Dominance War Crit
Hey guys,
So I entered the Dominance War mainly to see if I could complete my model on time... so I joined cgchat which sucked because no one gave me any critiques due mainly to the fact that I chose not to do a humanoid character with glowing lights all over it... So I was hoping you guys wouldnt mind tearing me a new one and letting me know how I could've improved my entry.
I rushed my texture at the end and since this was my first normal mapped character I ran into many snags including my UVs which ended up being a disaster
So here is my final image.
and here is a link to my entry on dominancewar.com
Here!
Appreciate it!
Kicker
So I entered the Dominance War mainly to see if I could complete my model on time... so I joined cgchat which sucked because no one gave me any critiques due mainly to the fact that I chose not to do a humanoid character with glowing lights all over it... So I was hoping you guys wouldnt mind tearing me a new one and letting me know how I could've improved my entry.
I rushed my texture at the end and since this was my first normal mapped character I ran into many snags including my UVs which ended up being a disaster
So here is my final image.
and here is a link to my entry on dominancewar.com
Here!
Appreciate it!
Kicker
Replies
I think you could improve textures for this one. You've some wasted space in texture maps. Also green color of the monster is "too" green for my eyes and it gives a bit cartoonish feeling (but that is just my opinion
I made an example, a bit darker and "muddier" works better for me.
But I like it after all!
btw. is that polycount face it's ripping?
I can see what you're saying about the green and I definitely agree, he is very monochromatic and comes off looking cartoony.
and yes...he's destroying the polycount homeworld. Take that! It says CGCHAT in the middle of his head too..
cool stuff man, and congrats first and foremost on getting your entry completed. that's a victory a lot of people on all sides never managed--even if its only a personal victory, it doesn't diminish the glory. as an artist, a lot of your first responsibilities are to yourself and your own knowledge of where you are and where you want to be, where you see yourself going. so here we go...
unlike a lot of models that might suffer from one or two glaring omissions or faults, i think the trouble you've had on this submission is a series of smaller mistakes and oversights; it's death by a thousand cuts. which is a backhanded praise, of a sort--work through and correct all the little stuff and a fairly middling entry becomes a great deal more interesting.
starting with the concept (as all great designs start), you've actually got a bit better consistency and presentation than what ends up on the final model. how much did you iterate on this design, by the way? always take that little extra time on the concept--it will always pay off.
the body shape is more readable and the aquatic/fishlike aspect is more readily evident; the final model bears very few of these characteristics and looks far more humdrum sort of humanoid in shape, and the tail is almost completely lost in the scene/posing.
also better considered in the concept is a the use of color. using the green on green accents on the body really makes the spray of red across the head really stand out; in the final version, you've devalued red considerably in the scene by splashing it all over the place.
this is a valuable lesson about color, about highlights, or about anything you want to draw attention to: like an over-CG'd science fiction movie with special effects in every frame--if you're seeing it constantly, it's not terribly special then, is it?
so whereas the concept has a recognizable and interesting character to it, kind of a space Cthulu/Deep One thing going on with that split head, your finished version is a mess of color and especially red. put red everywhere and it attracts the eye equally across the composition, which is to say it no longer attracts the eye to any one specific point. which is bad.
so aside from the red runic tattoos on the body that would have been better in green, i agree that the color map for the body green could have been better managed, more variation... also in tandem with a normal map that defines features in starker relief.
other key problems with the model include some questionable polygon distribution in the head (what are all those polygons doing, exactly, on the interior of that head crest? what detail are they providing? push those polys to the silhouette, to define some detail) and most critically on the head, the lack of depth with the mouth. it might be how you've lit the scene, but the mouth appears to almost a normal map application with no real depth--i want to see those teeth popping out, or the interior of the mouth receding back to show me how horrible and nasty that set of chompers is.
you also look like you've got a pretty busy set of maps there. lot of spec info on top of the normals, plus emissives on top of it all? just like we talked about with color, if you put highlights or special doodads everywhere, all the effects cancel themselves out.
which is my principle issue with characters that have glow-y bits--it's a cliche and abused visual shorthand for drawing attention to different parts of a figure, and when overemployed really dulls the overall reaction.
as i've commented on other peoples' domwar entries, when designing, try to avoid your first inclination, since it got there from somewhere else, and is usually the most obvious solution. in this case, detailing out the body by adding glow-y runes... think about how often you've seen that before. probably can't count the number of times, eh?
on a smaller design question, the head tentacles can work, but i'm not quite convinced that they are working in the concept and certainly not in the finished model. whereas the rest of him as a certain aquatic streamlining to it (at least in the concept), these tentacles jut out of the posterior of the head at a seemingly 90 degree angle. in the concept you've got more of them and they seem to be more sensibly arranged, but in the model with the reduced count and poor articulation, they're pretty lifeless.
and given that the viewer is going to spend much of his or her time looking at a character's face when it comes to character art, you've got a very ugly feature right next to a prime area of interest.
on a related note, the projections (tentacles? spikes?) from his back are doing a whole lot of nothing, other than distracting and confusing the scene. once again, if you repeat a visual element too much, it becomes devalued--you have head tentacles and then back projections both. in the concept, there appear no such superfluous protuberances, and it's a stronger looking character for it. maybe if you hadn't put the back bits on, you'd have enough polys to make the proper amount of head tentacles?
finally, we'll talk a little about how those other issues are compounded by a poor presentation/composition for the final. take a look at some of the winning pieces and, other than being vibrant and interesting characters that are well realized generally, they usually are well composed in their scenes as well. and if all else fails, just go with a simple flat color or basic gradient background--the last thing you want to do is overshadow, complicate, or stymie the presentation of your character in any way. if it's not reinforcing the character, ditch it.
for starters, the background: big glow-y nebula is a good idea, and this one looks good, but maybe not the best choice for this character. particularly with the lush reds and purples it adds to the scene, you're getting a whole lot of bright colors that spread the interest a bit too generally throughout the picture, fighting against the character. you don't want that, this is all about the character like we keep coming back to.
and as it's all about the character, we come to the execution of the pose on top of that poor ol' Greentooth. now, i give you some points for trying, but after thumbnailing out this scenario, you probably should have given up including it entirely--not everyone (and certainly not the many of the judges) are going to realize they're looking at a giant Greentooth that the character is tearing into... if you wanted to show this character's primal dominance over some hapless, lesser form, why not show him goring up some generic soldier or smaller form, like Goya's Saturn devouring his children?
you've got to consider what sort of context in which the viewer is encountering the piece.
some of the judges may have been familiar with the history and therefore the high visibility of the Greentooth smiley would make sense to them as a symbol for polycount, but i'm sure it made no sense at all to others. especially in a contest like this, where you might only get a few seconds of the judge's eyes resting on your piece before something in their gut or brain says idly, "what is that, is he murdering a giant space potato? a big fleshy asteroid or something?" and they click away and onto the next piece, and they're gone. they won't give your piece a second look.
this is not to say the winning gambit is to festoon your character with garish baubles and cheap tricks (though it's worked for some
quick rundown of what's going wrong in the various departments--
placement of the character in the scene: you're running afoul what my old art teacher would call "x-compositionalism", you basically have the character sitting dead-center in the frame. this only works for very specific types of compositions, and this isn't one of them. there is no dynamism, no motion through the frame. if you're not familiar with the "rule of thirds," look it up--divide the picture plane in to thirds horizontally and vertically; place areas of interest at the intersection of these points. later, good compositional habits will create a more intuitive sense for placement and pose, but these are capacities that must be nurtured and developed through good practices--don't feed bad habits. if you don't know good composition, practicing what you (don't) know won't help you.
character pose: related to the placement issues, you've got one arm up and one arm down, which creates a diagonal element--which is good. diagonal elements can help create a lot of motion in a scene. unfortunately for this space baddie, he's standing smack-dab in the center of the frame, which means those diagonal-creating arms of his are effectively just helping lock him dead in the middle. instead of adding to motion, they're actually taking away from it. no amount of motion blur or suggestion of motion from the blood spatters is going to help an essentially static composition like that.
and as i've talked about before, the pose and modeling have de-emphasized the interesting, more aquatic aspects to the design, so we can't really see the tail and the pose just isn't highlighting much of the character at all. it does help show the head crest area and multiple eyes well, but the lower body is murky and also partially obscured by the blood splatter. we didn't come for the effects, we came to check this guy out, so don't make us work to get a look at him.
a final note on color: red and green are complementary colors--opposing colors on the color wheel. which means if used strategically next to each other, they can produce some pretty intense visuals, but poorly applied, (once again) they cancel each other out. there is so much green green and red red in the frame that neither one really wins out, and the picture reads as surprisingly muddled. it shouldn't--you take the picture and desaturate it and you'll see a reasonably decent range of values, but values and contrast aren't where you troubles are, they are in the composition and in the colors.
alright, i think that should give you enough to chew on for a while. take some time to get to know polycount, you might find yourself staying
I really appreciate you taking the time to give me an honest critique of my work. I wish the guys over at cgchat would have taken the time to critique my work during my design process, though the only person I have to blame is myself and my inexperience. I learned a lot in the comp and will use what I've gained in my future projects and look forward to becoming a member of this community.
I think Gauss covered everything in his post. The Dom wars are a great learning experience for everyone who enters, regardless of skill to begin with. I learned more in the few weeks I was working on my dom war entry than I did in 6 months prior to it. Nice stuff and congrats on completeing your model