DomWar 2 – A Very Large Dump!

Wow, I have a website?! And nobody told me I need to keep it updated? (Oh wait, I was told, but too busy to do much about it).
And so there was Dominance War this year again and for some odd reasons I decided to participate, even though fantasy or sci-fi fighters are not my cup of tea, and neither is the whole background for the characters. Luckily, I have teamed up with a great concept artist, Chris De Joya, aka Snowfly, who came up with a concept for my entry. Here are some of the sketches he made for me:

And then, here are my final renders that were submitted:

I didn’t really get the chance to polish the presentation or even set up some good lighting, but I hope to revisit some of the textures and render out a turntable, since there is an iridescent shader on the shell and wings that doesn’t show really well in the stills. I will also post some close-ups when I render them out (or hopefully, when I learn how to write CgFX shader so that I can show a viewport grab). The main problem I had with final presentation is that hardly any details showed up, because the image is so small (800×900 pixels was the upper limit for all the submitted images), and also because there wasn’t any closeup that would really show all the detailing that went into the texture work, or even some of the smaller normal map details are still barely visible.

I might as well post my work in progress images, to show my high poly and the entire process I went through until I reached the stages captured in images above. So here goes…

Base female mesh done in Maya and Silo:

At this time, Snowfly came up with the nice shape for the mount, and was kind enough to create a nice side view sketch so that it was easy for me to model it (it was my first animal model, if I don’t count Mrs. Boaring who wasn’t fortunate enough to have her texture finished). Here I show how I modify the shape of the beast to add more bison-like features, and finally add wings (or at least placeholders for wings):

We are struggling with design at this stage. The initial idea was to do a beetle type character, then I suggested adding a rider (a big mistake! it cost me many extra hours of work, since I essentially worked on two characters instead of just one). So here is my paint over of my model in perspective, and a rider, mainly based on previous Snowfly’s sketches and some of the ideas we were throwing at each other:

Until we decide on the final design, I start working on the boot:

Then I create high poly in Mudbox:

Then there is pretty much final design for the beast, and I go ahead and add more detail to my initial model, and it looks something like this:

The next step was to create a quick rig (it’s not even a rig, just simple skeleton), and see how the two characters fit together:

I start adding bits and pieces to the base female model of the rider:

I take the arm geometry to Mudbox, do some sculpting, and add some details in Maya (the ornaments around the shoulder pad):

I then go back to Silo and try to finalize the medium res model of the female, but I have much troubles with the helmet. Colleague at work told me she looks too friendly with round shape on her head, as if she just begs to be patted, like those cute little animals Sonic takes care of – and he was completely right. But yeah, at one stage my rider looked like this:

I know for sure I have to do something about the helmet, so I look back at Snowfly’s sketches and try to find bits from all of them that would fit together and make the character look more menacing (Polycount’s characters are supposed to be evil!):

The final sketch of the rider had those things attached on the hips and they seemed to add some nice interest to the silhouette, so I decided to add them in. I take all that to Mudbox again, and do some more sculpting:

I add a few more details here and there, and call the high poly for the rider done:

Then I take the mount into Mudbox:

Still some more definition is needed, but I decide to move on to low poly, and here is the rider with messed up AO map, courtesy of their Slowness, mental ray:

At the same time I baked the colors assigned to high poly, to make isolating materials on texture sheet much easier:

I then add head gear to the mount and add some details to the wings at last:

Finally done with low poly and also finished the high poly of the rest of the gear (saddle and bags). And then, xNormal comes to the rescue, just at the moment I was about to go kill mental ray! I was so happy that this program worked great for me – thank you so much Santiago Orgaz, Sir! This is what normal map + ao in color channel looks like (pretend you didn’t see it in construction shot at the top):

I realize my characters are too large and are hard to fit in the frame so that all the detailing is actually visible. That’s the reason I grab some close-up shots to show:

I then start adding some colors:

All the areas have colors now, although some modifications of the hues are to follow:

I add iridescent shader that is alpha masked, so it only shows up on the shell parts. Need to add something similar to the wings as well:

Created new skeletons and trying out some poses:

Contemplating the shot from behind, and also added the rock to act as a pedestal, quickly running out of time:

It is already Sunday night and I’m out of energy. I post this render in my progress thread on Polycount:

I receive handful of good crits, but I realize they mostly address issues I already worked on and there comes painful realization that most of that work DID NOT actually show up in the render, since the details were so small and mashed with all the filtering and post processing. For instance, there was a suggestion to add iridescence to the wings, and in fact spec for the wings was gray, color was yellowish, and the colors seen in the render came from iridescent shader, but alas, it was a still shot and nobody could tell where the green and blue in the wings come from. I’m figuring that iridescence shows only from sharp angles, just like specular, so I have to find another angle for the final shots. And it’s already Monday night and I’m out of time, and it takes forever to render. I submit my stuff, sigh, and feel relieved but very crestfallen.

So in the end, I think it was great to participate in another challenge and make something, but I also feel that the whole contest was missing the point by having us make “real-time” (e.g. game specs) models that are intended to only be seen in one small render and eventually thrown in portfolio to catch dust. Another thing that surprises me about my work is that for some reason I had so much energy to put into this, and yet I knew nothing about these characters, or what do they do and what are they like, and I didn’t really even care about them, because they were so devoid of any meaning and personality for me, so it comes as a surprise to me that the craft was quite enjoyable without any background to give meaning to final product. And that is all beside the dubious setting that these characters come from: a war 1100 years from now in which the glorious forces (pardon – shadows) representing the empires of our favorite bulletin boards are fighting… for no particular reason (?!?). Now that I think of it, I’m really out of place in that war (and perhaps most other women). And so are my characters…
So the next thing I make will by all means end up in the game. And that’s final, YEAH!


One Comment

  1. hi .. i like your texturing skill very much … since i am not a texture person so i will really love to know how can the character keep in a low polly count but have the high Res. outlook … maybe we can exchange some skill .. since i am much more better on character Rigging heh … once again … owsome work you’ve done hehe

Leave a Reply

Your email address will not be published. Required fields are marked *