Light ROM 4/Light ROM 4 - Disc 1.iso/text/maillist/1995/121595.doc/[e-mail] _Mon Dec 18 17:34:05 1995.msg

Internet Message Format | 1996-01-16 | 4KB

Received: from relay1.UU.NET (relay1.UU.NET [192.48.96.5]) by keeper.albany.net (8.7.1/8.7.1) with ESMTP id RAA21064 for <[e-mail]>; Mon, 18 Dec 1995 17:34:04 -0500 (EST)
Received: from garcia.com by relay1.UU.NET with SMTP
    id QQzurx06008; Mon, 18 Dec 1995 17:26:37 -0500 (EST)
Received: from  (localhost) by garcia.com (5.x/SMI-SVR4)
    id AA24175; Mon, 18 Dec 1995 17:25:19 -0500
Date: Mon, 18 Dec 1995 17:25:19 -0500
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: Andrew Hofman <[e-mail]>
To: Multiple recipients of list <[e-mail]>
Subject: Re: Droplets
X-Listprocessor-Version: 6.0c -- ListProcessor by Anastasios Kotsikonas
Status: RO
X-Status:

First I wrote:

>>Using Surface Morph and a
>>repeating envelope, why not have each ripple area morph back to the larger
>>puddle surface in between ripples? That way the ripple areas don't stand out
>>when they are quiet.
>

Then you replied:

>It is necessary that you design your image sequences and surfaces such that
>they don't stand out anyway (other than the ripples themselves). So the
>morph is solving a problem that you should not have in the first place.
>

My assumption was that you'd have two things going on in this puddle: (1)
Raindrop ripples, and (2) General ripples from all this activity in the
puddle, lake, etc.

This is why, to keep things under control, I would probably render the
droplet ripple sequence on an otherwise perfectly flat surface. This allows
me to mix textures with more predictable results. (If I added general
turbulence to the rings, a mismatch with the overall puddle would be
assured.) Then in the final conglomerate object I would partially morph
surfaces in each droplet area. This would permit a mix of both droplet and
overall ripples. In between droplet activity for a given spot, the general
turbulence would be ramped up to 100%. This also means you don't have to
have the droplet sequence instantly resuming the moment it finishes -- with
morph envelopes you can just ramp it out for a bit longer to get more
apparently random droplet hits. And you can vary the duration of this for
each individual droplet area, to further mess up repetitive cues. If
pseudo-random sequences repeat too quickly, the eye can easily perceive the
"pseudo" part.


Then I wrote:

>>Or, greatly exaggerate the ripple depth and then only
>>morph 50% into it.
>

And you replied:

>You could do the same thing more efficiently by altering the bump amplitude.
>The only reason for a morph would be to alter the bump amplitude over time,
>which could ease the effort of making your puddle rings dissipate.
>

Altering bump amplitude is exactly what I meant, but in retrospect it could
have been read another way. My reason for even doing this ties in with the
method I described above. I'd want to have general turbulence occurring in
droplet areas, at the same time droplet rings are also forming. Otherwise,
the result would be quiescent areas with perfect little rings spreading in
the midst of a turbulent puddle. That kind of stuff is a dead 3D giveaway.
But if you're going to have a mix of these two things, it's going to cut
down on the amplitude of each in a 50-50 morph. Solution? Double your
amplitiudes and in the morph they'll each end up being 100%.


   ************************************************
   * Andy Hofman       |    LumaQuest Productions *
   * [e-mail]  |             3D Animation *
   * 513-643-7333      |  Dual Raptor 3 Rendering *
   ************************************************