323 replies to this topic

[historic_bruno]

Wijitmaker, on 17 March 2012 - 03:35 PM, said:

IMHO an IK/FK rig isn't required. I'm assuming that whoever has been rigging and animating the blender animals in the game was doing so without one.

Actually, I believe Zaggy has started using IK on his animations. It might be slightly more complicated but apparently leads to better results.

[Gen.Kenobi]

I don't know...
I'm really out of free time to take a more in depth look into this - [just got home from school (20pm). Tomorow needs to be there at (9am).]
I personaly work better with IK/FK solvers, but that's just me - I really feel that they make the animation more "realish" and "fluidish".

Peharps, Jason, If you get the time, maybe you could take a look in Blender? there are some prety foward begginer tutorials that may help you get started:
http://sagefans.net/
And related to Blender Rigging: http://www.blendergu...tion-to-rigging

[Wijitmaker]

Quote

Actually, I believe Zaggy has started using IK on his animations.

Ok, that is good to know. Maybe I could work with Zaggy to get something working in Blender then?

Quote

Peharps, Jason, If you get the time, maybe you could take a look in Blender? there are some prety foward begginer tutorials that may help you get started:
http://sagefans.net/
And related to Blender Rigging: http://www.blendergu...tion-to-rigging

Thanks man! I'll check those out hopefully by the end of the month.

[Zaggy1024]

Wijitmaker, on 18 March 2012 - 02:33 AM, said:

Ok, that is good to know. Maybe I could work with Zaggy to get something working in Blender then?

Well, go ahead and tell me what I need to do!

I've already tried importing one of the biped animations into Blender, successfully got it into the engine, but the problem is it turned out totally messed up... The problem, I believe, is that the current Collada importer in Blender (and the one in 2.49 too) imports the bones extremely inconsistently, sometimes rotating them to be totally horizontal or totally vertical, and sometimes appearing to put them in the right position... I tried joining all the important bones to the ones they were supposed to be connected to but it work.

You could export the skeleton to the BVH format, and let me try to import it. (I imported one of the motion capture files in the art repository, and it looked OK, but the structure of the skeleton was wrong)


About the IKs, it worked fine when I baked the animation to remove the constraints on the zebra (historic_bruno's idea ).

[Mythos_Ruler]

There has been some discussion on IRC of just making all-new dude and dudette meshes. We already have a new horse mesh.

[Wijitmaker]

Michael it is interesting you posted that comment today. It is your choice, of course.

Coincidentally, I met a guy today at a Microsoft store (of all places - first time I've been in one) who is an animator that works in a variety of different 3D programs. He said he would be happy to take a look at the 3ds Max dude and get it into Blender. I'm going to send the files to him tonight to see what he can do with them. I was excited to come home and share this news with you all as I know you are very interested in this.

A request though... could we move this topic to the development forum so that he could catch up on the challenge?

[Mythos_Ruler]

The reasons for making new dude meshes are many-fold.

Compatibility:
The previously mentioned Blender compatibility. Compatibility is one of the major issues holding us back in the animation dept. It makes it difficult (i.e. impossible) for new talent to jump in and begin making animations for our existing units.

Unit identification:
A wider array of meshes, representing non-armored and heavily armored dudes, would give us distinctive "silhouettes" that would assist in unit identification from max zoom. This one is important and difficult to achieve with the current meshes which all have the exact same silhouette. Basically, what we'd do is have a set of "skinny dude" meshes for archers, slingers, skirmishers, and then a set of "beefy dude" meshes for heavy infantry. The proportions of the new dude meshes would conform more closely to an "ideal" artistic human proportion. Lastly, heroes would have a set of meshes with "heroic" proportions. They would be slightly taller and slightly more exaggerated, but not overly so. The key words are "exaggeration" rather than "cartoony." Basically, the design aesthetic would be less cartoony and more "stereotypical." We stereotype ranged units as skinny weaklings, while melee units are beefier and more muscular. And consequently, heroes are beefier than that, but not freakishly so, just enough to give each set a distinctive silhouette from a distance. A big thing that would also help this would be to have a specific set of idle animations for each type of unit. Right now they all use the same set of idle animations, which makes them less distinctive. So a combination of "stereotypical" meshes and new sets of idle animations for different types of units would make our units look 10x better and even assist in making gameplay better.

Skinny, non-armored, ranged units:


Beefier, heavy infantry ("Ideal") proportions, and then "Heroic" proportions for heroes:

[Wijitmaker]

What I'm attempting to help you with isn't going to stop you from making a new mesh. You can still do that. I'm trying to preserve the dude skeleton, so it saves you guys from remaking the 50 plus humanoid animations.

[Mythos_Ruler]

Wijitmaker, on 30 March 2012 - 05:16 AM, said:

What I'm attempting to help you with isn't going to stop you from making a new mesh. You can still do that. I'm trying to preserve the dude skeleton, so it saves you guys from remaking the 50 plus humanoid animations.

Yeah, that's fine, brosephus. It might be possible to just edit the meshes and keep the existing skeletons. So, please continue.

[Shield Bearer]

Mythos_Ruler, on 30 March 2012 - 02:50 AM, said:

The reasons for making new dude meshes are many-fold.

Compatibility:
The previously mentioned Blender compatibility. Compatibility is one of the major issues holding us back in the animation dept. It makes it difficult (i.e. impossible) for new talent to jump in and begin making animations for our existing units.

Unit identification:
A wider array of meshes, representing non-armored and heavily armored dudes, would give us distinctive "silhouettes" that would assist in unit identification from max zoom. This one is important and difficult to achieve with the current meshes which all have the exact same silhouette. Basically, what we'd do is have a set of "skinny dude" meshes for archers, slingers, skirmishers, and then a set of "beefy dude" meshes for heavy infantry. The proportions of the new dude meshes would conform more closely to an "ideal" artistic human proportion. Lastly, heroes would have a set of meshes with "heroic" proportions. They would be slightly taller and slightly more exaggerated, but not overly so. The key words are "exaggeration" rather than "cartoony." Basically, the design aesthetic would be less cartoony and more "stereotypical." We stereotype ranged units as skinny weaklings, while melee units are beefier and more muscular. And consequently, heroes are beefier than that, but not freakishly so, just enough to give each set a distinctive silhouette from a distance. A big thing that would also help this would be to have a specific set of idle animations for each type of unit. Right now they all use the same set of idle animations, which makes them less distinctive. So a combination of "stereotypical" meshes and new sets of idle animations for different types of units would make our units look 10x better and even assist in making gameplay better.

Skinny, non-armored, ranged units:
human_figure_8_heads_tall.jpg

Beefier, heavy infantry ("Ideal") proportions, and then "Heroic" proportions for heroes:
prop_var.gif

Its too detailed of course, since it wasn't made for 0 AD. But I could get the poly count down by quite a bit, I believe

[Mythos_Ruler]

Not bad! Doesn't need fingers, so you can cut quite a few polys by removing those. Also, doesn't need a head, since heads are props.

[Shield Bearer]

Alright, what should be the max poly count?

[Shield Bearer]

Hmm, is it true that our current models are a thousand+ tris?

[Wijitmaker]

The count depends what your talking about... just the base model or a model with all the props and stuff?

[Shield Bearer]

I got the poly count down to 708(tris, not quads). Is this the best place for this discussion?

EDIT: @Jason: The base model only. I checked a model with a sling, so I didn't take that into account.

[Mythos_Ruler]

A fully propped dude currently in the game goes up to 1200 tris. If we go a couple hundred above that with any new body meshes (+props), we wouldn't have any problems.

[Shield Bearer]

So 700 is good?

[Mythos_Ruler]

700 is good. What knocked it down by 500? Removing his fingers and head?

[Ykkrosh]

Mythos_Ruler, on 30 March 2012 - 06:15 PM, said:

A fully propped dude currently in the game goes up to 1200 tris.

Prop polygons are cheaper than skinned polygons (since we don't have to run the skinning code for them), so we might need to be a bit careful here. It may help if someone could create some animated skinned demo units with varying triangle counts - the current dude.pmd looks like 246 triangles, so a range of like roughly 500 and 2000 and 5000 could be useful for testing. They don't need to look any good (they can be made with automatic smoothing or tessellation etc) and the animation doesn't need to be any good, but they should have a realistic number of bones (which should be as few as possible) and should be realistically skinned (in terms of number of bones influencing each vertex, in particular), so we can stick a few hundred units on screen and see at what point performance becomes a real problem on modern hardware.

(Incidentally, it's best if new meshes are fully closed (they don't have any holes in e.g. their necks or the bottoms of their feet to save a few polygons - I think some of the current ones do that), since closed meshes can help with some stuff like shadowing.)

[Shield Bearer]

Mythos_Ruler, on 30 March 2012 - 06:33 PM, said:

700 is good. What knocked it down by 500? Removing his fingers and head?

No! That model which you saw was made up of quads So it was actually 2k+ tris. I brought it down by 1500 I think by merging unneeded vertices

So, yeah, Philip I got a 2000 tris model and now a 700 tris model and I can smooth it to get a 5k one if someone is willing to animate it