[jep]

Takumi - DID YOU JUST REFER TO ICE?!?!?!?!?!?!

HAHAHA you just went up a few notches in my book

AWESOME! I learned how to composite on that package many moons ago - Nice reference!!!!

I totally agree, a nodal based material editor would be awesome - take a look at Shake - itt's the industry standard nodal based system. I think some of the things about hypershade are tougher than they need be - if we just stuck to the basic ICE model, we'd be SET!

This also brings up the question of a good curve editor (how else are you going to tweek the slope of your 24 hour day animation?) again - take a look at Houdini's and Shake's curve editors - top Notch (notice I didn't say Maya)

[sidenimjay]

Oh MY GOSH


Takumi Takahashi, u have used ice ????

i used that to do fx for a certain paramount space station tv show back in the day . . .

prisms rocked and that is the parent software to houdini.... which i think is the top of the game for fx, or anything one needs to do for that matter . . .

unlike maxwell im a tad biased . . . .narf

so im thinkin , houdini and maxwell and watch out world . . . .

procedural maxwell . . . . . yummy . . . .

go download a free apprentice copy of houdini 7 @ www.sidefx.com . . . .

[jep]

uh oh... don't even get Sid started on Houdini... it's a door you can't close!

I look forward to the day he has a proper Hoidini friendly copy of Maxwell - you all are in for a treat!

[sidenimjay]

yup hard to close a door on something you love . . . especially when its the best out there . . .sorry to all other 3d software . . .


im startin to feel the same about maxwell right about now . . . . .


mmmmmmmmmmmhmm



opinions are . . . .

[Takumi Takahashi]

Dear jep and sid,

SURE !! - ICE was my first compositing app (back in 93), but could not play with Prism itself that much, but I did a couple of experimental works on ICE.

Btw - I asked guys @ Alias for Maya's node based stuff before its inception by referring to SOPs concept of Prism, and I guess they were also looking at ICE for Hypershade (maybe)... since I worked at Alias for 8 years, I got a chance to hint them few times.

Conincidentally Greg's Sidefx and Rob (Burgess)'s Alias office were in both located Toronto, but not sure how close they are now since Alias went through bunch of changes..

sorry for the OT,

-tak

[Tyrone Marshall]

Oscar, just so we're not confusing out meanings, I'm not talking about opacity maps to apply on meshes in a physically incorrect way (making them fade to nothing) , I'm talking about material lapha blending for adding selective dirt, reflection or specular to base maps, adding color gradients to glass etc. that sort of thing. these are all real world effects.

Ah ok, then I think we were talking about different things. What you say is something that have to be done through the material editor from the stand alone, using maps to blend different layers into a new one material.

As we are developing the material editor, I invite all you to post your ideas about this or something related to it, in the wishlist, so that we can have in mind all your ideas ( I think my english is every day worse )

Best regards,

Oscar, your english is fine as your comments give us more pride and comfort. Comfort that we know without hesitance that Maxwell Render is a true revolution!

Thanks!

[oscarMaxwell]

As for the Material Editor, my idea would be to have a graphical node-based editor, where you can see textures, materials nodes as icons, so you can do blending operation (hopefully the blending node as a separate icon and selectable to see its properties and edit them ) so you can blend/layer materials visually.

My favorite examples would be Maya's HyperShade or ICE (an old compositing app for Prism). Here you can also see the connections between the nodes as arrows (selectable and deletable)... Also you can see each nodes having quick preview of its effects on each icon.

Well - just my idea.... hoping the first release of material editor would be some sort of WYSIWYG.

Is in our minds to do something like that. But a nodal based editor or some sort of WYSIWYG is very sophisticated for a first version. Our first interface and editor version will be modest, but we'll sophisticate it more as time goes on.

Best regards,

[tom]

thx, that would be great... i'm dreaming interfaces like simbiont
http://www.darksim.com/html/dt25_screen_shots.html

-- seamless tiling
-- looping procedures
-- fractal parameters

are very welcome!

[jep]

tom - cool link... that does look interesting!

[Mihai]

No no no......

You want node based materials, look at XSI's rendertree. It works so smoothly because the layout is well thought out. Elements can be expanded/contracted, they don't take up unnecessary space with fancy bevels and silly icons. Even complex trees are easy to oversee. Other things it shares with most node based IF's, like thumbnail preview of images. That would be nice to have in Maxwell, with a small text underneath the thumbnail specifying bit depth, type and size.

[tom]

yeah i agree mihai, there are several good examples... one is better than another always...
at the end, I believe M~R will have a different editor than all other because it's working with BSDF.

[Takumi Takahashi]

It is in our minds to do something like that. But a nodal based editor or some sort of WYSIWYG is very sophisticated for a first version. Our first interface and editor version will be modest, but we'll sophisticate it more as time goes on.

Thanks Oscar for your reply.

It is totally understandable - I just wanted to post ideas for your polling of Material editor. I don't expect the first version equipped with full-blown GUI like other 3dcg apps.

Instead I would rather ask you guys spend more time on speed (+quality) optimizations + mandatory feature sets of Maxwell Render, other than some fancy material editor. I am sure you guys come up with something great in a long run.

btw - I didn't mean to start a war on 3d app shoot outs!

[The Pixel Artist]

I think you guys have the right idea here. If I could make one suggestion in regards to the course to take on interfacing (and other), it would be that I hope NL and M~R follows a philosophy similar to the one the used by Luxology (with Modo). There are so many great software packages out there and they all have something they do better than the other. Unfortunately no one in the industry has yet figured out that what is really needed right now is a consolidation of all the best techniques into one great program. Luxology I do think is on the right path though and I believe you guys at NL are on that same path (if only you could merge, ah the thought ). I am curious what your long term plans are, I personally hope that they include a full modeling, animation, raster, etc. package.


Maybe what could be used here at least for now in the forum is a tread with several polls for what people think are the best interfaces for say rendering, materials, lighting setup, etc. My only reservation to this though would be that the majority users of one software will vote for there package even though they have no experience with another’s that maybe better. So I guess I'd ask that only experienced users of multiple software packages chime in.

[sidenimjay]

since i am a houdini user, i am used to the node based system for all aspect of cg animation



i find that for modelling, compositng, channel editing, particles, and animating, nodes work supremely


however, i have found that when it comes to shader writing . . . a simple constant or nosie or even basic shading is fine. . . but when you need to do a complicated nested for loop containing if/then's and 10000 nodes and wires . . . .you may just want to give up and write the code by hand. there is alot of overhead in a shader that is made up from a node based system.

i can handwrite the same code with about 5000 lines less code and get the same look . . . but in alot less time cause the code is cleaner . . .just gimme the functions i can use and/or let me make my own . . .

all lines of code and node counts are approximated . . .

[jep]

I just took Sid to the matte on this one - I feel that visual shader writing is exactly like compositing so a node based system is the way to go - Sid is worried that functions get to confusing visually - my point is, I'd rather be able to visualize my ops in a flowgraph rather than just be restriced to text based... His point is, that's all good until you start hitting a super complex shader that is doing displacement based on some if/then and while loops - I kind of understand where he's coming from, I mean I learned to write renderman shaders by hand and really wasn't happy with MTOR or MayaMan because the insert so much extra garbage - HOWEVER, I honestly don't feel like being restricted to a text only system... and I'm willing to bet many users on this forum agree... of course I'm willing to bet! that's why i'm going to Vegas Baby, yeah!

I also think that many of the overly complicated shader networks won't neccessarily pop-up as often since the materials are physically correct most of the extensive shader building would be to combine shaders not doing complex "if sample info >X then blah blah blah".

Again - just my opinion...