Comparison of 6th generation game console hardware

[Barone]

Ratchet & Clank (2002) used renders from Jak and Daxter: The Precursor Legacy (2001) and the Kinetica (2001) engine. So it doesn't seem like ultra-late NASA-only technology to me, but maybe in some alternate reality 2001 or 2002 releases are already considered late gen titles for a 2000 console, IDK...
Mental note: Must ignore the existence of companies which succeeded under Sony's umbrella. Must ignore how many companies went under due to Sega dropping support of its consoles and add-ons abruptly.

Haven: Call of the King (2002) is a third party game by Traveller's Tales which claimed: "An engine capable of 13 million polygons per second with one-pass fogging and point-lighting gives designers limitless control to build a completely immersive and interactive world.". But it doesn't count for the sake of not-counting.
Mental note: I shall request primary sources to invalidate the numbers which disfavor my narrative.

Somehow the bad programmers at Bizarre Creations failed to unleash the Dreamcast's Untapped Power from Hell^® for Metropolis Street Racer despite not using the Windows CE Toolkit and spending more than 30 months in development with full technical support from Sega.
Mental note: Must hide the existence of DC games which weren't rushed or which had big development teams.

In some alternate reality, around DC's death Sega would be developing a low-level profiler similar to the Performance Analyzer to be released around 2003 which would finally unleash the Dreamcast's Untapped Power from Hell^®, despite the Sony's PA system being known since 1997/1998 and the Dreamcast being a platform targeted at high-level programming.
Mental note: Must play with conjectures to even things. At the end of the day everything is shades of gray and I win.

Somehow Melbourne House managed to get Grand Prix Challenge (2003) running at 60 fps on the PS2 using its racing game engine which ran at 30 fps with Le Mans 24 Hours on the Dreamcast. Anti-competitive Sony must have payed Melbourne House to downplay the DC's hardware; that's for sure.
Mental note: CANNOT remember people that Melbourne House's engine was ported to different platforms.

Need for Speed Underground 2 (2004) was considered graphically-heavy, had simultaneous releases for PS2/GC/Xbox/PC and performed better on the PS2 than on the other consoles. But that must be ignored in order to preserve my narrative which is solely based on historical facts while disconsiders the development of middleware tools for game development over the years - evolving tools and techniques are considered when they benefit my narrative though.
Mental note: Must create, promote and repeat until become historical facts the disadvantages of the systems I dislike.

Thanks to the Sony's Hype Machine^® the "some games" of the PS2 which matched the Dreamcast's Untapped Power from Hell^® were also good selling titles and they sold much more than "most of" Dreamcast games. But those "some games" aren't even as relevant as the DC ones 'cause people playing them weren't gamers but just mainstreamers.
Mental note: Percentage or ratio of games is a good way to relativize the importance of PS2 games' technical achievements.

In some alternate reality, the worst selling and way-quicker-to-die console is the one that set the standards for the entire gen; so multipass effects are crap and they just don't count. And, thus, tile-based rendering is the king despite being the most limited architecture in terms of numbers of polygons/VRAM.
Mental note: Must use the multipass argument to invalidate the higher polygon counts of the PS2 games.
Mental note: Avoid direct comparisons at all costs. Keep calm and go around in circles.


Now, outside the infinite loop of revisionism and relativization, I've found these links:
- Briefly explains how some of the Shadow of Colossus graphical effects were implemented:
http://selmiak.bplaced.net/games/ps2...&page=makingof

- Factor 5's "we estimate most scenes at 12-15 million polygons per second" prior to Rogue Leader's release for the GC:
http://www.ign.com/articles/2001/08/...hat-transcript

[stu]

The Ratchet game above is showing 6 million polygons per second and is considered technically advanced for the PS2. Melbourne House's interview about Grand Prix Challenge says that "most" other games were doing 6 million polygons per second max compared to their 12-13M, or 18M with "multipass effects". But yes, I think most Dreamcast games get trounced technically by some PS2 games, especially after the performance analyzer was released to third parties in 2003.

The Dreamcast never would have had such a tool you see, it isn't possible to improve Dreamcast games in any way. It was completely maxed out by like two games.

Seriously though, I think both systems have a fairly steep list of caveats for maximum performance to be achieved. For the Dreamcast the developer had to focus enough on the platform to optimize for the tile based renderer and not use Windows CE and not port from the PS1 or N64. The PS2 needed the developer to be savvy to Assembler level parallelism and the game could_not consider multiplatform needs. From what I can tell based on developer interviews both conditions are likely to produce more "bad ports" and poorly optimized games than not. And here we are.

Considering that there was a version of the PA for the PS1 back in 1997 I would say that Sega had ample opportunity to develop a DC based PA of their own had they felt that it would of made a difference. From what I have read in this thread and the other DC vs PS2 thread it seems to me that Sega chose to make the DC dev environment as transparent as possible (ie exposing all of the features of the DC to the developer thus allowing them to exploit the DC's capabilities from the outset) I also seem to recall that the 5 million polys/sec is about the max the DC can draw on the screen before saturating the memory bandwidth of the system, it seems to me that these factors made a performance analyser for the DC redundant.

I'm not really seeing where a performance analyser would boost the DC's performance above what games like Le Mans already achieved.

[sheath]

We just saw that most scenes in Test Drive Le Mans can be targeted below 300k per second correct? Le Mans is supposed to be the only Dreamcast game to even approach 5 million polygons per second peak. This peak is only achieved with full weather effects, trails and all cars on screen. Meanwhile on the PS2 we have in game samples of over 6 million polygons per second and this is comparable to the 300k polygon per second "average" of Test Drive Le Mans. Could it, maybe possibly occur to a developer that a level of detail engine might possibly push more than sub 500k polys per second in a Dreamcast game in the average scene? Should games have been made for the Dreamcast for a couple more years would more games have targeted 5 million polygons per second peak while averaging below 1/10th that figure? I see plenty of room for improvement in Test Drive Le Mans alone.

That Le Mans and even games like Dead or Alive 2 can easily be found to be running below 1 million polygons per second actually plays directly into my earlier assertion. The entire generation was "floating around" the 1 million mark. Either that or five million or fifteen million polygons really aren't the super huge deal Sony and MS would have made it out to be. Remember earlier that the context to my comments was Resident Evil 4 for PS2, which was reputed to run at 900k per second peak on the PS2 version and still "looks fine" to Playstation 2 gamers.

I see a lot of problems implicated by these facts. Really though, I just wanted all of the documented facts in one thread so the truth could be known. If the PS2 really is every bit the "polygon monster" fans claim it is I would have thought I would have been inundated with facts that support that claim by now. This has not been the case, and not for lack of trying on my part. I also assumed that it would be apparent in the average game, especially against the Dreamcast but also against the Xbox and Gamecube. I am not preventing anybody from proving the PS2, on average, outperformed the Dreamcast's in game peak in most games. I am simply discussing what is on the table. Multipass effects add five million polygons per second without increasing model complexity? Fine. Palletized textured polygons at a lower screen and texture resolution are just the same? Awesome. More sparks are better than fewer sparks or other effects? Terrific. It doesn't matter to me as long as the claims are consistent across the board and have actual facts to support them.

[stu]

We just saw that most scenes in Test Drive Le Mans can be targeted below 300k per second correct? Le Mans is supposed to be the only Dreamcast game to even approach 5 million polygons per second peak. This peak is only achieved with full weather effects, trails and all cars on screen. Meanwhile on the PS2 we have in game samples of over 6 million polygons per second and this is comparable to the 300k polygon per second "average" of Test Drive Le Mans. Could it, maybe possibly occur to a developer that a level of detail engine might possibly push more than sub 500k polys per second in a Dreamcast game in the average scene? Should games have been made for the Dreamcast for a couple more years would more games have targeted 5 million polygons per second peak while averaging below 1/10th that figure? I see plenty of room for improvement in Test Drive Le Mans alone.

That Le Mans and even games like Dead or Alive 2 can easily be found to be running below 1 million polygons per second actually plays directly into my earlier assertion. The entire generation was "floating around" the 1 million mark. Either that or five million or fifteen million polygons really aren't the super huge deal Sony and MS would have made it out to be. Remember earlier that the context to my comments was Resident Evil 4 for PS2, which was reputed to run at 900k per second peak on the PS2 version and still "looks fine" to Playstation 2 gamers.

I see a lot of problems implicated by these facts. Really though, I just wanted all of the documented facts in one thread so the truth could be known. If the PS2 really is every bit the "polygon monster" fans claim it is I would have thought I would have been inundated with facts that support that claim by now. This has not been the case, and not for lack of trying on my part. I also assumed that it would be apparent in the average game, especially against the Dreamcast but also against the Xbox and Gamecube. I am not preventing anybody from proving the PS2, on average, outperformed the Dreamcast's in game peak in most games. I am simply discussing what is on the table. Multipass effects add five million polygons per second without increasing model complexity? Fine. Palletized textured polygons at a lower screen and texture resolution are just the same? Awesome. More sparks are better than fewer sparks or other effects? Terrific. It doesn't matter to me as long as the claims are consistent across the board and have actual facts to support them.

All you can say is that certain games within that generation of consoles were "floating around" the 1 million mark, to claim that the entire generation was at that level is patently false and you have previously gone on record as agreeing that it was false.

Right here in fact -> "Uh, yes, that later quote is plainly incorrect. I'm not even sure why I would have said that. I know why I was stating so plainly about the entire generation floating around the 1 million mark, that was based on Resident Evil 4 PS2. I was also sick of the myth around the PS2's polygon prowess constantly cropping up in any 6th generation discussion without any evidence, such as the developer quotes that have now been brought to the table.

-edit-
I just remembered why I said the second statement above, it was before the VUs being designed to stream polygon data to the GS instead of polygon data needing to be stored in VRAM was brought up. If the GS's 4MB of VRAM was to include polygon data the Dreamcast PVR2-DC would have been more capable of surpassing 3 million polygons per second. Since the VU1 can stream more polygons than that by quite a bit per second this point is wrong. This is exactly the kind of info I have always hoped this thread would bring up."

It has been proven by a number of games through out that generation that the 1 million polys/sec benchmark was well and truly broken, the only way that one can make that claim is by "cherry picking" the low hanging fruit like the games very early in that generation, but you can't extrapolate that across the entire generation and seriously claim that it proves true all the way across the generation. You even quoted a developer as saying that "most games were running at around 6 million polys/sec", yet you now are going back to this totally untenable position of claiming that 1 million/sec is where most games of that gen were at.
If you go through the PA presentation that Crazyace linked to and look at Slide 39 it will show you that as far as Sony were concerned that any game running at less than 7.2 million primitives/sec on the PS2 was running at a less than optimal level with 300k primitives/sec being sub-optimal (slide 34) and that is just on the PS2, what about games on Xbox and Gamecube? Are you claiming that these systems never got above that 1 million polys/sec level you have set?


Edit:- On Slide 44 of that presentation it shows that Sony is indicating that over 10M polys/sec should be obtainable - In Game on the PS2. That is presumably with all effects applied as well, which tallies roughly with what Melbourne House managed with the Formula One game for the PS2, among others.

[zyrobs]

Well, with a BIOS update in the GD-ROM the Dreamcast was in the same generation.

What is this supposed to mean?

[Barone]

"You must spread some Reputation around before giving it to stu again."

sheath has managed to drive around in circles for so long that now he's contradicting himself every time he posts.

sheath dismissed what Crazyace just said ("As the picture for the Ratchet capture looks like a quiet area ( no enemies close to the main character ) you cant assume that it represents the highest poly count - it may be more comparable to the 300k figure.") and the Factor 5's quote ("we estimate most scenes at 12-15 million polygons per second") in order to repeat his old stupid prediction ("The entire generation was "floating around" the 1 million mark") - which not only wasn't supported by any sources but also has been disproved a couple times already.
We shouldn't be surprised though...

sheath keeps dodging his own claims and ignoring facts while reiterating his absurd statements and trying to put the burden of proof on the others. That's a pretty bad attitude IMO.
It doesn't help that he never gets down off his high horse and keeps talking like if he was some sort of well-reputed historian ("It doesn't matter to me as long as the claims are consistent across the board and have actual facts to support them.") right after dropping another load of unsourced "facts" and esoteric assumptions.

Also, it sounds like sheath thinks Le Mans 24 Hours on the Dreamcast doesn't make use of a LOD algorithm; hehe.

[sheath]

Stu, it was a joke toward evilevox about an earlier thread he started regarding the DUO. Barone, you need to have your eyes checked or find a new hobby besides trolling me.

[TrekkiesUnite118]

Barone, you need to have your eyes checked or find a new hobby besides trolling me.

Disagreeing with you and proving you wrong != trolling.

[Barone]

Barone, you need to have your eyes checked

Your point?

or find a new hobby besides trolling me.

Self-pity much?

[sheath]

Disagreeing with you and proving you wrong != trolling.

Telling everybody that I will fabricate evidence, delete facts, and otherwise twist everything to some end I don't even have is definitely trolling. You should know, you love doing that kind of crap.

[TrekkiesUnite118]

Telling everybody that I will fabricate evidence, delete facts, and otherwise twist everything to some end I don't even have is definitely trolling.

You do twist the facts and ignore cold hard evidence for the sake of furthering your own agenda. Which is to put Sega and their history in the most positive light known to man. Sorry but this is a case of you being called out for shit you actually pull. That's not trolling. That's debating with you. If you don't like having your arguments broken down and disproved then I suggest you try making stronger arguments.

[KnightWarrior]

How many polygons can the Xbox and Gamecube do?

[TrekkiesUnite118]

How many polygons can the Xbox and Gamecube do?

A lot.

[sheath]

You do twist the facts and ignore cold hard evidence for the sake of furthering your own agenda. Which is to put Sega and their history in the most positive light known to man. Sorry but this is a case of you being called out for shit you actually pull. That's not trolling. That's debating with you. If you don't like having your arguments broken down and disproved then I suggest you try making stronger arguments.

Debating with me would go something like this, and I'll even use the caricature of me as an example:

Sheath: PS2 can't render more polygons than the Dreamcast because of the 4MB VRAM.

Respondant A: Actually, the PS2's VU1 streams polygon data to the GS with no need for VRAM.

Sheath: Well, that's news (roots around the internet) that nobody knows apparently. Good to know.

This has happened plenty of times, it is you Mr. Can't Be Wrong that constantly bends and twists everybody else's statements to continue the "argument". Classic skeptic nonsense. As for me, myself and I, I have already admitted to being wrong about at least one statement and am still working on the others. I guess some people just can't handle the process of actually looking hard at the facts and considering one's own background as bias.

[TrekkiesUnite118]

Debating with me would go something like this, and I'll even use the caricature of me as an example:

Sheath: PS2 can't render more polygons than the Dreamcast because of the 4MB VRAM.

Respondant A: Actually, the PS2's VU1 streams polygon data to the GS with no need for VRAM.

Sheath: Well, that's news (roots around the internet) that nobody knows apparently. Good to know.

This has happened plenty of times, it is you Mr. Can't Be Wrong that constantly bends and twists everybody else's statements to continue the "argument". Classic skeptic nonsense. As for me, myself and I, I have already admitted to being wrong about at least one statement and am still working on the others. I guess some people just can't handle the process of actually looking hard at the facts and considering one's own background as bias.

I have not twisted your own words. I've responded to them exactly as they were written. The same is true for Barone, stu, rusty, and every other "troll" in this thread. You claim to look at cold hard facts, yet you are the one completely ignoring them in this thread. The most recent example is on this very page. You are STILL clinging to the idea that most games of that generation hovered around 1 Million polygons per second. This has been disproved countless times and you still refuse to admit you're wrong on this one. And you have the balls to accuse me of always having to be right? Go take a look in the mirror for once.

When evidence is posted that proves you wrong, instead of going "ok, my bad" (something I actually have done a many times) you instead claim everyone's trolling you. Sorry but that's NOT trolling you. If I really wanted to troll you I'd post gifs and arguments that hit at you personally just to get a rise out of you. I'm not doing that. If Barone really wanted to troll you he could insert code into Game-Pilgrimage to redirect your IP to meatspin. That would be trolling you.

What's going on in this thread is completely tame. You are just having a very hard time accepting that your argument doesn't hold water.