A little more info would be helpful. Besides xNormal what software are you using. You should have a 3D modeling package at the very least, and probably a sculpting package. What is the ‘mesh’ for? Is it an environmental piece, interactive, animated or static? Character or equipment pickup?
Character models in particular can take a considerable amount of time, especially if they assume a prominent role in the game and have up close shots with dialogue.
If I’m working for someone else and they set a deadline of 2 days from now, I can get the character done in that time frame, but there will be some areas where the character is iterated less, and isn’t as polished as it could be with more time.
Characters are just like anything else. Setup a schedule for yourself and then you can manage how long it takes to create the character. For instance if you give yourself 2 days (20 hours) to create a general character to completion, you can plan accordingly. Base modeling 4 hours, UV’s 2 hours, Rigging & Skinning 4 hours, that’s one day right there. So you have 10 more hours to play with on Sculpting, Texturing and Animating.
These steps aren’t in order but you get the idea.
Indie developers have even more things to consider, because we are trying to create complete games. So maybe in a 3D game I’m creating with 4 characters, (example Left 4 Dead) I will give myself 2 weeks (80 hours) to complete the character assets. That timeline might be tight, but I have other things to consider if I ever want my game to be completed.
Scope based on how much of a role will the character/object play in your game. If this mesh you speak of will be the central point in the game that is viewable at all times from all angles, then you want to spend more time on it to make sure it looks perfect. If it’s a set piece that is seen several times throughout the game but is basically background filler, spending less time on it is better time management.
I think scoping your art based on how complex you want your game to be is good practice. You can also do the same planning process in reverse, scope the complexity of the game based on how beautiful and detailed you want the game assets to be. But if your game is beautiful work of art but limited in game play - it may get some negative reviews. Same result if the game play is great, but all assets looks rushed and lack quality.
I think there is a happy medium for this type of planning.
If I’m going for a complete high resolution (main character) this is my preferred pipeline.
- Create a low poly model in Max with generated UV’s.
- Pull low poly into Mudbox and sculpt away for days.
- Once high res sculpt is complete, generate normals, AO, for low poly character.
- Start/finish mapping textures - this could take days depending upon quality desired.
- Once textures are complete - test export into Unity and setup shader/material to confirm it looks awesome.
- Iterate as needed on textures.
- Rig and Skin - could take a while depending upon complexity of the character and extra components.
- Test Export to Unity with a random animation to confirm all rigging and skinning is tight.
- Iterate as needed on rig and skin weights.
- Morph or bone based phonemes and visimes for dialogue creation.
- Test export with extreme limits on facial gestures
- Iterate as needed
- Other areas to consider, cloth, fx, physics controlled parts, rag-doll, IK controls
- Start animating - high quality custom animations so the character looks and acts like no other that has come before.
1 high quality character nearly complete - on to the next one.
But before that - there is a lot of setup to do in engine before this character is ready for the spotlight.
A lot of artists jump straight into sculpting without creating low poly models first and then retopologize a low resolution mesh from the high resolution sculpt. This is a matter of taste regarding low to high, or high to low.
Hide the seams as best you can. On a character seams are created in places where they are less obvious, on the inside of the legs, back side of the head with the face and head being one complete shell and the neck seam being below the shirt line, under the arms and down the side of the body. And the less seams you have the better, but it requires some good unwrapping skills.
Keep in mind normal maps can also be generated from a 2D grey scale bump map. For lower resolution characters, maybe secondary characters who don’t have “speaking rolls” this is perfectly acceptable, though most character artists frown upon this technique and say the normal maps aren’t as accurate as they would be if generated from a high resolution sculpt. I see no difference once inside any game engine and neither will any players who play your game. Look into it.
Hope this helps.
