Originally posted by Lifthz
Yeah. Nothing LOD can't handle though. Also again, it's not like the whole model is looking that good.
Alright. Again, maybe true. But how do you know this yet?
If you've noticed MoH, there are alot more use of pre-rendering, especially when you notice the latest MoH: Rising Sun (thats the console version, not Pacific Assault..) trailer's beginning is all pre-rendered.
Originally posted by Speculator
That is a PRE RENDERED face. The lighting engine of any game, its anti aliasing, etc would not allow any in-game polygonal structure to be able to stand up to that quality. As a modeler myself, that face is more than any computer during gameplay can be capable of..even if using LOD models (most LOD systems are based on multiple models)
I don't see as an MoH fan that the game has absolutely no prerendered parts, when you can clearly see the opening scene in Allied Assault is ALL pre rendered
If THATs in-game, then I guess EA's MoH team dumped wanting the most for every audience and went for total next-next-gen graphics which only the highest end system could run on.
please note no Quake 3 was mentioned!!! (not even the slightest bit is there...) he also states its called MOH Tech. Oh... and once and for all IT'S A RENDER PEOPLE!!!!!!!!!!
)Why wouldn't they? Seems to be working on you. If they did go out of their way to make that really detailed face in Maya (or whatever), then you would fall for it and buy a crappy game simply because you believe they'd never do that. Little ironic, huh?Originally posted by commando
they wouldn't jus go out their way, make a really detailed face, show it to everyone and say, here is a soldiers face you will not be seeing ingame.
trust it will be ingame face
http://www.trepid.net/images/gl_wireframe_2.jpg - Yes it does.
Ahem, dynamic scaling of polycounts can work in-game. It was working in TF2 years ago... but when they bumped up the polycounts of the models (when moving over to Source) the calculations needed to lower polycounts on-the-fly took more time than the technology saved on rendering... so they scrapped it.
A complex model that is rendered does not mean that it is displayed on your screen. There is a difference between the calculations done to render a model and the calculations done to render a frame. At 60 fps, the scene is not rendered 60 times per second, but displayed 60 times per second. The scene is rendered once, and then the calculations done to display the scene 60 times per second. The resolution that you're running in determines the complexity of rendering a frame. If you are running at 1024x786, then 60 frames containing 1024x768 pixels are rendered. If you change the position of the camera, the entire model is not rendered all over again in a different position. The model has already been rendered; it simply has to calculate the new way to display it for the next frame.
If that is so, then why in this screenshot are there areas rendered that the player cannot see from the place that he is standing? If that scene could render as fast as you say it could, then why is it being rendered when the player cannot directly see it? Wouldn't it reduce a bit of CPU usage if it wasn't rendered now, and only rendered when the player moves outside?
A model does not have to be re-rendered when animated. The model has already been rendered, the positions of the polygon's vertices are just moved in accordance with the camera. It's just like the same thing as the camera moving.
Originally posted by commando
welll i gotta admit the MOHPA soldier face is looks mighty fine! i mean the dirt on face, the teeth! sooooo cool.
prob using the same BRAND new engine so hay both same i say.
(but hl2 have the better ANIMATIONS and better all around
Originally posted by DelaZ
ok.. but the rest looks good.. and the mouth.. with those teeth.. look better then HL2.. me thinks.
Don't get me wrong.. im a big fan of HL2. But i hate those fanboys who you can't argue with... CAUSE HL2 ownz teh moh tech in the hizzle!
Just trying to start a conversation here
Originally posted by nØthing
Why wouldn't they? Seems to be working on you. If they did go out of their way to make that really detailed face in Maya (or whatever), then you would fall for it and buy a crappy game simply because you believe they'd never do that. Little ironic, huh?
Trust me. You're not going to see that in-game. That face was not rendered using their engine. It was rendered using Maya or Max or some other modelling software.
The g-man in the tech demo was rendered using source. That was the whole point of the video. They were showing off what the source engine is capable of rendering. This is what the source engine is more than capable of rendering (Also note that this is an old screenshot. It'll be even better than this on release). It can also do this.
Now, compare this 'in-game' render with another one. The second doesn't even look mildly as good.
And don't give me this 'LOD' crap. Are you saying that if you were inside a building, the characters and building will be rendered using super high-poly models. Then, if you went outside, the scenery outside would be rubbish, using obvious sprites for bushes. Also, if the characters were to follow you outside, they would now look awful because they have been re-rendered using lower poly models. It doesn't work that way. Ever heard of a little thing called vis?
The entire 360 degree scene around you is rendered in one go. If you quickly turn around, the engine is not constantly rendering new models in a futile effort to keep up with you. The stuff behind you, including all the player models, have been rendered already. If you turn around you are simply rotating a virtual camera. The engine doesn't have to keep rendering everything new that the camera can see. If it had to do this, there is no way any game would render anything in time for you to see it when you turned aroud. Complex models are not just rendered instantly, you know?
As well as the entire scene around you, the scene and players around the corners from your position are also rendered. This way, if you quickly turn a corner, you are not going to witness the new scene rendering infront of you. However, the engine will now be rendering the next scenes that you might stumble upon. This means that by the time you get to the next corner, the scene around there will be fully renendered and you can step around it without seeing a void where the engine is quickly trying to render the scene. This is basically how the vis system works for pretty much all 3D rendered games.
For example, this might be what the player can see, but this is what the engine has rendered.
Now, by your 'LOD' system, if you were to look at a character from a distance, he would be rendered using a low-poly model. Are you saying that if you were to walk closer to him, he would be instantly re-rendered using a higer-poly model? I don't think so.
Conclusion: It's a pre-render for the purpose of fooling people into thinking that the engine is much better than it actually is. They won't tell anyone that it's a pre-render, but the wont tell anyone that it's an in-game shot either. Nice and legal.
EDIT: Damn colours.
That's not my screenshot. You think I own http://www.trepid.net ? lol - I wish.
