I was planning to job-apply more actively after finishing up my latest piece. I took advice from last time, and cut down the number of pieces in the folio to 3-4 pieces, but I think I gotta present them better to make em pop and shine.
http://www.pyr3d.com/
The Facebook link also goes to my art page, not my personal page. (it's very new, though :P almost empty atm.)
If I want to do game characters as well as 3d print stuff, then should I have sculpt pieces in there as well ?
I notice Carbonmade isn't so user-friendly that ppl can save images or pin them online. Is that a big deal breaker ?
Replies
As for the work, I think the biggest thing that sticks out to me is the texturing and material definition. The wizard for example has 2k and 4k maps, but I cannot see why you would need that resolution for what you are showing. Also the textures themselves do not seem to be using the PBR setup in marmoset to the fullest advantage, you are getting some really muddy shadows, and flat looking metal. Can you post the maps for him?
But alas, something to keep in mind.
Anything else I can get comments on ?
Untextured projects always feel unfinished. Resist the temptation to post works-in-progress and be aggressive about culling the weaker projects over time.
Overall, your textures could use the most work. They're missing many of the finer details that have become common over the last several years (since the PS3/Xbox 360 days, really). Cloth should have distinct textures, metal should shine, and glass should be partially reflective and transparent. You did a bit better on "four-armed", but the material definition on the non-cloth portions is lacking.
On a similar vein, you can't get away with solid clumps of hair anymore except for stylized designs. Hair - and feathers - need textures that appear to show individual strands.
Your hands could use some work. Looking closely, I think part of the problem is that they appear to be simply cylinders with rounded ends. It varies from person-to-person, but overall the fingers taper slightly from base to tip, with distinct bulges at each knuckle when viewed from above. When viewed from the side, the top appears mostly straight, but the bottom has a distinct taper. I'd strongly recommend doing a project that focuses entirely upon hands; I think something like Escher's hands would be interesting in 3D:
Whenever you use somebody's concept, make sure you include an actual link to the project. Just mentioning the artist's name makes things too difficult for the viewer.
I'm not able to tell from your about section what you want to do. Most of the experience isn't really relevant to an art-related job, and your skills don't point in a specific direction, either - there's no mention of any game engines, for example. I'm personally rather terrible at talking about myself, but it is useful if you can tell people about yourself and what you want to do.
The Kitten, for example, there's no clear definition, especially in specularity, between the metal and the fabric. It overal feels flat, which feels weird still even with the anime-style it is going for.
Really take the time to polish your existing characters textures.
Also, here's the clothing texture. The layout isn't great because I cut out some pieces last minute <_<" But yeah, I think the Spec map is too flat...
@DWalker Thanks man. Got it. I will only post WIP on my FB or such, but not on my folio. Hair strands " Feels like a shame, now that I think I start to sculpt better hair in ZB. Oh well, there's still 3d print I can do that. I will work more on that, and on the hands too.
Note to self, post texture map for other projects too.
Thanks a lot man. I will update the about page as well.
Well, most likely I gotta look for a new folio site since this one doesn't let me save img or pin stuff...
@JadeEyePanda Thanks man. I will work more on material definition now. When I'm not sure, I'm gonna exaggerate it and see how it goes.
I would highly recommend reading these posts by marmoset, and really understand the information.
http://www.marmoset.co/toolbag/learn/pbr-theory
http://www.marmoset.co/toolbag/learn/pbr-practice
Then from there I would recommend experimenting with some small swatches of different materials on some cubes/spheres. Use the chart on marmoset that gives rough albedo, reflectivity, and roughness values to starting nailing down your material definition. Stick 100% to real world materials, and trying to create them as realistic as you can at first. This should help get you started with pushing your textures to the next level. Then from there you can push the stylization as far as you want.
Question: So, this chart is not the new PBR then ? (Roughness + Metalness). Since Metalness is grayscale, but this chart has Reflectivity with color. This means I have to use Spec + Gloss for this reflective material ?
You just tag things as metal (white) and non-metal (Black) and use the time you've saved perfecting your microsurface (gloss/roughness) and albedo.
That chart is still useful though, for microsurface, which is super important.
EDIT: I also have comments and crits, just at work right now. Please hold. :poly142:
EDIT 2:
Additional map-related comments: No lighting information (AO OR highlights) in albedo, spec or gloss. Things don't get less shiny/reflective where the light hits them less-- the light just hits the same, equally shiny material less. The only situations where this occurs is where there's dirt and grime in crevices.
Lighting information in the albedo kinda undermines the entire point of PBR. The AO is separate for a reason, and highlights stop looking good in different lighting situations.
Gloss should be greyscale, not color.
pbr does not require metalness, metalness is just a different way to do things with different pros and cons.
The metalness workflow is just a different way that PBR is implemented in some engines, so the base knowledge and understanding is still the same. The metalness workflow benefits from saving texture memory by basically getting rid of the spec map all together. The final results and the actual shader is basically the same as using a colored spec map.
The metalness map works by labeling what is metal and not metal on the material, which is a white and black texture. Anything that is metal, will use the information in the albedo map as the spec information, then automatically set the albedo value to 0. Anything labeled non-metal will get an automatic spec value of around .25, because the spec difference between non-metals is so small that this value can just be constant.
Also, guys, thanks for explaining the PBR to me. I think I understand it a bit better now ^^ Gonna have to do some research on my own next !!
My good sir, thank you for explaining this so I don't have to.
Pyrzern: Listen to everything this guy is saying, he's giving some great advice.