EnigmA Amiga Run 1997 July/EnigmA AMIGA RUN 20 (1997)(G.R. Edizioni)(IT)[!][issue 1997-07 & 08][EAR-CD IV].iso/lightwave/lwmlist/96.lightwave-0422/[e-mail] _Fri Apr 26 13:20:30 1996.msg

Internet Message Format | 1996-05-03 | 4KB

Received: from e55.webcom.com (e55.webcom.com [206.2.192.66]) by keeper.albany.net (8.7.5/8.7.5-MZ) with ESMTP id NAA10887 for <[e-mail]>; Fri, 26 Apr 1996 13:20:27 -0400 (EDT)
Received: from localhost by e55.webcom.com with SMTP
    (1.37.109.15/16.2) id AA199389142; Fri, 26 Apr 1996 10:19:02 -0700
Date: Fri, 26 Apr 1996 10:19:02 -0700
Errors-To: [e-mail]
Message-Id: <[e-mail]>
Errors-To: [e-mail]
Reply-To: [e-mail]
Originator: [e-mail]
Sender: [e-mail]
Precedence: bulk
From: [e-mail]
To: Multiple recipients of list <[e-mail]>
Subject: Re: moving balls on water
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Status: RO
X-Status:

(Read the whole thing, it evolves....)

On Fri, 26 Apr 1996, [e-mail] (Gonzalo Garramuno) wrote:
>[e-mail] writes:
>
>>     2)  Why does it work?
>
>Basically, you have to understand
>that every 3D program in the market has a priority for deformations
>(they usally never tell you, but they do).  So, certain deformations
>take place before another one (if done carefully, deformations will do
>what you expect them to do... although there can be problems).
>
>Therefore, from this example, you have learnt that LW performs
>displacement maps AFTER morphs, which is logical (if it were the other
>way, you would not be able to perform the effect).

    Wait.  Isn't it:

    1)  Displace very small box points(add displacement vectors), then
    2) morph them into place ?

    Because if it morphed first, the box would ripple, no?

    I _think_ (vaguely) you got it backwards.  My bwain hurtz.

    Now, a ripple applied to a morph target will not affect the morphing of
a morphed source, so that would imply that displacement DOES follow morphing, as
you say.  Actually, I remember having a similar discussion a couple of years
ago, and it is probably saved >somewhere<.  Allen explicitly outlined the order
of deformation. Possibilities:

    Morph-->Displacement-->Bone
        Seems like this must be the way, since neither ripple nor
        bone displacement affects morphs.

    Displacement-->Morph-->Bone
        This is how my confused mind thinks it must be for the beach
        ball trick to work (and it does), but contradicts the other
        fact.  <See below for elaboration + correction)

>Anyway, displacement maps need to know where the original points lie so
>that the program knows how high or low will each point move.  In this
>case, the original points (although already morphed to a ball shape)
>were originally in the origin.  The displacement map takes that
>information to calculate the displacement, so it knows that the
>displacement will be the same to all points (since they are all together).
>However, when the displacement is ACTUALLY performed, the current point
>positions are transformed (this is correct, so that any previous
>deformations are not destroyed).  So, the points that are in a ball
>position move up/down according to the displacement values, which
>is the same value for all points.

    AHA!  This is the crux:  The displacement vectors are CALCULATED using
the original points, but are APPLIED to the morphed points.  I find this rather
counter-intuitive, but possibly very useful.  This is the level of detail one
rarely finds in manuals BTW.  Did I miss it?  (page num?)


>Gonzalo Garramuno

    Thanks Gonzalo!  Informative as ever!  (Well, except when you contradict
Ernie....>8^)


******************************************************************************
** [e-mail]     |     Synergy Graphix & Animation -- Seattle     **