*UPDATE
-----------------------------------------------------------------------------------------------------------------------------
OLD
Hey guys and gals, I've been working hard on this sniper rifle for the game "The Repopulation". Don't worry, it's not under any NDA so I can share the progress. We've given the individuals that backed our kickstarter campaign the chance to submit their own designs for weapons they'd personally like to see in the game.
Here's the initial concept art that was submitted along with the final silhouette I drew up.
I tinkered with other variants before landing on the above silhouette. I've never done silhouettes before, but after a few attempts I started getting the hang of it
Final hp and basic color scheme.
TB2 is just, wow :thumbup:
Let me know what ya'll think!
Replies
Holding it looks like it would be extremely awkward since the stock is so thick and the handle is so far forward
Marmoset Viewer
FLATS
HIGH POLY
LOW POLY + WIRES
did you explode your model? would you mind showing an image of it, if yes?
also am i right that you didnt bevel any of the edges in the lowPoly model?
You are correct about not beveling. This is for a last gen MMORPG so I was needing to keep the tri count down. I could've reduced it more than I did. Any added bevels would've shot that count up, but would've made uv'ing easier since I could use less uv splits at hard edges.
Thanks for your insight, i'll wait for your images. Would you mind sharing your uv maps? I'm trying to learn everything i can about baking, but's it's quite difficult for me sometimes^^
and i found that i can learn quite a lot trough other peeple's work
In the future, no rifle would ever look like that. Form follows function, and there is no function in 80% of the weapon. If you want to make a good weapon design, take 80% of a existing weapon (remember they put 1000 times the time and thought into it than you ever will with your concept as an artist no matter how good you are and theirs actually works, is produced and sold and many people critiqued it.) then add the features your futuristic version would have. If it has lasers, add a battery, a lens and cables to the power source. If it uses plasma, add a appropriate canister and gaz stuff. If your scope is using computers check out some existing CPU tech. Change form a bit maybe, add your own charme. Also there are many many guns out there being sold for 20 years that look more futuristic than most of the sci-fi concepts made. Authenticity is the key to making a badass gun, don't waste time by not profiting from thousands of hours of work put into real guns. You also learn a lot faster following a reference as you will just do the shapes that are convenient for the moment and ignore the difficult ones if you model your own thing.
Keep it up, the modeling skills are promising, but dont waste your precious time and portfolio pieces with concepts someone sketched in 30min.
I think you would be in much better shape if you just reduced that surface noise! Also try to breakup your materials a bit - everything in your roughness map seems to have the same base value.
Keep at it!
@Shrike I totally feel ya on everything you're saying. We've been giving people the opportunity to send in sketches of weapons they want to use in the game. Clearly we don't expect masterful artwork lol. I didn't want to deviate too far from the original design only because I don't know how that person would've felt if I had reworked it entirely. That's some solid advice for my next weapon! I'll use that general ~80% of a pre-made weapon and ~20% to future it up some. Get's most of the design work done for me.
@Deforges I'll take another pass at my roughness. I can see how everything appears of a similar base value. It needs more contrast. Still trying to get better at material definition, but I think I'm on my way. I think the surface noise is a result of either the brush size I used or possibly the sharpness I have turned up in marmoset The plastic might be a tad overly done. I'll try toning it back as well.
@mrjimmy Blast! I don't think this gun would pass field inspection :poly136: It could do some damage though!
@Staubkopf Here's a breakdown of my workflow. Gonna refine some things to make it more efficient so don't take everything for gospel
Exploded bake
The main idea behind explode baking is to move elements far enough away from one another so that an element's cage projection doesn't overlap with anothers. This would cause ray misses which is just a technical way of saying that area didn't project correctly. The green pieces are the low poly and the grey are the high poly/floating geo.
Determining what to mirror and what to leave as a unique element
Anything red is unique while anything green is something I mirrored. This has been an evolving process for me. When deciding what components of your mesh can be reused just imagine:
a) What type of game is this for? That will give you a general idea of the triangle budget which potentially frees up how you allocate any unique texturing.
b) The camera view. For example, if it's a FPS you don't want to mirror the sight areas because that will be what the player sees the most. The forend of the gun won't be seen by the player so that area can be mirrored.
Did you know that you can mirror specific faces just as easily as you can mirror the entire element! I never thought of that until commador filled me in on this revelation and blew my mind. I always associated symmetry with that lovely symmetry line running down the model. You'll notice I don't have a symmetry line like that anywhere. I kept the inner part of the model unique, an area I'd have typically placed symmetry in only to deal with a symmetry line. This isn't to say you have to always apply symmetry this way. I just wish to avoid having that symmetry line. Some engines have a work around. UDK had a nifty solution where you could fix the inverted shading you get with mirroring down the middle. I can only assume UE4 can do the same thing?
UV map
You want to make sure you have enough padding for each UV shell in ALL of your textures. That's something I didn't quite have enough of in my Albedo, Metalness and Roughness. I should've extended the borders for each shell outwards by like 4 or 5 pixels. At different intervals of the game textures are mip mapped, another technical term meaning your textures are down res'd. This happens when you're far away from the model since there's no reason to need to see a 2k texture from far away
I've learned to place my UV shells together based on their relation to one another. So if the stock is in one area I'll put the shoulder pad and any other parts of that gun in that area of the UV map.
What to do with mirrored elements
You want any mirrored UVs to be placed outside of the UV space. Any overlapping can cause artifacts and other issues when you bake your normal maps. The baker will ignore anything outside the 0-1 UV space.
Setting up material selection map
I highly recommend that when you bake out a selection map to do so by material!
I started creating a rifle in brush and now I'm Unsure if I'm going about it the right way. Anyway, the gun looks really good and I'll use your workflow for baking!
I haven't updated it in a while as I just revisited the asset today, had a few other projects I had to do but I'll put up the latest update when I'm done for the day. I was just curious how people usually approach hard surface, as I never see the same artifacts in other peoples works as I see in mine (hard surface that it).
And yes, the Zmodeler looks extremely promising and the next project I do will probably rely that. As it is I mostly stick to using shadowbox, clipping curves and deformation to do hardsurface so the new tools will probably flip the script for me :P
Thanks for your breakdown, it's very helpful. There were thing's i didn't know, for example the fact that the uv's of the mirror geometry don't have to be on the 0-1 uv space.
Thanks for your tip's
keep it up
Remember your art fundamentals- even complex items need some rest areas. Balance out those high density areas with some larger panels to give it some visual diversity- it'll make those detail areas pop even more, and make the design more believable.
My biggest concern about the design though is actually the thickness of certain areas- especially that grip area. The whole action sits on a wafer-thin panel that the grip is attached to. Other areas could be looked at too- remember that handle on top would be supporting the whole weight of the weapon.
Staubkopf - No worries!
Neox - I'll double check to make sure I'm using the right tangent space. How would I go about that? Picking the baker you used isn't in the mesh properties anymore. I've got my Y flipped correctly, but other than that I don't know of any options to choose my baker.
Wasteland Melody - I guess I was too cautious about what and how much I was changing. With the consistent feedback of the overall design I wonder if it's something I should totally go back and rework or just apply to my next model. As a portfolio piece I want this to demonstrate my absolute best skill. If these design areas are as problematic as they sound it might be worth my while to go back to the drawing board. Decisions...
I have to agree with whats been said. I think you should focus on readability and functionality before complexity. Certain parts look good, but I think the receiver suffers from unnecessary detail. As mentioned above - there is no place to rest the eye; it needs visual diversity along with functionality.
Personally I would study existing guns and concepts, and examine games like Destiny, Halo, etc. Try to replicate the important aspects and then apply your own. Something I noticed in futuristic shooters is that many of the guns look familiar yet they have aspects that make them look unique. The important attributes are there, but there are unique details built on top of that.
Designing is hard, but you definitely have the skills to create a great model if you get the design right.
edit: By the way, I hope I didn't sound too harsh. Keep in mind that the rifle does look good, but reworking a couple of things will take it to the level you want.
I've been reworking this puppy over the past few days to make everything more along the lines of what you all were suggesting. I did a collage of sorts and think I came up with something that's a marked improvement. Started from one end of the gun and really put myself in the mindset of holding it. I thought of comfort, usability, functions, and a lot less busy randomness. Here's what I came up with! Just gotta model it all out now.
VS
Looking at the strap placement I think I'd have to extend the front for the location to make sense.
Neat silhouette
Call-outs that make even better sense
:thumbup: ?
The biggest issue I have with the newer design is the "barrel" or lack thereof where the energy beam shoots out. Why is it so long and empty? If I were to run with that I can picture the top and bottom slapping together or at the least bending from its own weight over the years.
Plan on doing a few decal variants too
Worked on the orange paint, some of the coated metal, and rubber. Trying to be frugal with my scratch placements.
Here's the marmoset view with other render angles chasemarkham.com