Can Wii achieve the same level of Xbox's Doom3?

[Richard]

Btw guys you might be interested in this for context:

GP: On the PC, Doom 3 is going to be something of a resource monster. When it was first proposed as an Xbox game was there any concern about getting it to run on a console?

KB: John Carmack made some really good decisions in architecting the engine for scalability. We knew that PS2 and Gamecube were out of the question due to particular graphics requirements (global illumination, normal mapping, shadow volumes etc.), but we knew Xbox had a shot. The key concerns were memory and performance. With only 64MB RAM on the Xbox and a 733Mhz processor, we knew it would be quite challenging ?- but not impossible. The team here was really excited to be working on this project -? a chance to work on Doom! -? so we were definitely going to take it on and prove that a great Xbox version could be made. Now, after many months of development, I think it?s safe to say that players will be stunned when they see the type of graphics fidelity that Doom 3 is able to present on the Xbox.

VV ported D3 to the xbox, JK2 to the gamecube, etc.

 

[Fafalada]

As I said before, it might go beyond simply specs in this situation.

Like blast processing?
Are there any known examples of this? All I'm aware of are relative comparisons between XBox and PS2 where the two do exactly as you'd expect from respective spec (latter beating the former easily).

The Gamecube, and so far the Wii, has had zero games in total.

As I've noted, to my knowledge at least Spiderman2 did this (and that's outdoors, which is pretty much worst case scenario for volume usage). But I know little about rest of GC library - are you actually basing this off anything other then PR releases touting high-profile titles?

 

[Statix]

Like blast processing?

Like the "back door hacks" and "hidden instructions" and "custom shaders" unique to the Xbox's GPU that Vicarious Visions was referring to. I don't know much about the Xbox's hardware, or the GC. But I do know that there are some things unique about the videocard used in the Xbox, such as extra vertex units, and the aforementioned hidden instructions/optimizations, that improved performance majorly, especially in later games in the Xbox's lifespan. It also has programmable shaders that (I think) ultimately has some bearing on the system's ability to render dynamic light and shadows as compared to the Gamecube.

As I've noted, to my knowledge at least Spiderman2 did this (and that's outdoors, which is pretty much worst case scenario for volume usage).

But even in that game, notice in the screens, that only Spider-Man casts such a shadow; not the environment, the buildings, or even other character models or objects. They're probably not even using shadow volumes at all, but simply some extra gray-colored polygons mirroring Spidey's model in order to "fake" a shadow being cast. Even Quake 2 and some N64 games used this method for character shadows.

 

[Fafalada]

unique to the Xbox's GPU that Vicarious Visions was referring to.

They were comparing direct access to XBox GPU to being handcuffed by API restrictions on PC. All consoles that allow metal-access benefit from this to various extents.
And I totally agree about programmable-shader advantages, but this affects the other half you mentioned. Volumes just need the fillrate and something more flexible then shaders(usually the CPU) to run the geometry.

But even in that game, notice in the screens, that only Spider-Man casts such a shadow; not the environment, the buildings, or even other character models or objects.

The buildings were volume shadowed, with a moving light source (day-time changes dynamically). You're right about non-player characters, but this was a free-roaming game with a very large city to fly around, some concessions had to be made.

 

[sfried]

I remember John Carmack once saying Doom 3 would've been possible on GC if only it had more memory. Knowing fully well the Wii has more RAM than the GC, all that's needed is for someone at id to recode the engine to utilize TEV architecture (as opposed to shaders).

Stenciled shadows? Didn't parts of Metal Gear Solid: The Twin Snakes utilize that? (I believe the basement area did.) And din't Dewy's Adventure have normal mapping on some enemies?

So in summary, most likely yes. We can't can't just go about clamoring how Wii can't do it "simply because I haven't seen anything it on the console".

 

[swaaye]

On the other hand, wouldn't it be rather laughable if the Wii can't manage what the 2001-built NV2A could? Uhg.

I'm interested in seeing what The Conduit pulls off (and how fast it pulls it off) when it's released.

 

[Mobius1aic]

On the other hand, wouldn't it be rather laughable if the Wii can't manage what the 2001-built NV2A could? Uhg.

I'm interested in seeing what The Conduit pulls off (and how fast it pulls it off) when it's released.

We can list off one thing: any resolution above 480p.

 

[Dargakis]

so the conclusion is that the wii can never achieve this level of detail?

 

[Nano]

so the conclusion is that the wii can never achieve this level of detail?

Most aspects are possible, its just some people quite frankly can't answer.

I think the thing Wii would struggle with is normal mapping on everything due to the differences between the TEV and NV2A's shader architecture.

Either way, what's to say Wii can't make a game that looks a lot better than Doom 3 did back on Xbox? Its just one title with its own way of rendering things. I think the platform could potentially do better.

 

[wsippel]

We can list off one thing: any resolution above 480p.

Some developer mentioned a while ago that it would be theoretically possible to do 720p, but it would be somewhat hacky. I guess it might work by splitting the viewport (rotate the camera by 90°, render each frame two times with a slight offset, combine the two resulting framebuffers to a single 640 x 960 image, rotate - 90°, scale to 1280 x 720). Most likely not worth it, though.

 

[Kiloran]

I remember John Carmack once saying Doom 3 would've been possible on GC if only it had more memory. Knowing fully well the Wii has more RAM than the GC, all that's needed is for someone at id to recode the engine to utilize TEV architecture (as opposed to shaders).

Stenciled shadows? Didn't parts of Metal Gear Solid: The Twin Snakes utilize that? (I believe the basement area did.) And din't Dewy's Adventure have normal mapping on some enemies?

So in summary, most likely yes. We can't can't just go about clamoring how Wii can't do it "simply because I haven't seen anything it on the console".

Both Id and Vicarious Visions said the port wouldn't be possible, for all the reasons listed above (global illumination, normal mapping, shadow volumes) I think it's pretty safe to say that achieving the same level of realism on the Wii (as the Xbox) is impossible. I'm sure that any one of those effects would be possible to mimic on the Wii, but rendering generous amounts of all three together just isn't possible on that architecture.

 

[pc999]

Both Id and Vicarious Visions said the port wouldn't be possible

Just to note that those statements are after the D3 port are annunced, they had to be sure to promote their game. Anyway it would probably strugle with it anyway, there are much better ways to use Wii HW, D3 like SW isnt what they dev GC/Wii for, that are the reasons why things like EMBM are so easy on GC/Wii.

 

[vanquish]

We can list off one thing: any resolution above 480p.

A question; the Wii has a set framebuffer right? so it can't do 480p with AA?

What about the 360, did any notable titles have anti aliasing?

 

[Dargakis]

so in the end, my conclusion is"it would look like wii farcy compared to xbox farcry.
although the game is more fun because it's on the wii, the graphics would lack certain effects, which are crusial to farcy it's look, the same as why doom3 would look like crap also.

 

[pc999]

so in the end, my conclusion is"it would look like wii farcy compared to xbox farcry.
although the game is more fun because it's on the wii, the graphics would lack certain effects, which are crusial to farcy it's look, the same as why doom3 would look like crap also.

Far cry Wii is just a rushed port in no way indicative of Wii potential, just think in how many games in GC have much better water fx, (even than any XB game?) textures and even jungle scenes (SW:RS).

XB and Wii are really diferent pieces of HW, if you just port over (no matter which way) you will end up losing a lot of qualitity. If they did a remake for Wii the end result would be much better (and costly, both in money and time). Something that is much close to what we saw on GC/Wii is probably UC2 or others UE2X games.

True Wii HW is probably weak for somethings that are on D3, that is why it is so strong in things EMBM, TnL and such, in the end games are supossed to look diferent. If better or not is a question of taste.

If you prefer Wii (even GC) does have somethings that we never saw a XB game pulling (physics, particle fxs, and some high poligons scenes with special fxs and/or EMBM), while XB had made somethings that we never saw on the Wii like D3, or a 32bits frame buffer).

 

[Fafalada]

(global illumination, normal mapping, shadow volumes)

D3 has no form of GI (dynamic anyway).
In regards to normal maps, Wii has considerably more usable memory, so one would expect texture quality to be higher as well. What can actually be done with them in lighting model is another question.

 

[pc999]

D3 has no form of GI (dynamic anyway).
In regards to normal maps, Wii has considerably more usable memory, so one would expect texture quality to be higher as well. What can actually be done with them in lighting model is another question.

Just one question, considering that the normal mapping in D3 seems so poor anyway, couldnt they just use plain bump mapping, in most cases seems as good, yet it saves a lot of performance. Or would they lots more than one thinks in terms of lightning?

 

[Squeak]

Just one question, considering that the normal mapping in D3 seems so poor anyway, couldnt they just use plain bump mapping, in most cases seems as good, yet it saves a lot of performance. Or would they lots more than one thinks in terms of lightning?

Normal mapping is bump mapping. I don't know what you mean with "plain bumpmapping".
If you mean emboss mapping, then it would have equally high cost to normal mapping (two passes and lighting calculations) and be less flexible.

 

[Nano]

Normal mapping is bump mapping. I don't know what you mean with "plain bumpmapping".
If you mean emboss mapping, then it would have equally high cost to normal mapping (two passes and lighting calculations) and be less flexible.

Why would emboss bump mapping be as costly as normal mapping?

I couldn't imagine Rogue Squadron III with the same performance had its embossing been replaced by normal maps. In some places it covers entire environments in the stuff (Hoth, Endor).

 

[pc999]

Normal mapping is bump mapping. I don't know what you mean with "plain bumpmapping".
If you mean emboss mapping, then it would have equally high cost to normal mapping (two passes and lighting calculations) and be less flexible.

Doesnt it need a few more CPU cicles and memory? Plus there is some hardwired HW for emboss mapping in GC right?.

Found it

Method 4: Emboss mapping

Are more subtle but nevertheless important method is bump mapping where a height field mapped onto a surface describes its elevation per pixel without adding geometric data. On the Nintendo Gamecube two different methods are straightforward to implement. Emboss mapping computes light values per pixel. It is not possible to compute “bumped” specular highlights and reflection with this method. “Real” per pixel bump mapping using the indirect texture unit is capable of doing so (c.f. method 5).

The hardware has direct support for emboss mapping.

Method 5: Bump Mapping

Visually better results can be achieved using “real” bump mapping as supported with the indirect texture unit. Using this method the hardware computes a normal per pixel and uses that to lookup different textures including a diffuse light map (containing all directional and ambient lights), an environment map (as described in method 3) and even a specular map. Thereby all those shading effects are computed correctly in a bumped way. However, since the global lights are now fetched from a texture instead of being computed by the lighting hardware, the texture needs to be generated dynamically as soon as the camera orientation and/or the lights change (again, one can find an example on how this is done in the demo section of the Nintendo Gamecube SDK).

In addition, the height field needs to be pre-processed into a “delta U/delta V texture” (which is an intensity/alpha texture with four bit per component) and therefore needs (without further measures) twice as much memory for texture storage than the emboss mapping method described in method 4.

The delta-texture is fed into the indirect unit where it is combined with the surface normals, describing the orientation of the bump map. In the last stage of this three-cycle setup, the diffuse light map is looked up and the result is the bumped light color for the global lights. Note that the local lights are still computed per vertex (because they have a location and the normal used as input data does not give this information) and are added later in the texture environment.

But I didnt remember this

Emboss mapping does not support the computation of specular highlights. However, one can just ignore the emboss map and add non-bumpy specular highlights. Nevertheless, by doing so, the dark edges of the bumpy surface will be removed (due to the adding) and the effect falls apart to some extent (not to mention that the specular highlights by itself ignore the height field completely).

That means that in games like SM:G or The Conduit, here we see massive amounts of bump+specular (and in the later additional fx) they are all done via indirect texturing, nice, unless there is some extra HW just for it. Anyway it answers my question.