GB Foxhound Military Vehicle for VBS
polycounter lvl 5
I created this model of the British Army's Foxhound for Bohemia Interactive Simulations as a show piece for VBS3.
I spent 12 working weeks (60 days) making this model. The 1st 6 weeks were spent on creating the high poly model (including interior), 3 weeks were spent creating the low poly and 3 weeks for texturing.
This is the 1st vehicle I've made so pretty proud of how well it turned out.
It was meant to be high detail so I went a bit crazy on the polys, something which became known in the office as "going full retard" (Tropic Thunder Reference). The exterior is around 70k tri and the interior is about 50k. There will be Lots of LODs for it in the end. The renders below are using the full 8192 textures, it will be 2 x 4096 textures in engine. One set (diffuse, normal and spec) for the inside and one for the outside.
I spent 12 working weeks (60 days) making this model. The 1st 6 weeks were spent on creating the high poly model (including interior), 3 weeks were spent creating the low poly and 3 weeks for texturing.
This is the 1st vehicle I've made so pretty proud of how well it turned out.
It was meant to be high detail so I went a bit crazy on the polys, something which became known in the office as "going full retard" (Tropic Thunder Reference). The exterior is around 70k tri and the interior is about 50k. There will be Lots of LODs for it in the end. The renders below are using the full 8192 textures, it will be 2 x 4096 textures in engine. One set (diffuse, normal and spec) for the inside and one for the outside.
Replies
why
Because it took that long?
ok, let me rephrase that: how?
i don't mean to be rude but 12 weeks is ridiculous, even for something very detailed. the high poly is nice, but i'd only expect maybe two weeks at the very most for what you've shown plus an interior. for the low poly i'd maybe expect a few days at most, and to be quite honest the textures shown look very simple, like maybe a day and a half worth of work. mainly because 8192 is uselessly annoying to work with and i feel your pain there.
all i'm saying is don't brag about it.
"I'm not racist, but..." lol
You seem nice.
He's clearly not bragging, it's fine for anyone to say they're happy with the results, that is not bragging. Bragging would be something like "I made this totally awesome vehicle and I'm pretty sure it's the best one around this forum". Instead of telling him what a normal work pace is like, and saying how weak the textures are, I suggest giving him some valuable feedback/critique he can use to improve his work. I know you're a good artist but this is just a pointless comment.
I do agree the textures could use some work. As of now it's pretty plain in terms of colors and actual texture, and it's hard to read the materials. Try getting some cool shapes in (spots of dirt, rust, bolts etc), more decals/text and add some points of interest like bullet holes or something. Maybe this tutorial is for you:
https://www.3dmotive.com/f101201
I'mma stop you there. Not everyone is a super pro. I actually do almost this exact work (VBS2 dev) and if where he works is anything like my job when we made the jump to doing hard surface high poly bakes (which was actually just in the past 6 months) I didn't have anyone to show me how to properly explode a model for a clean bake or even hard surface workflow in general. Sure, I've done some high poly before but it was usually organic stuff and in my opinion, setting up a hard surface bake is a different beast and if you don't have someone showing you the ropes, it can take quite a bit of trial and error to get right.
The issue with this is that most army sim stuff is preferred to be really clean. I've had to go back and clean up textures because they were too dirty even though the vehicles were usually stationed out in the middle of the desert. =/
Anyways, Coontang, I think the high looks really good but I think the game model didn't take some of the normal maps edging as well as it could have. The edges looks really sharp and thin and it gives the idea that some areas are a lower poly than they probably are. This can happen if your high poly edge loops are too sharp in your bake. In any case it's cool to see some VBS3 stuff get posted. We're looking forward to the upgrade.
The only thing I can say against it is the Textures sizes are to big for my liking.
Something I try to remember when ever I make something that will need lots of detail is a 2048 map is huge. The characters for Dominance Wars only used a 2048 I thnk. Perhaps you should try halving your texture sizes and see if theres any difference but its up to you. Anyway Great Work!
Personally, I would not have gone with that meshing over the top and doors. Still, great work!
i was just baffled how any company would accept 12 weeks. i can understand taking 12 weeks if you are learning (as the OP explained) but that's a crazy long deadline in any production environment. i'm not attacking the worker, rather attacking the companies.
but in a game industry context, surely one must realize saying it took them a long time doesn't make them look very good. this isn't fine art where you're putting love and soul into every brush stroke. that's a waste of time. we are workers, given a task and producing a product for a client. the faster your output, the more money you make or the more you can laugh off deadlines and crunch times at your comfy desk job.
also you're really telling me how to give feedback? don't worry i was getting to it
the high poly is generally pretty nice looking. show some pictures of the interior, that's where things get exciting. the edge widths could definitely use some beefing up but i could see milsim art directors complaining about things looking soft if you did. odd standards but it's a possibility. i wouldn't have zbrushed the wrapped up cloth bits or the straps on the front of the vehicle. you could have achieved pretty much the same result within max or maya without dealing with the hassle of getting it into zbrush, sculpting, decimating, and bringing it back out for baking. that alone is a big waste of time for not a huge gain. use cloth simulation to it's fullest, it's very fast!
the biggest waste of time is that mesh wrapping. it's impressive in the high poly renders but ultimately doing that does a few things: firstly, it drastically increases the tricount of the high poly, giving potential performance issues when it comes time to explode it. secondly, it's a huge pain in the ass to place everything and make sure it looks correct. you could kinda cloth sim it, but it's still not fun. thirdly, baking stuff like that only works when you have TONS of resolution--which you do, but it's not something i would suggest to anyone. i would have just low poly modeled it and made the pattern in photoshop.
so i've done some milsim work before and texturing for those clients is... interesting. it's true, they like things very flat and sterile, but there's plenty you can do to make it nice. often times they're working on very outdated engines with just diffuse or something. the main keys here are to keep your materials reading very simple and give the edges some nice treatments. avoid noisy overlays like on the tires and some of the darker metals. make good use of color to define materials if you don't have functional spec or gloss. watch out for your AO too, it looks like it's too dark and needs the gamma adjusted. get some dirt in crevices instead, that should help. you can also add some subtle baked upwards lighting on the diffuse if you know you're working to outdated specs.
your render doesn't help the flat look either. there's no specular i can see, so yeah get some of that going.
anyway keep at it, your high poly looks quite decent. practice a bunch and clients will be lining up to hire you.
yeah
No hard feelings I hope.
this is a bit ridiculous to hear--it's absolutely relevant. we are not fine artists, we are workers on an assembly line working to deliver parts of a product. i don't feel like very many people in this community understand that, but it's the hard truth. as i said, the faster you can deliver your product, the more comfortable you will be either in money or workplace comfort.
btw the point of the posts was to basically say: hey, revealing such a long delivery time on your asset makes you look bad. i don't mind being rude.
Just out of curiosity, was this done in max?
"If you aren't in the industrie right now, you have one benefit: time, use it! Over the time you're speed will increase. I rather hire an awesome artist who needs a little bit longer, then one who creates average stuff in short time."
So from this, take your time as long as your not in the industry, speed will come automaticly. This didn't mean that you shouldn't work activly on your speed (giving yourself deadlines is a good way to start).
Also he only said 12 weeks, he didn't said f he works every day 8 hrs or maybe only at the weekend 3 hrs every day.
I know it from myself that sometimes, if you're still in the "hobby" phase, a project can take a while, simply because you have to work, meet friends, play some games in between etc.
For the model:
The modeling part looks good, even if I would add some damage here (to reflect it's a "used" vehicle).
The texture looks a litte flat. I would add Dirt a few scratches, color variations etc. (also add them to the spec map as dirt didn't have the same spec as metal paint^^). I would also check real world photos of military vehicle to see where scratches appear, dirt assembles etc.
That does look helpful thanks, I'm definitely going to work on the diffuse some more. That's partly why I posted it because I knew people would point out any issues.
I just need to check if the interior is sensitive info, I have permission for the outside but not sure on the inside.
This is something I'm aware of too, but I could possibly dirty it up some more just for my own version.
I will definitely try 2k to see, however our engine compresses textures down a lot so I think 4k should be fine in the end.
Thanks! The meshing is fitted when it is out in the field so it had to be added, and I added it as well as I could. This has been an ongoing project for 5/6 months. I have been working on it in between client work so I was able to take my time and make it as detailed as I wanted and also try out various things like making the cloth netting. I actually remade the straps in max in the end to get them cleaner, I'm not sure why I thought Zbrush was the way to go at 1st. There wasn't really a lot done in Zbrush, it just gives a nice easy render. I agree that was overkill and could have been done in the texture, but I had the time and I wanted to see if it could be done, and it looks nice in the high poly. Mostly, one or 2 small pieces were done in Zbrush. I have actually done this but maybe it's been lost a little, I'll try to make it stand out a bit more.
I was working full time but as I said earlier it was a low priority in downtime between other projects. The 60 days is the amount I have logged in our time tracking system in total. I am only a junior at the company so it was sort of about learning on the job.
[ame="
you should have mentioned that! hopefully no hard feelings. the high poly does look nice.
atleast for your own sake put in some edge highlights.