D
Deleted member 86764
Guest
Maybe we need another spin out thread, this time suggesting that a 2D crowd is definitely better than 3D. Maybe another that says sprites are more computationally expensive than dynamics lighting.
Freeing up the Kinect Reservation *spawn*
- Thread starter Starx
- Start date
- Status
Maybe we need another spin out thread, this time suggesting that a 2D crowd is definitely better than 3D. Maybe another that says sprites are more computationally expensive than dynamics lighting.
Maybe someone just needs to understand that billboards are not the same as 2D sprites
and feel free to spin off on your opinions, I only pointed out that texture asset saving is not the reason here, unless you feel like arguing that real photo takes up less memory than junk quality textures usually seen on the crowds. You are also assuming that quads are not lit, but they actually are.
Last edited by a moderator:
2D or 3D can take up more RAM if you spend more on the resources. 5 duplicated 3D peeps with 256x256 textures take less RAM than 1000 individual 1024x1024 photo sprites. While a mix and match of 30 body parts each with a 1k texture takes more RAM than a few dozen low-res photo sprites. I have no idea what the argument is here, nor what anyone's trying to price! 
D
Deleted member 11852
Guest
Yup. Unless you have resource to waste, like MLB 2014.
D
Deleted member 86764
Guest
Help me understand your opinion, as far as I can tell you're suggesting that 2D crowd is better than 3D because it's simultaneously better to have photos rather than a 3D render while it's also illogical to waste resource on them? And that 3D models look like zombies...
Why stop there, maybe we should have 2D trees and all other environmental details. I mean it's pointless having them there - you're only ever looking at the road anyway.
D
Deleted member 86764
Guest
I guess Mortal Kombat 2 is the peak of graphical prowess.
It's a shame that Eurogamer used a foucosed replay camera to capture an ugly screenshot to feed the trolls across the net. In forza if you use zoom in Photo Mode the background 2D trees became bigger and bigger like what you can see in this screenshot. During gameplay you can't see this huge ugly three at all because it's far far away from the game camera and far far smaller than what you can see in your example.
Also Photo Mode won't change the textures or other part of the image. In photo mode you will have the highest polygon model (interior details when you are on the ciruit) and it also uses better AA. He talked about textures and I show him some textures.
I know how it looks during the gameplay. It looks silky smooth and fantastic for what it is (a lunch sim racing game with 1080p@60fps locked and much other stuff).
D
Deleted member 86764
Guest
I happen to think Forza looks very good, but that's not to say it couldn't be 'better'. I genuinely want Microsoft to do better than they currently are, freeing up some of the Kinect reservation is another step in the right direction.
I read through the dumbness of this thread and nowhere does he say that a 2D "billboard" crowd is better. All he said is that it probably isn't done that way to save on texturing compared to a 3D crowd. He called out people for calling them sprites, because they are not sprites.
What I suggested is that it's just "cheaper" on the poly count than trying to get the same kind of look with 3D meshes, especially when the crowds in racing games are mere backgrounds, where it's not logical to spend that kind of budget on things that you'd just swoop by anyways.
In case you haven't noticed, racing games do use "2D" trees and other stuff like grass, and I'm not necessarily talking about Forza. There are also LODs that uses billboards as imposters when objects are far away.
Last edited by a moderator:
As no-one uses sprites as they originally were, and all sprites now are 3D quads (certainly in 3D engines), the use of the term sprite as a billboarded quad should be accepted as a convenient vernacular.
Sure, except at times the context is suggesting that the use of "sprites" are done in the traditional 2D sense (see ref:MK), as if they are done in the same way in while missing exactly 1 dimension.
Hence I don't like the dilution of the terms. And there is a difference on quads that face the camera always (billboards) versus those that don't in 3D scenes, versus just quads that layered onto the frame buffer like UI elements.
Last edited by a moderator:
- Status
Similar threads
- Sticky
