wow thats a really neat gundum well done! Ive seen quite a few gundam models but they are often simplified or loose the essence of the original but this looks really Gundam to me. It will be nice to see what you can do with some lighting and textures, though the original characters werent very interesting in terms of textures were they?
Just a quick experiment with rendering layers and positioning 3d space to fit into a photography. I'm planning to UV-map it right after the Christmas weekend cause then I will have the time.
Have a look at this in the meantime.
dang i think once the final textures are done and some really bad ass lighting is added to this guy and a quick photoshop paintover and this piece will be a total bad ass
[ QUOTE ]
dang i think once the final textures are done and some really bad ass lighting is added to this guy and a quick photoshop paintover and this piece will be a total bad ass
[/ QUOTE ]
I second that. Can't wait to see it when it's done. You've got a good start on the compositing allready.
I'm sorry to say that but these UVs are likely never going to work... Seems like you UVed every single piece as an indivudal gameart objet - but I'm sure you could reuse gradients and color all accross the model... And since this model is more like a film model in term of complexity why not take the whole leg and UV map it from the side for instance? Not perfect, but so much faster.
Plus the way you organized your layout does not make much sense especially with the peices fighting each other and sitting at an angle. You could do yourself a favor and just put them 'upside up', it will make them much faster to texture (if you even plan to paint the texture traditionnaly I mean). Aslo your pieces are way to close together for ingame use.
I'd say ... with what you have so far, just run a layout script that separates the UVchunks much more to avoid mipmap bleeding and simply bake your render to the texture. You could have saved yourself *hours* by just running an automatic UV script and baking the render to the automatically done layout - automatic layouts are usually a no-no for game models but in that case it wouldn't make much of a difference considering how highly split the current layout is.
Hope it makes sense?
Mesh looks cool still :P
Edit - obviously my crit was made with gameart in mind. But still if it is intended for film-like use, why taking the hassle of UVmapping that way? Just use procedurals, simple big UVs chunks from the desired angles for the textures overalys, and voila...
I think I have trouble understanding this model because 1 its not working as an ingame model and 2 not as a film model either becaue of the very hard edges (normal splits?) defining the blocky shapes. For fil m use you are better off with a microbevel to catch light. But if it's just to produce one still image ... why bothering with the UVs? You could just overlay the dirt textures in PS!
I am quite curious to hear what you intend to use it for!
Well... I could say it's for film use then. For the layout I think I'll split the map into several parts instead. Think it will be easier to work that way, and having it at higher resolution in Maya.
And why bothering mapping it? Well, for the challange, for the practice and everything I might learn from it. I want to at least have a turn table made for it for portfolio purposes and therefore have it looking good from all angles.
And you suggest me going the easy way and making look "not perfect" which isn't good enough. I want it to look perfect!
Who wouldn't want that for their work?
And the beveling? I've given that a thought but I don't want to bother with that since I want my model as clean as possible. And still, for softer edges, we have anti-aliasing. Not that soft perhaps, but still.
cooool, but that UV layout aint gonna help you even if you are using massive maps, painting it wil be so slow.
personaly i'd split it into material types (ie armour, bare metal, plastics etc) and lay the uvs so things are more grid like, not to get the most out of the space as thats not realy the point (is it?) but just to make drawing on and lining up decals, panels and other linear details that much easier and faster.
bugo: Approximately 1.5 weeks. Didn't use any plugin. Everything is done manually. :P
SHEPEIRO: Yep, as I said before. I'll split it into several maps instead, though I wills sort them in parts instead of material type since they're all going to have the same base surface, which later will be covered in paint etc.
I'm planning to have a map for the head, torso, arm and leg. Arms and legs will have individual maps for their own decals n stuff though I think I'll make them from the same base.
my gosh! 1 and half week? duuuude! did it give you crazy headaches, didnt you? man! 1 and half week doing a UV, well, maybe in some years we will have to make that, who knows. But hey I do agree with Pior, there´s alot of separated pieces, i know you could be doing this for film, etc, but anyway, this is effort your time anyway, it shows patience.
lol 1.5 week to uvmap maaan,i'd rather trash my model than trying to do that uvmap ehehe,anyway the model looks cool indeed,even though i'm no fan of gundam...at all....
The thing about the 'perfection' of it is - if this was to be a game mesh you would get soooo many splits that even with your theoretically accurate layout the mipmaps bleeds and oddities it would create would be just as problematic as what you would have with an 'unperfect' layout ie simple side projection of the whole leg for instance. Difference is that one technique takes weeks for the UVs, the other a mere split second :P
BUT it's a film mesh... :P Still I highly doubt that in a film production environment anyone would spend this amount of time on UVs. You could reuse materials, use procedurals for most things, maybe do smart things with UV sets... But then again I'm not sure.
I mean that's the catch for folio pieces ... you'd rather show that you can think outside the box rather than spending a very long time on a pipeline bottleneck that could be solved differently. Well it shows you've got patience so that's not bad :P
Good luck with what's up next!
(ps - what's going on with the boards lately? each time I type m.o.d.e.l it gets deleted once the message is submitted? Is this just on my side?)
The layouts are a heck of a lot less chaotic now (though I can barely see them, ha)
Maybe you already planned this, but you might want to consider combining them into one sheet, with area occupying a quadrant ,instead of there being four different maps
As textures on the model, they will be separate, because then I can keep everything at a higher resolution.
If I will have them all in the same PSD file I've not decided. Probably not since the file will be 8k x 8k then, and that might get heavy when the layers are increasing in numbers.
Getting along with texturing. About 75% complete. Left is eventual wear on the paint and some sort of phase shift effect.
I'm also planning on texturing the sword.
i dont get what you tried to accomplish, you made this really messed up uvw, so i assumed that you wanted to get really unique detail in it , but it seems that what you wanted is only a metal texture slapped in it ? also greyscale ? i would go for the color concept you had in the mockup with the photo. this is really weird right now
Well, it was a bit more complex than just "slap in" a metal texture.
And the colors will come plus decals. I guess I made myself unclear when I only said "phase shift".
right now you have some nice darks tones, mid tones, but your 100% missing any light tones. I suppose if you add some off white tones along with your color (like the first image in this thread) as an overlay it should help that out.
Once thats done I imagine some edge highlights will make this model SING. but- I also imagine that will take forever.
craziest unwrap I have ever seen. how long did that take just out of curiosity?
Oh yeah, and whats the count on this, plus do you feel like posting wires?
Definitely the occasional paint chipped areas, especially such as around the edges of the feet, knees, elbows, and anywhere else that might see wear and tear... maybe around the cockpit (chest) area, from pilots clambering in and out...
Well, I will stop working on this for the time being so I guess you could call it the final. I know there are some things that could've been done better, and I choose some pretty inefficient ways to work. Still it was a lot of experimenting and I've learned a lot from it. So now I at least know what I should not do on future projects.
Here's a huge amount of pics, and now I'm going to warmup for the coming dominance war!
The only things I would add might be paint chipping around "frequently contacted edges" like the feet and things, and a bit more grime right at the seams of things...
Yeah...some metallic worn edges on the bevels would definitely make better use of the texture size...by today's standards you can be surprised how much of that you'd find in a texture size like that.
And some chip off and mud traces along the feet where it touched the ground would work nicely too.
It's blood? Gundams are like 55-60 ft (18 meters) tall. It seems a little hard to believe he went around slaughtering infantrymen or something until his 50 ft long sword was soaked in their blood.
Replies
wires? polycount?
Now texture the scratches and wear.
Polycount is at approximately 60 000 tris.
Sounds great that you want a dirty and worn out texture on it since that is what I'm planning.
I'll see if I get some time to throw together some nice construction shots.
Cheers.
Have a look at this in the meantime.
dang i think once the final textures are done and some really bad ass lighting is added to this guy and a quick photoshop paintover and this piece will be a total bad ass
[/ QUOTE ]
I second that. Can't wait to see it when it's done. You've got a good start on the compositing allready.
Model is ace (even though i hate gundam designs), uv is... well... will see what is coming out of it, anyway: GOOD LUCK !
Good stuff nonetheless though.
polycount is 60k, why?
Plus the way you organized your layout does not make much sense especially with the peices fighting each other and sitting at an angle. You could do yourself a favor and just put them 'upside up', it will make them much faster to texture (if you even plan to paint the texture traditionnaly I mean). Aslo your pieces are way to close together for ingame use.
I'd say ... with what you have so far, just run a layout script that separates the UVchunks much more to avoid mipmap bleeding and simply bake your render to the texture. You could have saved yourself *hours* by just running an automatic UV script and baking the render to the automatically done layout - automatic layouts are usually a no-no for game models but in that case it wouldn't make much of a difference considering how highly split the current layout is.
Hope it makes sense?
Mesh looks cool still :P
Edit - obviously my crit was made with gameart in mind. But still if it is intended for film-like use, why taking the hassle of UVmapping that way? Just use procedurals, simple big UVs chunks from the desired angles for the textures overalys, and voila...
I think I have trouble understanding this model because 1 its not working as an ingame model and 2 not as a film model either becaue of the very hard edges (normal splits?) defining the blocky shapes. For fil m use you are better off with a microbevel to catch light. But if it's just to produce one still image ... why bothering with the UVs? You could just overlay the dirt textures in PS!
I am quite curious to hear what you intend to use it for!
And why bothering mapping it? Well, for the challange, for the practice and everything I might learn from it. I want to at least have a turn table made for it for portfolio purposes and therefore have it looking good from all angles.
And you suggest me going the easy way and making look "not perfect" which isn't good enough. I want it to look perfect!
Who wouldn't want that for their work?
And the beveling? I've given that a thought but I don't want to bother with that since I want my model as clean as possible. And still, for softer edges, we have anti-aliasing. Not that soft perhaps, but still.
did you use a plugin to layout it?
personaly i'd split it into material types (ie armour, bare metal, plastics etc) and lay the uvs so things are more grid like, not to get the most out of the space as thats not realy the point (is it?) but just to make drawing on and lining up decals, panels and other linear details that much easier and faster.
cant wait to see it
SHEPEIRO: Yep, as I said before. I'll split it into several maps instead, though I wills sort them in parts instead of material type since they're all going to have the same base surface, which later will be covered in paint etc.
I'm planning to have a map for the head, torso, arm and leg. Arms and legs will have individual maps for their own decals n stuff though I think I'll make them from the same base.
BUT it's a film mesh... :P Still I highly doubt that in a film production environment anyone would spend this amount of time on UVs. You could reuse materials, use procedurals for most things, maybe do smart things with UV sets... But then again I'm not sure.
I mean that's the catch for folio pieces ... you'd rather show that you can think outside the box rather than spending a very long time on a pipeline bottleneck that could be solved differently. Well it shows you've got patience so that's not bad :P
Good luck with what's up next!
(ps - what's going on with the boards lately? each time I type m.o.d.e.l it gets deleted once the message is submitted? Is this just on my side?)
Anyways, here's the new maps.
Head
Torso
Arm
Leg
Maybe you already planned this, but you might want to consider combining them into one sheet, with area occupying a quadrant ,instead of there being four different maps
If I will have them all in the same PSD file I've not decided. Probably not since the file will be 8k x 8k then, and that might get heavy when the layers are increasing in numbers.
but those look alot more managable now.
can't wait to see this textured
good work there, looking sweet.
I'm also planning on texturing the sword.
it will be all grayscaled?
And the colors will come plus decals. I guess I made myself unclear when I only said "phase shift".
Once thats done I imagine some edge highlights will make this model SING. but- I also imagine that will take forever.
craziest unwrap I have ever seen. how long did that take just out of curiosity?
Oh yeah, and whats the count on this, plus do you feel like posting wires?
details, woven carbon fibre bits, different material types, damage, you have the space use it
as it stands it is not coming across, though I am looking forward to when it does.
Here's a huge amount of pics, and now I'm going to warmup for the coming dominance war!
Cheers all!
Renders
Wires
Textures
The only things I would add might be paint chipping around "frequently contacted edges" like the feet and things, and a bit more grime right at the seams of things...
And some chip off and mud traces along the feet where it touched the ground would work nicely too.
Other than that...the sword seems fine
Keep us updated.
Still its looking great
And I've put up a turntable.
http://www.vimeo.com/621849
That is blood...
[/ QUOTE ]
It's blood? Gundams are like 55-60 ft (18 meters) tall. It seems a little hard to believe he went around slaughtering infantrymen or something until his 50 ft long sword was soaked in their blood.
Very impressive modeling. Wish I could pull something like that off
I think it's from when he killed Godzilla. It's Godzilla's blood on the sword.
[/ QUOTE ]
Exactly.