Unreal engine 3 on Wii... possible?

Status
Not open for further replies.
Probably upscaled or something like we've seen with some Red Steel screenshots. I do think such gfx should be possible on the Wii, it doesnt look that good. Not exactly high res textures or high poly models, only nice lighting and shadows.
 
The value of the UE3 is not just in its renderer, but also (some would say mostly) in its toolset. So even if you heavily downgrade the renderer, if you use the UE3 toolchain to author your game, you'd be better off than if using UE2.

Well if you optimize it you may get "high end" performance.

It will cost you though.

How about if you...

1)scale to 32fps
2)or scale to 28fps
3)scaling off-screen objects
4)scaling on screen objects
5)scaling the draw distances
6)USB GPU like Asus XG Station+ Hollywood(latency issues?)
7)USB Harddrive(20GB, $60 retail)
8)Use of Displacement and parallax maps.
9)HyperNURBS
http://www.creativemac.com/2004/02_feb/tutorials/koc4ddisplace0402242.htm
10)sub-polygon displacement
http://www.c4dcafe.com/reviews/R9/r9p4.html
11)PowerNURBS
http://digitalproducer.digitalmedianet.com/articles/viewarticle.jsp?id=33973

Certianly there are options. Its just what company(nintendo or any other)are willing to throw money at the problem.
 
Last edited by a moderator:
Well if you optimize it you may get "high end" performance.

It will cost you though.

How about if you...

1)scale to 32fps
2)or scale to 28fps
3)scaling off-screen objects
4)scaling on screen objects
5)scaling the draw distances
6)USB GPU like Asus XG Station+ Hollywood(latency issues?)
7)USB Harddrive(20GB, $60)

Huh? This didn't make any sense...
 
Probably the only reason you want UE3 on Wii is because its easier to port games. I believe gearbox said that before. I have no doubt that UE3 will run on Wii, the question is actually how much of UE3 is left after ripping our everything the Wii cant handle and if you can still call it UE3 after that.
 
Probably the only reason you want UE3 on Wii is because its easier to port games. I believe gearbox said that before. I have no doubt that UE3 will run on Wii, the question is actually how much of UE3 is left after ripping our everything the Wii cant handle and if you can still call it UE3 after that.

How about use a external GPU like I posted above?
 
Is that going to come free with your UE3.0 game? :???: Even if a USB GPU was viable, which it isn't due to bandwidth and latencies (that ASUS XG station connects via Express Card Connector = PCI Express), you can't write UE3.0 for Wii on the premise that the users have an extra add-on, without limiting yourself to that addon. Otherwise all those UE3.0 games will need people to shell out $100+ or whatever on a GPU upgrade that'll see nonimal adoption. It'd be cheaper to release a Wiimote for XB360 than Xenos for Wii!
 
Probably the only reason you want UE3 on Wii is because its easier to port games. I believe gearbox said that before. I have no doubt that UE3 will run on Wii, the question is actually how much of UE3 is left after ripping our everything the Wii cant handle and if you can still call it UE3 after that.


No, the question is actualy how much of UE3's benefits the developer would have retained, producing art assets with a good toolchain and levels with a powerful editor instead of developing both in-house, and spending some time to fit the UE3 runtime in the Wii's limited memory and to rip out the renderer.

UE3 has a reputation of being hated by the programmers, but loved by the artists, and being fairly bulky in terms of memory use. The answer to the question whether there is any sense in trying to get it to run on the Wii is not as clear-cut as it seems.
 
Is that going to come free with your UE3.0 game? :???: Even if a USB GPU was viable, which it isn't due to bandwidth and latencies (that ASUS XG station connects via Express Card Connector = PCI Express), you can't write UE3.0 for Wii on the premise that the users have an extra add-on, without limiting yourself to that addon. Otherwise all those UE3.0 games will need people to shell out $100+ or whatever on a GPU upgrade that'll see nonimal adoption. It'd be cheaper to release a Wiimote for XB360 than Xenos for Wii!


The Wiimote a nintendo product cannot be made for a microsoft product.

Well we can hope for Wii2 to have decent hardware.
 
First of all, why bother? As I understand it, the Wii doesnt support pixel shaders via hardware, making one of UE3's huge assets useless. Wouldnt sticking with UE2 be a much better option?
 
First of all, why bother? As I understand it, the Wii doesnt support pixel shaders via hardware, making one of UE3's huge assets useless. Wouldnt sticking with UE2 be a much better option?

1) because of the tools
2) Wii and GC does have pixel shaders in its HW, just not equals to DX ones (or nearly as good as what it asks on the PC), you may have a bigger problem on the vertex shaders
3) if you realy want to use Wii/GC potential, neither you made your own engine, not that many other engines wouldnt be better than UE2/3...
 
1) because of the tools
2) Wii and GC does have pixel shaders in its HW, just not equals to DX ones (or nearly as good as what it asks on the PC), you may have a bigger problem on the vertex shaders
3) if you realy want to use Wii/GC potential, neither you made your own engine, not that many other engines wouldnt be better than UE2/3...

Well there can be some options but most of them be a longshot.

If you use UE 3.0 you can(maybe) I guess you would need to optimize this to high latency. Your going to have to just live with the higher latency.I assume a Hollywood running 28fps it will be less of a problem.

1)Compress(into a .dol or a .lzo file) the contents of the RAM. Decompress at execution
2)scale draw distances
3)limit frames per second to 28fps(use of motion blur,repetition etc to smooth over the animation)
4)Take out everything Wii isn't capable of or does poorly
5)Optimise for use of the TEV unit.
6)Optimize the renderer for low polygon models(techniques useful for low polygon models like displacement mapping,normal maps,bump maps,hypernurbs,powernumbs,parallax maps,noise generators etc)
7)Use of memory cards or USB flash cards(512MB,1GB ect) as extra memory(yes I know the latency issue hopefully as 28fps the latency will be less)
8)scale texture color presicion at farther distances

I know there will be latency issues but the end result will be better. :devilish:
 
Displacement, normal, and parallax mapping isn't going to help you on Wii. You also aren't going to see a USB-based RAM expansion. It's just too slow for something like that.
 
1)Compress(into a .dol or a .lzo file) the contents of the RAM. Decompress at execution
You can't effectively run the entire contents of the game from compressed data files. You an use compressed textures and audio formats, but on the whole the data needs to be in RAM in a ready-to-use form, unless you have lots of CPU cycles to spare.
4)Take out everything Wii isn't capable of or does poorly
By this point, there's the question of why use UE3 then!
5)Optimise for use of the TEV unit.
One would hope that a UE3 engine for Wii would be optimized in every way. The question is a matter of when optimized, is there enough horsepower to drive the UE3 engine effectively.
7)Use of memory cards or USB flash cards(512MB,1GB ect) as extra memory(yes I know the latency issue hopefully as 28fps the latency will be less)
These have the same sorts of latencies as an HDD, and can only be considered as a streaming media format. It's not useable as working memory.
8)scale texture color presicion at farther distances
:???: What do you mean by that? Have 24 bit textures for up front, and 12 bit textures for small mipmaps? That'd only save you a bit of bandwidth. The processing is still going to be in the frame-buffer precision or whatever the hardware can cope with.

I know there will be latency issues but the end result will be better. :devilish:
Trading delays on input latency for eye-candy will make for a pretty rough game. Also 28 fps will either have jitters on 30 fps displays, or you have unsync'd vertical tearing. Neither of which is ideal. Though more and more common in console games I think.
 
How about instead of porting one of the Unreal Engines to Wii they just build an engine entirely around the system. Unreal Engine Wii if you will :)
 
If someone can port Far Cry to Wii...:LOL:

In fact, Unreal Engine is not all about the graphics, which is only a small part comparing to the entire engine. The script system has a lot of advantages (like latency function supporting); the data are self explained via metadata; the framework of IO/network subsystem is very well designed; and most impotant, the Unreal Editor is an excellent integrated working environment for artists and level designers. So, stripping out some complex shader system, it's a good software platform for Wii titles.
 
If someone can port Far Cry to Wii...:LOL:

In fact, Unreal Engine is not all about the graphics, which is only a small part comparing to the entire engine. The script system has a lot of advantages (like latency function supporting); the data are self explained via metadata; the framework of IO/network subsystem is very well designed; and most impotant, the Unreal Editor is an excellent integrated working environment for artists and level designers. So, stripping out some complex shader system, it's a good software platform for Wii titles.

Seeing how the latency(going from what they guy above shared) would make it nearly unplayable. I think releasing a UE 2.0 modified for the TEV unit with limited draw distances at 32fps (below 28fps you get a video desyncing? Why is that?).
 
You'd have either jitters or tearing with 32fps just the same becuase that is the console has to output at 60hz. Why are you suggesting things like 28 and 32 fps anyway?
 
How about instead of porting one of the Unreal Engines to Wii they just build an engine entirely around the system. Unreal Engine Wii if you will :)

Ha I actually suggested that same idea in a thread on ign about a year ago. I said that if Wii was really successful that some company with make "unreal engine thWii".

:smile:
 
6)Optimize the renderer for low polygon models(techniques useful for low polygon models like displacement mapping,normal maps,bump maps,hypernurbs,powernumbs,parallax maps,noise generators etc)

Umm correct me if I'm wrong but won't hypernurbs and powernurbs require a geometry processor to do it at any reasonable speed? ( ala DX10 )
 
Status
Not open for further replies.
Back
Top