and tools to make animations. However, I don't think we
are going to burn you at the stake for this one transgression.

-- 
Ken Tyler


Post a reply to this message

From: Rune
Subject: Re: Request: Rune's foot placement algorithm ( 0kbbu!)
Date: 31 Oct 2001 09:22:42
Message: <3be00932$1@news.povray.org>
"Greg M. Johnson" wrote:
> Rune, you were kind enough to share with us your actual
> code for that spiral walk you had.

Well, a strongly simplified version of it meant more as a hint than as a
finished solution... :)

> If you have time sometime soon, could you pseudo-code this
> (verbal description) for us over-35 types who can't follow
> your code?

Well, I can try.

...

Ok, I tried, but I couldn't without having any real code to demonstrate
with.

So instead I've added lots of comments to the code example and posted the
new version to povray.binaries.scene-files in the same thread as before. I
hope that helps. If not, just ask more and I'll be happy to answer as good
as I can!

Note that the code example only demonstrates the very basics of the
technique, with just one leg, very simple motion etc. That's not very
useful. You're supposed to rewrite it completely to adapt it to your needs!
;)

> With this setup, I can imagine it being really hard, really
> easy, to implement a walk cycle along a winding path.

I figure it'll be some work for you. You approach uses more or less a still
standing body with the ground moving backwards beneath. You can't do that my
method, but have to move each part of the body (including the torso) with
individually calculated transforms. You *may* be able to still use your
forward kinematics rotation-based system, but I think it'll be quite
difficult to calculate the correct rotations, so it might be more work to
adapt it than to just scratch it and use transforms instead.

> Q's:
> i)  Do you absolutely require the foot to stay put, with
>     no rotation slippage, no matter how sharp the curve??

What do you mean by the word "require"? What requires what from what? It's
all just a matter of how you code it.

> ii)  How do you do yours?  The huge favor I'm requesting
>      is just the algo, probably no more detailed than above.

Let me know if the scene file I've posted is understandable enough. :)

> Apologies again for being so dense as to not understand your code.

No need to apologize - I should have commented the code in the first
place...

Rune
--
3D images and anims, include files, tutorials and more:
Rune's World:    http://rsj.mobilixnet.dk (updated June 26)
POV-Ray Users:   http://rsj.mobilixnet.dk/povrayusers/
POV-Ray Webring: http://webring.povray.co.uk


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.