Forum - View topic

Koei Tecmo's Berserk Game Launches This Fall in the West

At the speed its coming out I doubt it will get an English dub

I wonder how the US voice acting is going to be, in all sorts of ways. The publisher(s) could technically use the existing film / tv cast and fill a few empty chairs on the own, but Musou is infamous for their VA decisions. Not that i will use the US voices.

As much as I dislike musou games, I must say the game excels on the art department; I wish they used these models for the current "anime" instead of the bland ones that attempt to emulate a sort of drawn art style and fail miserably at it (though they're somewhat better than their previous attempts).

If you're going 3DCG, do it correctly, the game itself prooves you can do it, come on...

chronos02 wrote:

As much as I dislike musou games, I must say the game excels on the art department; I wish they used these models for the current "anime" instead of the bland ones that attempt to emulate a sort of drawn art style and fail miserably at it (though they're somewhat better than their previous attempts). If you're going 3DCG, do it correctly, the game itself prooves you can do it, come on...

The CG models have to match the hand drawn animation, or match as closely as possible. you know when they show close ups, they have a range of facial animations that I doubt he rig is capable of.

People are more tolerable of these models because they are in a game where they match up with everything else. If they were actually in the Anime, people would hate them a lot more than the current models.Trust me, you would not have a single word of praise for them if they were in the Anime.

Unfortunately that is absolutely not the case, I work as an artist and rigger, and what you say makes absolutely no sense. A rig, in the first place, is not done inside a game engine, it's done within Maya/Max/Softy/ and why not, Blender. There you make a rig which allows a massive number of facial expressions, almost unlimited, with small details that can be transfered without problems to a game (within the game engine though, mapping all those blendshapes can be a hell, moreso if one tries to mimmick a true facial expression and not an aproximation). So you basically do a rig that allows for many, many expressions, but then you use only those the animation director aks you or ultimately the art director or even the director for the project; the fact you see fewer expressions inside a game doesn't mean the rig didn't allow for the full range.

And the sentence "3DCG models have to match the hand drawn animation" is all but nonsense, simply by changing the shader you can achieve that "hand drawn animation" style you talk about, after all, it is but the shader that makes the anime look animeish, and nothing else (and maybe a postprocess). In this case, the 3DCG anime doesn't seem to be using as much detail as the game is using, such as the different maps the game has shown us: albedo, specular, gloss, diffuse, heightmaps, AO, normalmaps, etc. plus a PBR shader which makes the final result something short of amazing (for the budget and time they've had to make it at least).

And you say I should trust you on a matter regarding the use of far better models and shaders would not yield better results and thus praise than the current ones? Well, I must decline, I do know the amount of animations within the game are fewer by far than those in the anime, BUT, I was not talking about that, they should have simply used the models and shaders from the game and make custom animations for the anime (not like the current ones aren't custom made, after all they anime uses only cusotm animations for each scene, safe for, maybe, the background characters). The amount of praise in such a case would have been far, far greater than the amount it has accomplished with its current models and shaders. Still, it is true that the retopology used in the game might lack a bit of detail, but that's not a problem, since adding a few more subdivs on the retopology should solve the problem (it actually takes a bit of work, it's not a simple subdiv...), after all, what's important in the end is the high poly model done in Z-brush or a similar program for high poly sculpts, from which we extract all the maps (aka textures) to use on the final model (aka the retopology of the high poly model).

And well, I almost forgot, but the current shader is not enough to get that "drawn art style", well, no problem! adding but a simple paragraph of code would be enough to achieve that result though.