PDA

View Full Version : future of 3d app u.i. and methods of working



cresshead
01-24-2005, 03:55 PM
hi

how's this for a "thought of the day"....

i've played/used quite a few 3d apps and a number of plugins for various apps such as 3dsmax and lightwave..plus recently xsi...

now where are we heading in the nr future of say 2-3 years ahead...
there seems to be 2 trains of thoughts on implementing new and more advanced apps or their plugins...

take these two as an example...

1.lightscape 3.2
2.final render for 3ds max [gi renderer]

both of these apps set out to get nice and realisic renders and discreet even tried to incorporate some of lightscape into max 5 and 6...

lightscape now even though this is a now "retired" 3d renderer it has some unique capabilities and some rather surprising cool things to offer users...
the main cool thing i like about it is the render wizard...you can if you like set up the renderer by hand or you can simpley answer a few yes/no questions and lightscape will set up the render settings for you...it asks if the scene is lit from only sunlight thru openings for example...once the wizard is answered you can hit calc and see your lighting grow in the viewport....
another nice feature is you can select eithe know ies lighting models OR generic lights such as a 100w bulb or a 4ft striplight...it's nice to have a simple choice sometimes...

final render approaches the challenge of photoreal renders from another angle...giving you a "under the hood" feel to the render panels with about 150 dialogue boxes or entry ranges from 1-100..feel more like a nasa style approach where you really need to understand what all the gizmos are up to and how they relate to one another...not very friendly but quite capable in the right hands..


so where are we heading?

will future plugins/apps just get more and more complicated to use learn or will the underlying tech soon get a nice "frontend" that let's us get on with creating cool 3d scenes and character animations?

i'd really like to see a move toward some nice front end and maybe wizard style implementations to new apps or plugins but they seem not to be a hi priority to most vendors right now..

most simultion plugins say they are for instance, a "cloth simulation" but in essence they are a spring based polygon deformation simulation that if you get the settings just right give the appearence of cloth...so really they are not from the ground up actually trying to simulate cloth..it's just a possible by-product of some of the settings that the plugin can reward you with a cloth like result...there is a subtle difference if feel that means that the plugin and u.i. can be quite a learning curve to new users as they would be looking for things like a type of cloth to load or a stretch limit to keep the sim from going out of bounds of what most cloth will do...

i hope that someday we are in the same position to say where we are in modeler in lightwave...we have points/ poly and we just move them and whatever around as we's expect to...a huge gap to where this started back in tron days with graph paper and entering co ordinates into a database to build up a mesh with numeric entries of vertex locations...

right now we're still in that tron stage to a great degree with rendering/lighting and simulations and even to a point bones and rigging though they are at least mostly predictabel/controlable in most apps nowdays.


let's hope that kray tracing and fprime not to mention lightwave's renderer continue on a quest for better images but not at the expense of usability...

in the end i want to make "pretty pictures" and not become a "lab coat technician" which is where such apps as mental ray and xsi seem to be headed off toward with ever more complicated things to learn...

my view is that the app shold do the thinking and not rely on the user to dip his/her head into a mountain of tech books just to get a nice picture out..........

...comments?

cresshead
01-24-2005, 05:44 PM
hey maybe i'm talking to a crowd of lab coated 3d technicians!....opps!

seriously though i don't want to "dumb it down" just make it user friendly, as currently the more i look into such program as xsi the less i'm liking doing 3d in that app at all....i'm not propsing a poser/bryce but just give users more choice with either a simple u.i. option as well as the "under the hood" option...

currently some things would be like this analogy...

you want the car to go faster....

either....press your foot on the gas pedal
or...lift the hood up and adjust the petrol/air mix valves to enrich the mixture....

both get the job done....one is a the simple method the other a more refined way.

can we have both please! :D

Gui Lo
01-24-2005, 06:03 PM
I agree Cresshead.
Recently I have begun learning another app that is thought of as high, high end. Because of this I thought it would be quite easy since I have about ten years experience in 3d. As it happens I had difficulties as the workflow has a high degree of precision which needs to be adhered to for decent results.

This seemed backward to me since the the app is still very expensive compared to current market trends. I thought that the higher the degree of sophistication the easier it would be to generate good looking 3d. Also the better the app the more easy it should be to do things that are difficult or fiddly in other 'lower' apps. It is mostly the opposite that happens, with lots of new options are added so apparently the user has more "control".

For the future what I would like to see is more development for things like IK Boost where the app knows what the user wants to do with IK and enables them to do it. Also rigging is a very basic "Tron" system. For most purposes a bone has a joint action rather than a generic deformer so we should have tools that produce body joints.

For the more general aspects I love the developement of more user control over things like selection of edges, loops or whatever. I hope more are added like selecting connected surface. Also more developement needs to be put into simple navigation so that the user can select the point or poly they want easily. This includes adding phantom point recognition intogenerating weight maps and vector maps.

As a 3d app like Lightwave develops it should be easier to learn and use and the high end option should come lower down so that the high end keeps getting pushed higher. With the recent developements we have seen this, but of course we want more.

Gui Lo

policarpo
01-25-2005, 08:15 AM
Well I think the issue is larger than that. As we progress into the future, we will gain a more emotional attachment to the software we use. Software will need to connect with us on a technical as well visual level.

I think we are quickly approaching a balancing of technologies and features in the 3D market place. All the current tools offer pretty much the same features. We can get the work done in any of the chosen top 5 apps in my opinion. While some allow us a quicker turn around, the solution is there if we bother to learn it.

What the future of 3D is about, is about how extensible the 3D app is. How quickly can the developer tack on new features which feel integrated (not slapped on as we experience with 3dsmax and LW). It's about designing a modern core that is adaptable, robust, and geared toward modern workflows.

I recently had an opportunity to give modo a whirl, and the experience I got from it was something bordering on blissful. The experience of the tools and the experience of the UI made me feel empowered to do my work. It reminded me of how modeler felt when I first began moving vertices around in version 6.

I have always believed that tools are tools, but as I look into the future, I realize more and more that as I invest my time and intellect with my tools, I want a some certainty that my software is growing alongside me and not fighting me as I mature as an artist. At the end of the day, what matters is the art, but it's good to know that the software has you, the artist, at the forefront of its intent.

Cheers.

Nemoid
01-25-2005, 09:47 AM
Agree totally on this statement : i think that in the future, software will be more and mor with the artist in mind, not always the "technician."
coming from a traditional approach to art, i often feel myself wanting to obtain results that i clear see in my mind, but sometimes tools are obstacles and require way more work than a normal artist could think. artist in mind, is a key point on 3d and 2d software developing, the more the artist is connected to the app, and the more the app and technology allows to use natural gestures to get the job done, the better art will be.

or at least we'll use waay less time to produce crap!! :D

modo seems the right step in this diretion, but it's approach is still classical, but most important, even with great advantages not so different from modeler right now. we'll see in the near future.

z brush seems to me more innovative and good for a ytraditional artist. right now is awonderful companion for every 3d or 2d artist indeed, working good with different apps too.

i really hope it will evolve greatly even for animation purposes.

the same could be for Lw. it could be greatly enhanced, and become a really powerful and user friendly app, :)
with a very serious and i'd say long work of development with the artist in mind.

just innovate.

Rory_L
01-25-2005, 09:07 PM
in the end i want to make "pretty pictures" and not become a "lab coat technician"
I so agree! In recent years (let`s not talk about 3DS4!) I`ve experienced working with both MAYA and Lightwave 3D. As the two apps stand at the moment MAYA is winning in the Lab technician friendly stakes and LW is way up in the warm and fuzzy league. Both apps need to learn from each other.
All the main 3D applications should develop towards having robust, black box style plug `n` play functionality, with the option to delve inside those boxes when necessary.

There`s a thread (started by Cresshead) on these boards currently arguing about nodal and layer based surface editors. People`s comments are circling this thread`s topic, so I thought I`d mention it: you might want to cross-refer.

Cheers,

R