Model Optimization Thoughts

Discussion of gameplay issues and experiences, game balance and suggestions for tweaks and modification of the existing systems.
Post Reply
User avatar
Max
Posts: 3
Joined: Wed Jul 30, 2014 4:45 pm
Location: Great Lakes Region
Contact:

Fri Aug 01, 2014 9:54 pm

I took a look at the player models today. (I was thinking of popping them into ZBrush and seeing if I couldn't come up with some additional armor sets or maybe even a robe or cloak that I could bind to the tail joints). I personally like the models, for the most part they are well done.

I was surprised that the poly count was ~5500 with 72 joints (+- it was a quick count), YIKES! :o Add to that items, and clothing/armor/hair multiplied by 20-40 models and that's gotta be a heavy load on the game engine. I'm impressed you are getting as good a frame rate as you are! 8-)

As an artist, I love that the models are so detailed, having all their fingers and toes, fairly developed ears and all, But in the interest of maximizing the fame rate and maybe allowing for a higher model count, I have a couple of suggestions to consider.

Take the four smaller toes and "sock" them (merge them together). Many player character's are going to be wearing boots after the 1st few days, so it shouldn't be glaringly noticeable, and who pauses to examine starving bandits feet? It should also save ~300 polys per model, while that doesn't seem like much, it adds up with larger team/mobs. You could probably gain a few more polys from flattening the ear insides as well. Once the normal maps are created for the ear detail, the lower polygon detail probably won't be noticeable. Again, ears are a target optimization area because there is a good possibility that they will be covered by hair or helmets.

A few more polys could probably be carved out in the neck, etc.

The bottom line being, the game performance could probably benefit from a poly diet without compromising artistic integrity too badly ;)

Likewise joints/nodes can significantly impact engine performance as well. Consider cutting a few from areas such as the tail, and maybe combine the last fingers into one joint chain instead of 2? They would move together but with the thumb and forefinger moving separately, I think the illusion of being fully functional would still be there, especially at a distance.

I'm not sure of the design plans for having 2 pigtail joints. Maybe it's going to be something WAY-COOL that I would love, but at a quick glance - couldn't you use one joint chain instead of 2 and simply rig both pigtails to the same central joint chain?

Anyways, food for thought! :mrgreen:

User avatar
Captain Deathbeard
Developer
Posts: 1831
Joined: Thu Jul 03, 2008 12:04 pm
Location: UK
Contact:

Tue Aug 05, 2014 6:09 pm

Already planned actually, not poly reduction but proper LODS, and the face and hand bones are going to be taken out completely and replaced with morphs. Bone count has been identified as the major bottleneck, polycount isn't too much of a problem.

Revage
Posts: 67
Joined: Sat Mar 23, 2013 6:49 am
Location: Coos Bay, OR

Tue Aug 05, 2014 9:50 pm

Reading this made my brain hurt. I like turtles.

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 7 guests