Portfolio Feedback (3D Modeling)

Astrocoolbug

Hello Everyone,

I've made a few posts on Polycount before and got some great feedback for some characters I did. I made a ton of progress through that process so I wanted to share my portfolio as well.

I'm really interested in heading into the games industry and I do love 3D modeling, I do have skills in other areas but modeling is where I really get my energy. I spend most of my time making characters, even more specifically stylized characters. I just graduated from a state school a month ago and now I'm in open waters and it's quite intimidating. I've been applying like crazy but I know my stuff still needs A LOT of improvement.

I'm not really expecting to become a character artist as my first job, so I'm thinking about what I should do. Whether that be diversify my work to include more environmental pieces or do more realistic. Basically where I'm a bit confused and would really like some guidance is how specific versus broad should I go?

I don't want to be too specific that I limit my opportunities but I also don't want to be so broad that I come off as a master of none.

My current portfolio link is right here, please give me any feedback you have. I would greatly appreciate it.

https://www.artstation.com/andrewdowns

BagelHero

Mmn. I have trouble weighing in because character artist probably IS riskier (WAY less jobs), and possibly a bit harder. But imho? Your eye for character work is MUCH stronger than your eye for stylized props. Your character work still needs work, but I can see where you're going with it and especially with the priestess there are some forms and details I'm quite impressed by even if I feel there are some areas that are lacking. But your props near all each have some bothersome flaws/odd interpretations.

Ellundra - While I feel there are some tiny issues with some clarity and silhouette, and the lighting could be worked on, I think this is great. Probably the best piece on your folio right now other than the torb gun.

Regalian Longsword - Your concept slaps! It's very clean and nice and I like it. I also did not think it was yours because the shapes and construction of the 3D model feel very different. Given the switch in art styles I don't expect it to look 1-to-1 the same but I feel like key forms are just.... skipped over, perhaps in favour of things that are easier to model? I don't exactly know why some things were changed, but it doesn't look like the result of an intentional art style shift. I'll draw over some of the changes that concern me the most. Immediate concerns in blue, confusing areas in orange & pink.


Torb Gun - This is fine! Textures seem a tad unpolished but its a keeper. Like the comp.

Wildebeest - ...is a character who is mostly made up of props, and raises some concerns in its current form. I feel poorly about this late crit on this in part because I keep meaning to follow up in your thread and I kept forgetting.

Problem number 1 is absolutely that it's just not finished. He doesn't have legs! They're seethrough, and it's very obvious. Unacceptable and needs fixing asap; even if you just put in some hold-over primitive to cover it up.

Once again your direct comparison is hurting you and raising questions, so many different areas have been broken down in ways that are less appealing or less functional, without there being a clear technical limitation. Many of them have in-game mirrors you could have referenced to figure out what was intended by Tsang as well.

Some points that immediately jumped out -



Part A: Modeling

1. Shapes in legs are neither overwatch nor matched to the concept. Gotta go with one or the other-- the legs are a bit vague so I get having to wing it a bit but there are certain kinds of rules for overwatch boot armor. They're pretty minimal, often one solid piece or at least pieces with large, soft detail-light surfaces, the shoe is a separate piece underneath this, sometimes hinged sometimes completely separate. Examples that would make good refs:
   
2. You can't just make up the areas that are the most clearly defined. Breakdown the shapes before modeling and block out your model before detailing/separating it into parts. This is a broad, fairly simple codpiece with a flat extrusion in the middle of it and two hip guards, and you've complicated it so so much. The additional side panel bits you've added also change the silhouette a lot. Overwatch relies strongly on the rhythm and interplay of concave vs convex shapes, so this is something you should put a lot of thought into before changing.
3. I think this is underarmor. It should probably be pretty unintrusive and serve as a point of rest for the eyes. The lines seem more like panel lines and less like extrusions or larger forms/pieces, which is why the side panel bits are so jarring to me. even though he had some small details in this area, final Reinhardt also had a similar place of rest around his waist.
   
4. this area you've simplified TOO much. But it's also an area you can easily take liberties on so it not too bothersome-- the way it connects around the side and the overall shape/silhouette is a bit more of a concern. I think the side bit was an overhang that flares out a bit, maybe like the ankle pieces on characters like soldier 76, but its a little odd and probably not great for animation so changing that isn't inherently an issue. Probably something similar without the flare, though, would have been better.
5. the upper chest and head of this dude is just like... a ball. The chest armor, as such, just really needed to be more curved/rounded overall. Right now it feels like a flat plane only curved on the x axis, but not vertically.
6. I can't decide if these thigh parts are armor or fabric covers, and it kind of looks like you couldn't either. I'm thinking fabric, if so they needed more work and some of the pattern you could see at the top in the concept. also! only the one crease line in the front, 2 makes it look like a panel. There are pretty severe silhouette issues in this area, the shapes are pretty different in ways that make him look way stubbier than the concept. in part because the way you've changed where the fabric ends changes the impression of where his hips are and how wide they are. If you were going for armor the pieces are too wibbly & I'd remake w/ hard surface, tbh.
   
7. Another area I can't fault you for not matching exactly, but also another one where more references from the final iteration of overwatch would have come in handy I think. Both for the materials you went with and for the kind of shapes and joints used. The overall silhouette and forms have lost their chunk, but the individual panels/armor pieces have lost their chunk when compared to the concept. It's kind of hard to nail down exactly why this looks wack. Sorry.
   
8. Lots to address here. The angle of the shoulders is really important to nailing both the silhouette and the intimidating nature of the pauldrons. If they aren't wide and low (like they're about to charge) they look... a bit goofy, but also weightless. This could likely be addressed in the rigging and posing process though. What can't be is the horns not being a fairly contiguous addition to the overall "head" piece. Overwatch appears to be a universe where they can easily manufacture contiguous metal parts individual to human-sized people and plug 'em together like a little cast figure haha. they're attached via a fairly simple panel line in the concept, and the hole in your final model looks weird. The eye-hole should also be a bit more natural-looking I think. having it be artificial takes away again from its intimidation factor. the underside, the connecting joint also looks a little funky. Another piece I'm not sure would work in the concept but could surely be a bit closer/more overwatch-y than what you've gone with.
9. Finally, in the future make sure pieces like the necklace and beads are proportional to the concept. Your beads look like you just made them really low poly for some reason because the bevel and gap between them aren't enough to read at a distance, and the necklace is floating. I think you couldn't tell if the shape was foreshortening/perspective or the actual shape, and I'd say it's both. I also forgot to change the color but I wanted to point out some shape/silhouette differences in the helmet.

Part B: Color

1. Material breakdowns are important to nail down BEFORE you get into texturing, hell, before you even start modeling. Feel like this image kinda speaks for itself; your textures are samey because you haven't defined as many different materials as the concept has. This also just changes the interpretation of forms pretty severely. Makes the silhouette differences pop out too. there are some areas I'd argue might actually be raw metal where I've drawn them in white paint here, etc, but the point still stands.
   
2. (while i did notice there seems to actually be a brown material on yours, it should be painted metal I think, not metallic itself. it blends too much with the other raw metal and becomes muddy and undefined.)
   

Also, again I kept meaning to comment this earlier and I'm sorry for not doing so, but the weapon he holds is actually a flail. the head is an orb that sits within the handle.

I think both not matching it to its own concepts AND not matching it perfectly to the static fullbody concept where it could be mistaken for a club/mace hurts it. It would have been cool to have an action shot where he's using it too, which shows off you can make more complex props than just one-piece static pieces. But that's something to take away for another project, I suppose.

Sorry for being so straightforward about this. I know this is... a wall of text, most of it crit, but please please know it's because I think you have promise. I was following that thread pretty closely and you were clearly working very hard and listening to everyone as much as you could, you approached things pretty boldly and didn't get too hung up on bullshit and that's super commendable. But in order for you to pull yourself out onto the other side as someone hirable, you need to get invested in the kind of details I pointed out.

Sloppiness when you have all the time in the world to work on a portfolio will not get you hired for an environment where you will almost always need to be sloppier due to time and technical constraints. You need to plan a bit better, push a bit harder, pay more attention, and really see things for what they are. Invest the time into making sure you do things right.

Oh! and you probably need another project or two. Esp if you can improve or replace some of what you have right now. It's not super cohesive and I think that's just a reflection of your lack of direction.

...Aaaand that brings us back around to deciding what you want to do. This is about you. What do you like to do, what do you feel like you could really work on forever? Is it more organic humanoids, is it guns, is it cool armor dudes? Any and all? Regardless of the jobs available or how likely you think you are to get them. There isn't a perfect portfolio for everyone, so it's honestly better to focus on things you love or are good at, as you'll likely improve quicker if you're able to focus on something you can do without hating it or getting really tired.

Ellundra is good and shows a lot of promise for stylized character art-- a slim area for jobs, but you seem good at it. If you have the luxury, taking a few months to a year to really really push yourself and level up in this area may actually allow you to get in on a character-related job. But that may be a luxury you don't have. & you might have to get a side gig if you don't already have one. It may limit your chances to land a job in the industry on a generalist portfolio, if you'd like that kind of work just as much.

Conversely, your fire hydrant is stock standard, but if you could pump out a bunch of props or small environments to that standard or higher + some good looking hero props, that might be a faster way in. You may not have a lot of time to improve your character art portfolio after that, though. Would you be okay with that?

You could always do stylized props, mobile might be easier to get right to freelancing for, you mentioned you were working on a mobile game on your folio. Maybe that route? Sell some packs of generic props or something? I dunno. It would be work, if small. Something to show.

Lots to consider, and not a lot I think we can help with easily. But... IDK. I hope this helps a bit. Both on direction, and in terms of your portfolio quality itself. Good luck out there in the real world!

Astrocoolbug

@BagelHero First of thank you so much for the feedback, one of the best crits I've had. I really appreciate how forthright you are, it makes it extremely clear and motivating so thank you. I'm not going to reply to every point mostly because I agree with everything you said but I wanted to add a few things and clear something up on my end.

I'll start with the end actually, I think the question of what I like to do is organic characters. And I'm glad that seems at least to be the one that is showing right now. I've spent definitely the most time shoring up those skills. I enjoy organic characters far more than the other things I've modeled. That being said though, I did like working on Wildebeest but that project was a real struggle for me and I think I know why now thanks you and just me sitting with where I ended up with that project. So I think overall I enjoy making characters in general, Organic slightly more than hard surface, but I do like both.

Speaking of WIldebeest, I tried to rush this project to the finish line for no good reason. I fumbled and it shows. I don't think I really thought about that when I finished it, but you are 100% about everything you've said. In fact I've already gone back to rework the model, and this time I have a plan for fixing him. I learned probably the most ever from that project and I think that's why Ellundra turned out a lot better, but I also learned things from Ellundra's project that are ESSENTIAL to Wildebeest. So I am reopening this project and I'm going to be making more updates soon. And just so I'm not blowing smoke, I'll summarize my plan and have a more detailed one in the actual thread soon.

1. Match the concept, way better (This is something I struggled a ton with, but I learned a lot from Ellundra since her forms were a lot easier to digest and see. But I think I have the skills now to execute it properly on WIldebeest)

2. Making sure my high poly pieces are clean enough for baking before just thinking "eh it wont show in the low poly, why worry?" I think there are some cases where I can get away with that but for the most part I think that's a terrible excuse.

3. I did a lot of texture experimentation recently and bought a few tutorials from professionals that really helped me learn a lot more about stylized materials specifically so I really want to try and match the Overwatch textures as best as I can with the programs I have.

4. And lastly, and this is a big reason why I got impatient and discouraged when I said I finished this character last time, but the rigging and animating part I just do not get a lot of energy from. I don't know why but I really just feel like I'm wasting time by doing it. I've actually turned to auto riggers for characters like Ellundra which I feel like helped me just get a character out and allowed me to do more tweaking to the model itself rather than getting stuck in the weeds. I was constantly being pushed at school to do rigging but, I think I just don't like it all that much. I'm very grateful to know what I know about it but I really feel like that it sucked all my energy and I just said that the character was done instead of doing the hard work of going back and fixing him.

5. And lastly lastly, I learned a lot more about lighting recently too, but this is still one of my weaker areas. I need to just really study this and execute it well. If you or anyone has any good tutorials/reference for making good lighting I would really appreciate it. That one is bit harder for me to see for some reason but I think it'll come with practice.

Again if I left something out in my reply it's because I agree with you. Wildebeest feels like a bit of stain right now, and I've felt that for a bit so I hope I'll be able to turn this around and get a version out that I'm really proud of. Thank you again for you feedback, I hope I don't disappoint with future Wildebeest updates.