So I was wondering how many UVs should I use for a Character I have always used only one. However I was wondering if I should be using more for animation, which I would like to have higher quality textures? Any help is greatly appreciated, thanks.
How Many Is Too Many?
Let me try to explain…
For production wise, it all boils down to render time, efficiency, etc. So normally we will look at the shots and how much your character will be shown on screen, and finally the output resolution of your image.
Lets say, you are rendering in HD1080, which is 1920 x 1080, and u have a close up shot showing the whole head of your character. In order for the texture to look nice, it will have to be at least 2k res. Imagine if your head texture is only 1k, the render will definitely look blur and pixelated.
And lets talk about UV now. If you squeeze your whole character into one UV map, eventually you will need a 8k map just to make sure that the head is of enough resolution. However, you might not need as much details on the body compared to the head.
So, this is what I will do. Separate the UV of the head and body, so I will have a good texture resolution of the head for my close up shots. If I am doing HD output, my textures will be at least 2k, or even 4k. I will try to optimise the UV space and squeeze in as much in the 0 to 1 space, so that I am not wasting any unused space.
One more tip is to have a low-res texture for test render and lighting, maybe 1k texture in jpg, and switch to the hi res version when doing the final render.
Hopefully you get what I mean 
By UV’s I believe you mean UV maps?
In that case you can increase the size of the texture to allow more detail. For instance you can use:
-512x512
-1024x1024
-2048x2048
-4096x4096
Alternatively you can as you said increase the amount of uv maps for a model. Be careful you need to be sure that the uvs sizes are universal to ensure that the pieces all have the same resolution when it comes to texturing.
For example you can have a jacket on one 512x512 map, stretching across the whole map, and a ring on another spread across the map. The jacket would be low res compared to the ring. This is just hypothetical you probably shouldn’t split the maps up that way.
Any more questions or need further clarification I’m here!
-Ryan
Thanks very much, I thought that was the case, I just didn’t know for sure. So just so I can make sure I have a clear concept of this… for a parts of the body that will need the most detail i should use bigger maps on them, or just increase their layout on the map in comparison to other portions of the model? Also I am creating a robot ideally for a short animation of my own, I intend on having some close up shots of the robot. I was thinking after seeing how armor was done for “Snap” in “Round 6” from Snowball studio was done separately on a 4k map that I should break the robot down into 11 different portion, I see this is a bad idea now, I was excited to try something different and was getting carried away without doing any research. What would be the best method of moving forward. I’m thinking maybe just two maps, one for upper and one for lower each 4k maps? Oh the robot is about 4 to 5 stories high and I intend on having a Human Character near him at certain points.
Also whats the standard res for main characters in the film industry.
that depends on how much ressolution you need… highend chars are using up to 200 uv tiles or more… take a look at the mari blacksmith asset… the upper body alone contains 40 4k textures per channel…
http://www.thefoundry.co.uk/products/mari/training/#blacksmith
if your camera is often close to your character you need the max resolution on the whole body… so if your face needs one 4k map the rest of your char needs also 4k maps for each part with the same size the face has… if the there are parts the cam wont be close you could make them smaller… but i would aim for the same size for all parts…
There is no “standard res”. As explained above, the resolution used for each UV tile depends on how hero the asset is - if it’s going to be seen super close to the screen, the textures are going to be big (8k and over). But it’s also a balancing act between UV tiles and texture res - sometimes it’s better to use a higher number of UV tiles with lower res (4k) textures than to have fewer UV tiles and larger textures. It depends on the asset.
Thanks guys, this is making much more sense, I’m really just now getting my feet wet with creating textures and it was honestly a bit overwhelming at first, but you guys are really making this easier to understand.
So please let me know if I’m really getting this or if I’m completely mistaken.
Ok wanting to have my Character Asset be seen close in its entirety I should use high resolution maps, dedicating uv tiles as need to ensure each bit of the mesh has enough uv tiles so nothing is favored. (basically my applied checkered texture would have the same size squares all over the mesh.)
Now if I am going to favor a certain portion of the character asset I can use the same size maps as the rest of the body, just allot more uv tiles to the portion of the mesh that’s getting a close up shot. Which can also be done by using this method, http://www.neilblevins.com/cg_education/multiple_uv_tiles/multiple_uvs_tiles.htm.
Ok guys am I correct?
Thanks dude i have been looking for a mesh I could see to stack my work and methods up against.
Just had a chance to open that file that’s insane, and that was my original idea kind of, except for larger portions of character. Thanks that really helps me get it much clearer than I had ever before understood it.
So is it common to cut up the layout like that, and by that I mean its not something frowned upon right also is it better to have the mesh one piece or is that a game thing?
that depends also on your needs and pipeline… sometimes you have to split up your mesh into many parts… for example you have to animate a knight… alle those parts of his armor have to be single meshes cause all are needing there own pivot for animation, some parts have to be ready for simulation… sometimes its easier to share one texture over several parts sometimes its easier to have a texture for each object or shader…
OK that makes a lot of sense, thanks, I’m sure I’ll have more question once I get going into this thanks, guys. Also that’s the mentality I went into making this character with, so perfect example.
So here is the robot, I was only able to use 2k res maps, as my comp wasn’t playing nice with the 4k. Let em know what you guys think, I really feel I need to improve the texture, any suggestions?
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.