Maya Particles.. My thoughts. 2005-05-05 - By Andre DeAngelis
Back
Brad,
I have to disagree with you.
In spite of the fact that Maya particles may require expressions to
control seemingly simple parameters, I consider Maya particles to be
much easier to get to look good from the get go - even the default
parameters give nicer results.
For example, create a sphere and make is a particle emitter, and leave
the default. Do this in both apps and see the difference. The version
in Maya looks like a stream of particles, the version in XSI looks like
a bunch of dots. From here, it just gets worse.
Good controllable forces and sub-frame sampling (and flexible noise
params) are at the core of particles, and Maya does both of these much
better. These are what produce organic behaviour, which is by and large
what we usually try to achieve. Another example is Particle Flow in
MAX, which also puts XSI particles to shame. The control in excellent,
as are the forces, the noise controls and the sub-frame sampling.
Without these controls, you end up fighting to avoid what otherwise
looks like a bunch of dots emanating mechanically from a source.
The first time I delved into Maya particles, I was able to get a sim I
liked almost immediately. Conversely, it takes me (and I dare say
others) a lot of tweaking in XSI to get to something we consider
acceptable. Now, there are some who will argue that "you just have to
know how to tweak it", but that old argument was used by Maya zealouts
for years in defense of their appalling renderer, before they finally
got Mental Ray.
>From what I have observed, the community appears to be in agreement as
to what they want from particles. An open node based system along the
lines of what is currently available in Houdini or Max.
What I, and some of my colleagues, find disconcerting however, is that
for years now, we have been very proactive in not only providing
feedback to Softimage, but also providing examples, demonstration, specs
and even DVD's/tutorials to Softimage to illustrate what we want/need.
Sadly, very little has come of it.
AD
-- --Original Message-- --
From: owner-xsi@(protected) [mailto:owner-xsi@(protected)] On Behalf
Of Bradley R. Gabe
Sent: Thursday, May 05, 2005 1:25 PM
To: XSI@(protected)
Subject: RE: Maya Particles.. My thoughts.
The impression I've been getting is that if you are non technical, you
can
get further with XSI particles, mainly due to the ability to treat each
point of the cloud as a typical vertex point.
However, if you are technically savvy, you can get much further with
Maya
particles due to the openness of the system to standard expressions. And
the reality is, particles and effects dynamics are often so need
specific,
they require constant customization. Thus the continued popularity of
Maya
over XSI for effects work.
I'm pretty scripting savvy, and I've had a tough time dealing with XSI
particles and trying to access the cloud to influence forces and
particle
attributes. Though at this point, I doubt Softimage needs to hear any
more
complaints about the particles. Seems as though we all have different
ideas
and suggestions on what can be done to improve them, the barrage of
which
must seem like an awful lot of noise. It may be more constructive for
the
XSI community to come to some kind of consensus on what we'd like to
have
particle-wise so they don't just keep patching on band aids to the
current
system.
-Brad
>Let's not forget that apart from forces, what makes Maya's particles
>"work" is the sampling, which is way ahead of XSI.
>
>Also, from a scripting/SDK standpoint, Maya's particles are
considerably
>more open.
>
>AD
>
>-- --Original Message-- --
>From: owner-xsi@(protected) [mailto:owner-xsi@(protected)] On
Behalf
>Of Matt Hollingsworth
>Sent: Thursday, May 05, 2005 12:26 PM
>To: XSI@(protected)
>Subject: Re: Maya Particles.. My thoughts.
>
>Hey,
>
>I've been doing particles for the Fantastic Four film here at Stan
>Winston, using XSI, and doing Maya particles at home, so doing both at
>the same time. So, some thoughts on this as well:
>
>There actually are a few things that XSI can do that Maya cannot;
>delete individual points in the cloud, deform clusters of points, use
>shape animation on clouds if they live forever. Maya can deform the
>entire cloud, but not a section of the cloud. No shape anim. No inter
>particle collision without trickery. But, yes, Maya's Fields (same as
>Forces) are way way better. XSI's turbulence is directional and is
>really wind. Maya's is actually just a noise function, but it works
>really well. Expressions are indeed needed, but they are very easy
to
>
>use. For instance, on a turbulence field, youc an animate the movement
>of the turbulence in the PhaseX, PhaseY channel. Like: PhaseX = 3.3 *
>time; Simple. Also, it makes it easy to script individual particles
>attributes suing other attributes. So, say you want the mass of an
>individual particle to increase with it's age, jsut write "mass = age;"
>
>Pretty straight forward. And, using ramps (grads) to define the
opacity
>
>and color of an individual particle based on it's age is pretty cool.
>Most of this stuff you can do in XSI, but the fact is that there just
>isn't quite as much power or control over forces in XSI, and that ends
>up hurting a lot. In Maya you can apply a "Newton" field to a particle
>cloud, and have it repel or attrach another cloud, ona per particle
>basis if you wanna wait around for it to compute.
>
>Anybody ever play around with Houdini?
>
>Now, don't attack me. I am one of those fairly rare individuals who
is
>
>neutrel in this whole XSI versus Maya debate. I jsut like both for
it's
>
>own strengths.
>
>Eric, share any other experiences you may have.
>
>cheers,
>
>-M
>
>Eric Lampi wrote:
>
> >So I've been working for a couple of days in Maya,
> >trying to figure out it's particle system.
> >
> >In a nutshell, it sucks. There are some nice things
> >about it. The playback is super fast, since it mostly
> >operates in what we would call "live" mode. Their's
> >actually works, where as in XSI, well, rarely works
> >well if at all, and when it does it's not much faster
> >anyway.
> >
> >They also have some issues with shaders, some things
> >need to be rendered in Hardware.. or software.. It's
> >not all that clear to me.
> >
> >The forces at first seem to work quite well right
> >away. Often in XSI you apply a force, and it doesn't
> >do much, or the results are not very good.
> >
> >I was amused to see that in order to add variance to
> >any value, you need to add it as an expression, which
> >is an incredible waste of time. In general you seem
> >to have to have a decent scripting background even for
> >the most basic functions. Overall it's been a
> >frustrating experience, and Maya in general seems to
> >be kind of messy. Navigating your emitters forces etc
> >is much easier. For example, it's simply a matter of
> >highlighting a force in the list to apply it to an
> >emitter.
> >
> >It's nice that you can script functions and
> >attributes, but I think it's pretty lame to make it
> >necessary for really basic functions. Which seems to
> >be the theme in Maya.
> >
> >So in general, XSI simply needs to improve workflow,
> >speed, and forces. There are some additonal things I
> >would address, but those are the biggest as far as I
> >can see.
> >
> >I was pretty impressed with the ability to select an
> >object, add a turbulence force, press play and watch
> >it drift away. Which is probably the only thing I can
> >manage to do in 3 clicks in Maya!
> >
> >E
> >
> >Freelance 3-D Animator, F/X Artist, Particle Man
> >---
> >Unsubscribe? Mail Majordomo@(protected) with the following text in
>body:
> >unsubscribe xsi
> >
> >
> >
>
>---
>Unsubscribe? Mail Majordomo@(protected) with the following text in
>body:
>unsubscribe xsi
>
>---
>Unsubscribe? Mail Majordomo@(protected) with the following text in
body:
>unsubscribe xsi
---
Unsubscribe? Mail Majordomo@(protected) with the following text in
body:
unsubscribe xsi
---
Unsubscribe? Mail Majordomo@(protected) with the following text in body:
unsubscribe xsi
|
|
