Could Dreamcast et al handle this/that game/effect? *DC tech retrospective *spawn

[Cloofoofoo]

What's the debate here? Yes, the technique was used again. No, it wasn't used the same way as ttt. Yes sc3 is overall a better looking game than ttt. I don't get where the debate is here

Not really a debate on anything. I was just saying to say it the fake material technique was not replicated isn't right. That's all . It isn't even that special , like the dojo floor thing , a slightly simpler version was used in lets say cosmic smash on the DC for the specular highlights on the walls and ceiling of the white rooms. It just seemed ridiculous to compare ttt to PBR when it's just a mix of good hand painted and clever use of env map-like techniques( mixed with lighting) .

 

[Sonicfan]

Not really a debate on anything. I was just saying to say it the fake material technique was not replicated isn't right. That's all . It isn't even that special , like the dojo floor thing , a slightly simpler version was used in lets say cosmic smash on the DC for the specular highlights on the walls and ceiling of the white rooms. It just seemed ridiculous to compare ttt to PBR when it's just a mix of good hand painted and clever use of env map-like techniques( mixed with lighting) .

Reminds me. Space channel 5 part 2 did some interesting things with materials. Even mods to doa2 show great floor fx

 

[Nesh]

Not really a debate on anything. I was just saying to say it the fake material technique was not replicated isn't right. That's all . It isn't even that special , like the dojo floor thing , a slightly simpler version was used in lets say cosmic smash on the DC for the specular highlights on the walls and ceiling of the white rooms. It just seemed ridiculous to compare ttt to PBR when it's just a mix of good hand painted and clever use of env map-like techniques( mixed with lighting) .

All material techniques were fake. But you still haven't provided a single good example that challenges the quality of TTT's presentation to which even Digital Foundry agrees has not been replicated by any other game on PS2.
Your argument is at best, a generic one making a point that specular effects existed in other games. Nobody argued whether specular effects existed or not in other games. So far you have been arguing for the sake of arguing.

 

[XboxKING]

I remember TTT having a silk-like dress on one of the characters; that was really impressive to me! The rocks and shadows were also extreme and for example the school level at golden hour looked like I had never seen in realtime graphics

 

[Shifty Geezer]

May I remind everyone this is supposed to be a technical discussion about the hardware potential. As I understand it the argument is TTT's application of varied specular maps was unrivalled? So the debate is 1) Could DC use the same technique and 2) could it apply it to the same degree?

If the argument is PS2 could do it much better than DC, the fact only TTT is cited as doing this and no other game comes close even on PS2 makes one wonder why and should point to a technical reason why DC couldn't do it? So rather than just saying 'nothing looks as good as TTT', reasons need be given why TTT is doing something other games isn't and why DC couldn't follow suit.

 

[Nesh]

I remember TTT having a silk-like dress on one of the characters; that was really impressive to me! The rocks and shadows were also extreme and for example the school level at golden hour looked like I had never seen in realtime graphics

Yes, a few of them had. Lei and Xiaoyu specifically had a much more convincing depiction of silk than Setsuka in Soul Calibur 3. It wasn't just the existence of speculars, but the values in conjunction with the lighting quality did wonders. I wish Namco tried to experiment more with the engine and aimed for those aesthetics in their future games.

 

[Cloofoofoo]

All material techniques were fake. But you still haven't provided a single good example that challenges the quality of TTT's presentation to which even Digital Foundry agrees has not been replicated by any other game on PS2.
Your argument is at best, a generic one making a point that specular effects existed in other games. Nobody argued whether specular effects existed or not in other games. So far you have been arguing for the sake of arguing.

All material technique werent fake, for example that screenshot of cosmic smash up there has the "fake" specular but also some models( the player character had real material such as actual specular calculation) thats a material. Same with some other games. In some games both are used simultaneously.

In the case of ttt they did it by choice. Even in sc2 they had characters using the exact same technique fully on the entire body , some used none or just on small details. Stage wise in sc2 only 1 stage does it to make glossy marble material. They purposely turned it down , reasons who knows but ID bet its artistic resasons whether they felt maybe it was over used in ttt? There nothing stopping them from doing it again, heck they still extensively used it after ttt.

Yes, a few of them had. Lei and Xiaoyu specifically had a much more convincing depiction of silk than Setsuka in Soul Calibur 3. It wasn't just the existence of speculars, but the values in conjunction with the lighting quality did wonders. I wish Namco tried to experiment more with the engine and aimed for those aesthetics in their future games.

Xianghua/ yoshimitsu(alot of others) pretty much did the same as those two for their silk costumes, the most interesting use was on emblems or patters on clothing. Heres a couple example from a model viewer, metal and sil material. Wont look exactly as in game but its just to illustrate.

Metal (base texture, Env map angle 1, env map angle 2)

Silk + emblem(base texture, env angle 1, env angle 2)

 

[Shifty Geezer]

In the case of ttt they did it by choice. Even in sc2 they had characters using the exact same technique fully on the entire body , some used none or just on small details. Stage wise in sc2 only 1 stage does it to make glossy marble material. They purposely turned it down , reasons who knows but ID bet its artistic resasons whether they felt maybe it was over used in ttt? There nothing stopping them from doing it again, heck they still extensively used it after ttt.

As I understand it, the issue is simple one of scale. TTT uses different specular maps a lot, whereas other games not as much. Is there a particular technical reason, hardware limit/feature, that gave TTT scope to use more specular maps than games on DC? And why didn't more fighting games on PS2 do the same as TTT?

 

[Cloofoofoo]

I remember TTT having a silk-like dress on one of the characters; that was really impressive to me! The rocks and shadows were also extreme and for example the school level at golden hour looked like I had never seen in realtime graphics

Yeah that stuff was impressive in ttt,sc2, sc3. Its amazing what a simple 64x64/32x32 env combined with a couple textures with an alpha channel can do. Look at ivys costume.

env map off, just diffuse texture:

Env map on , angle 1:

Env map on angle 2:

 

[Nesh]

All material technique werent fake, for example that screenshot of cosmic smash up there has the "fake" specular but also some models( the player character had real material such as actual specular calculation) thats a material. Same with some other games. In some games both are used simultaneously.

In the case of ttt they did it by choice. Even in sc2 they had characters using the exact same technique fully on the entire body , some used none or just on small details. Stage wise in sc2 only 1 stage does it to make glossy marble material. They purposely turned it down , reasons who knows but ID bet its artistic resasons whether they felt maybe it was over used in ttt? There nothing stopping them from doing it again, heck they still extensively used it after ttt.


Xianghua/ yoshimitsu(alot of others) pretty much did the same as those two for their silk costumes, the most interesting use was on emblems or patters on clothing. Heres a couple example from a model viewer, metal and sil material. Wont look exactly as in game but its just to illustrate.

Metal (base texture, Env map angle 1, env map angle 2)

Silk + emblem(base texture, env angle 1, env angle 2)

I don't think we are discussing the same thing. At this point you are playing with semantics or discussing the generic use of materials and use of speculars. I still dont see in any if your examples anything that approaches the quality of PBR materials that TTT manage to approach. You are only showing the presence or effort of multiple materials in other games. Again the discussion was not whether different materials existed or not in other games. It is regarding the quality and how closely they approach a CGI'Ish/BPR presentation.
While you are posting a whole essay to make a good case of materials existing in SC3, and trying to argue with me, you need to show actual comparisons that show a more realistic implementation in SC3 and also argue with Digital Foundry's same conclusion. The images you post constantly prove Digital Foundry that such presentation was never recreated by Namco, not your case. I don't see it either in any of your images.

 

[Cloofoofoo]

As I understand it, the issue is simple one of scale. TTT uses different specular maps a lot, whereas other games not as much. Is there a particular technical reason, hardware limit/feature, that gave TTT scope to use more specular maps than games on DC? And why didn't more fighting games on PS2 do the same as TTT?

Well I talked to the guy who did the soul calibur 2 model little scene demo on dc( you remember that on the topic?).He did specific test regarding that in ninja2 ( pretty much same effect and screens up there, env map+ alpha maps for selective shine). He said the biggest draw back ( he felt) wasnt even the transparency performance but the fact the model has to be drawn twice and since dc already somewhat limited in that regard depending how dense of a model it can be a serious hit. He did a teapot with a base texture + alpha and on top an env map. for example is the teapot is 12k vertices and after culling 7.5k vertices are drawn now with env is x2 . 15k vertices are now drawn!. And this was on ninja2 which he mentioned was tuned for better performance on geometry(much more than ninja1) and transparency( seems to mention it uses the secondary accumulation buffer and exposes it but ill mention that in a minute.) Alas no game uses it on dc since i guess its sdk cameout too late to matter.

Sega did release a demo very early on doing the same fake material effect that ttt does way before that released. Its the vase demo and it leverages dreamcast specific hardware meant to reduce fillrate hit for blending. It leverages the secondary accumulation buffer. Almost no games used it but could have been used to reduce fillrate cost of bump maps and other multi texture effect. But you still have the cost it requires multiple passes to use to increases geometry cost of the scene.Developers probably didnt deem it important enough to make use of it or early sdks may have not expose it to them.

Quote from redream author who implemented it on his emulator:

Secondary Accumulation Buffer Support​

Support has been added for a rarely used PowerVR feature - the secondary accumulation buffer. The secondary accumulation buffer was a secondary render target that could be drawn to, and then sampled from, during the same frame. This could be used to provide multi-texturing, bump mapping and trilinear filtering but in practice it was only used by a handful of retail games (Evil Dead and a few 2K sports games).
NHL 2K2 used the secondary accumulation buffer to composite the ice texture with a decal texture and then blended that back into the main scene:

Sdk(machine translation):

This is used when compositing polygons with the secondary accumulation buffer. If you want to properly draw trilinear filters or bump maps with alpha blending, you must use the secondary accumulation buffer. However, in that case, it takes a total of three passes (the load of three semi-transparent polygons), so the rendering load increases.

Segas demo on env maps using alpha using the secondary accumulation buffer:

 

[Shifty Geezer]

discussion was not whether different materials existed or not in other games. It is regarding the quality and how closely they approach a CGI'Ish/BPR presentation.

Actually it's about whether other games could match it, not whether they did. Other games on PS2 could have matched TTT's rendering because they were running on the same hardware. The fact they didn't isn't a hardware limit but a software/art one.

So we need to determine how TTT achieved its look and whether games on other hardware could accomplish the same. Cloofoofoo's argument is the same techniques were used and so different art could have gotten closer on DC's hardware to TTT's look.

 

[Cloofoofoo]

Actually it's about whether other games could match it, not whether they did. Other games on PS2 could have matched TTT's rendering because they were running on the same hardware. The fact they didn't isn't a hardware limit but a software/art one.

So we need to determine how TTT achieved its look and whether games on other hardware could accomplish the same. Cloofoofoo's argument is the same techniques were used and so different art could have gotten closer on DC's hardware to TTT's look.

But I did tell you how its done, exactly the same way Sc 2 and 3 does it. So i went back into the ps2 emulator to poke around the textures, the models get highlighted green when texture is chosen then i dump it. Heres lee skirt section of costume, base texture ( most are either 128x128 or 256x256) along with its 32x32 environment map( squiggly lines). Now lets find jins pant leg with the fire, heres its base texture... and its 32x32 env map texture ( spherical map). The stages which i didnt post essentially do it the same way. Theres nothing especial about the method, their artistry is really good is all and how things are lit but honest gamecube( who basically does this as an afterthought) and xbox(doesnt even have to do this sincei t can do actual specular maps and normal maps) would eat this for breakfast and not break a sweat. And As i mentioned before its been done on the dc as well.

The only thing I noticed about ttt was that its everywhere which is unusual even for ps2. Maybe what makes it look different isnt this but perhaps how its lit? maybe its not standard lighting but spherical harmonics or something for a softer look ? And the other namco games dont do it? but this is purely speculation. Personally i think its traditional lighting and ppl hold this game to unrealistic pedestal.

Lee specular selected(thats why his skirt is green:

lee base skirt texture:

lees env map texture:

jins fire pant leg selected:

jins pant leg base texture:

jins pantleg env map:

 

[Shifty Geezer]

But I did tell you how its done,

I know. I'm trying to get Nesh to engage in the technical discussion by talknig about techniques and what the hardware requirements are and what the limitations are. With the claim DC couldn't manage the same, why? (That question to Nesh as he's the one saying it can't)

The Snowblind Engine shows some pretty impressive material effects. Skin is rendered very differently to fabric and stone and metal in Champions of Norrath.

 

[Nesh]

I know. I'm trying to get Nesh to engage in the technical discussion by talknig about techniques and what the hardware requirements are and what the limitations are. With the claim DC couldn't manage the same, why? (That question to Nesh as he's the one saying it can't)

The Snowblind Engine shows some pretty impressive material effects. Skin is rendered very differently to fabric and stone and metal in Champions of Norrath.

In the quote that he debated I pointed that TTT was the only game that approached the closest the quality of what you would expect from PBR pipeline than any other game including PS2. His argument was that other games reached it or surpassed it. It wasnt about why DC could or could not reach that quality.

WIth available evidence neither me or him have any technical know hows on the hardware functions on why it could or it couldn't do it. I have know how on texture maps and BPR materials due to my work. But that's a software than a hardware thing. What info is available right now is that no game has surpassed it even on PS2 and that we have no evidence that can show as whether the DC could reach it or not. The evidence though do exhibit that the PS2 was handling better more effects and materials but then there is always the debate if whether the DC could achieve those effects if it had that support and didn't die soon and that we will never be able to find out

 

[Shifty Geezer]

In the quote that he debated I pointed that TTT was the only game that approached the closest the quality of what you would expect from PBR pipeline than any other game including PS2. His argument was that other games reached it or surpassed it.

I'm seeing an argument the technique TTT used was used on DC and as prolifically. The only visual discrepancy I'm seeing is in the artistic application of the the effect - the SC3 shots show similar material application.

WIth available evidence neither me or him have any technical know hows on the hardware functions on why it could or it couldn't do it. I have know how on texture maps and BPR materials due to my work. But that's a software than a hardware thing. What info is available right now is that no game has surpassed it even on PS2 and that we have no evidence that can show as whether the DC could reach it or not. The evidence though do exhibit that the PS2 was handling better more effects and materials but then there is always the debate if whether the DC could achieve those effects if it had that support and didn't die soon and that we will never be able to find out

That's kind of it. We can't really say how DC would have fared as we don't know the cause of the differences. DC using per-material maps does indicate the general premise. However, TTT could have been using more passes with more 'light sources' or various blends. Someone would need to dig out a Namco interview!

 

[Nesh]

I'm seeing an argument the technique TTT used was used on DC and as prolifically. The only visual discrepancy I'm seeing is in the artistic application of the the effect - the SC3 shots show similar material application.

That's kind of it. We can't really say how DC would have fared as we don't know the cause of the differences. DC using per-material maps does indicate the general premise. However, TTT could have been using more passes with more 'light sources' or various blends. Someone would need to dig out a Namco interview!

Soul Calibur 3 didn't exist on the DC and it is still a pointless argument since the argument wasn't about the technique (as the simple application of maps) but about the quality of materials which is not just the application of maps (technique) but also the quality and handling of lighting by the engine which interacts with the maps. You can apply the same maps on many engines. But the quality won't be necessarily of the same quality.

The mere existence of maps doesn't tell the full story.

Edit: I m not at home so I can't post images, but when I get back I can show some examples of areas that make TTT look more convincing compared to it's contemporaries. None of the techniques are unique to TTT, but it's how it handles them and applies them differently in every surface.

 

[Shifty Geezer]

Edit: I m not at home so I can't post images, but when I get back I can show some examples of areas that make TTT look more convincing compared to it's contemporaries. None of the techniques are unique to TTT, but it's how it handles them and applies them differently in every surface.

But does TTT handle it differently because of a hardware difference, or just artistic application? The fact no other PS2 games got the same results suggests the latter, with Namco just being ahead on the superior execution of conventional techniques.

Edit: And I don't understand Cloofoofoo's point to SC3. I thought that was a DC game but it wasn't. What are the DC games that used the same techniques?

 

[Cloofoofoo]

But does TTT handle it differently because of a hardware difference, or just artistic application? The fact no other PS2 games got the same results suggests the latter, with Namco just being ahead on the superior execution of conventional techniques.

Edit: And I don't understand Cloofoofoo's point to SC3. I thought that was a DC game but it wasn't. What are the DC games that used the same techniques?

Sc3 had nothing to do with dc. It's the point same company same technique , used probably as heavily if not more than ttt. This was just an answer to ttt being unmatched even by later games claim he made. Point was it's doing the same thing. Namco didn't really ever change what they were doing. The only real difference between the two they stopped applying to skin. That's definitely an artistic choice though

 

[Shifty Geezer]

I see. Would be nice to bring this back onto the topic of other platforms running the same games and techniques.