Out of curiosity, would it be easy to upgrade the graphics if someone say ported a...

[DeadlyNinja]

...PS2/GC/Wii game to the 360? We're talking about completely redoing/upgrading the graphics to 360 standards. Would development cost just as much if you ported the game with enhanced graphics as well? Is porting with enhanced graphics just as easy as adding more polygons and more shader effects?

 

[Shifty Geezer]

Would development cost just as much if you ported the game with enhanced graphics as well?

Pretty much. content is the number one expense. You can get a few people to create your engine, and porting a last-gen engine that isn't going to stress the CPU means devs can be pretty 'sloppy' about it (and we all know what a lazy bunch of wasters devs are... ) making it even cheaper to create the engine.

Now if the last-gen game had high resolution assets from which the low resolution in-game assets were created, it'll be cheaper to use them and retarget for next-gen. But I doubt many games are in that position.

 

[DeadlyNinja]

Ok, but is it easy to do? By easy, I mean actually putting effort into making it up to the 360/PS3 standards.

 

[Shifty Geezer]

Well...um...as easy as it is to create gigabytes of models and textures and such. For artists who do this thing for a living, it's easy enough, but it's time consuming and expensive!

 

[Blackraven]

Need for Speed Carbon comes into mind (that game released on all next-gen consoles along with the PS2 release).

 

[cloudscapes]

case per case, I think.

in my experience, most artists on our teams kept PSDs of the textures that were double or quadruple of the intended final TGA/DDS resolution. so if for a PS2 game, where we couldn't really go above 256, a large number of the textures were created as 512 or even 1024. this makes porting quite a bit quicker if the higher res versions are at hand. for a decent texture artist, it doesn't really take that much extra time to make a 1024 than it does a 256, if it's a non-specialized texture (like a character or specific design)

models are a bit harder. you cant simply put a meshsmooth on everything and re-export. it requires a bit of work, though it is possible. tescelate here, chamfer there, etc. characters are harder though, because chances are you kill the skinning and have to do that part over. I'm not 100% sure though.

if the engine has procedurally instanced content, like vegetation, decals, detail grass, particles, then you can simply up the number or max distance

 

[Sonic]

It really depends on the game and if the engine needs to be overhauled to work on the new system.

I agree with other posters here that the content/art would be the most expensive part of the upgrade. We should also take into account that we would want better animations, and that layers some costs onto the production. Would new effects not possible before on the last gen consoles be used in the conversion of the game? If so then there's even more tweaking that needs to be done. The list goes on and on.

It's like a racing simulator that has the Laguna Seca track in it. Do they use the original model for it they had in previous games or do they model it from scratch all over again?

 

[Akumajou]

...PS2/GC/Wii game to the 360? We're talking about completely redoing/upgrading the graphics to 360 standards. Would development cost just as much if you ported the game with enhanced graphics as well? Is porting with enhanced graphics just as easy as adding more polygons and more shader effects?

Why "port", would it not be better if the game engine was completely reprogrammed for the higher spec console hardware in question?

Bugbear made their Flatout games (1 & 2) for PC, PS2 and XBox 1 and recently they announced or released info saying that their next Flatout game would be made exclusively on XBox 360. Given that console's assumed dev tool maturity I would hope that they (Bugbear) are able to show us what they are capable of, then again I really don't know how much time they invested in making that game on the 360 but AFAIK the game is not called Flatout 3...

If what I am saying is what you were trying to say then in my humble opinion I feel that completely reprogramming a 3d engine to a higher spec game console like Playstation 3 or XBox 360 (I would prefer one platform so that the dev can focus more) while at the same time making use of the extra hardware power/memory to implement A.I., physics, shaders, etc in case the game engine did not use any of those extra features, I feel that it would be easier for a game dev to AFTER learning about the new console hardware to tear down and rebuild their old 3d engine from scratch as they already know what their previous game featured and was limited to but although I feel this is great for some game series, it all depends on the game developer to actually do this and while they would be reprograming the 3d engine it could almost be seen as making a new engine really that their series will benefit from.

Examples I can give is SEGA AM2 with Virtua Fighter, going back over 10 years ago, all the way to today and they have torn down and rebuilt or better yet reformatted their old 3d fighting engine several times to its current level of being able to run arcade perfect on PS3 thanks to their arcade hardware using a lower spec Nvidia GPU in comparison to RSX, the XBox 360 port only really benefits from this as well although that console uses an ATI GPU mainly due to being modern as well.

Another example would be Tecmo's Ninja Gaiden Sigma on PS3 as opposed to where the game originated from on XBox 1, however since Tecmo's dev team had as far as graphics go from Nv2X level graphics to the G7X/RSX level mainly for easier implementation of 1080P since both graphic chips are Nividia gpus, however I really cannot comment much on that game until all of the differences/upgrades (graphically, etc) have been documented, I can tell you that to me this example must have seemed easier for Tecmo to upgrade their old game even though they would only be using a puny fraction of what Cell+RSX can do.

When I think of that "porting" Tecmo did I could compare it to, lets say Nintendo remaking/porting Legend of Zelda III A Link to the Past on NeoGeo's Gold cartridge based console as an upgrade over the SNES since they both are 2d consoles.

 

[TheChefO]

It's like a racing simulator that has the Laguna Seca track in it. Do they use the original model for it they had in previous games or do they model it from scratch all over again?

This is the biggest waste of resources we currently have going on in the games industry. Modelling and remodelling the exact same assets for different companies and different hardware.

The same cities and roadways and vehicles have been modelled to death, but still will need to be done again and again come ps4 and xb720.

A 3rd party should model these popular assets at photoreal (using nurbs, b-splines etc) detail levels and push them down to whatever detail level necessary by the dev for the specific hardware and game.

For fantasy based games, sure this effort would be useless. And for the large portion of games which are based on reality (racers, tact shooters, gta types etc) this effort would be a huge undertaking and very expensive initially, but the end result will be assets which can be used from now til 20 years from now for every game and system that would benefit from such real world based assets.

With modern capture techniques and devices becoming more affordable and better equipped to gather this data, perhaps this dream will one day become a reality.

 

[Xenus]

That reminds me of something I read a little ways back in that the GT designers were trying to make deals with car companies related to getting a hold of the models they built for the production process.

 

[ERP]

That reminds me of something I read a little ways back in that the GT designers were trying to make deals with car companies related to getting a hold of the models they built for the production process.

In a previous life, we actually did get CAD models from manufacturers for several of the cars we used in a racing game. But they we're largely only useful as reference. It's nice not to have to guess at dimensions from pictures, but the CAD models aren't really intended for rendering at all.