TEXAN*s SEGA enthusiast thread - past and future hardware choices

Status
Not open for further replies.
@ function:

i will try to post my demo sometime this year as i do not have much time right now. In its current state it uses a single dynamic light and a prebaked ambient occlusion texture. It looks good but i want to add some extra layers of environment mapping, bump mapping and maybe cook some shadows as well.

80k verts @60 fps means i will have to dedicate another 1/60th of a second for game logic, so for such levels of geometry i think that 60 fps is not feasible. 30 fps are achievable though, and with the use of a good triangle stripping library like nvTriStrip in my exporter i can bring the geometry dataset down to half the size with no quality loss...so i would be pushing the equivalent of something like 150k verts @30 fps along with plenty of time of time for game logic to execute...

also someone said lemans pushes around 4.5MPolys/sec...not even close, just run nullDC and look at the number of vertices it sends to the gpu...
 
I ll be as brief as possible because I hate replying to chopped posts.

I checked almost all top games available on the DC a couple of months ago to remember what it was like. The performance improvement was less significant than that found on the PS2 withing the same period (2 years into their lifespans).

Visually Dark Cloud (dec 2000) > Phantasy Star (Nov 2000).

The PS2 didnt have games to compete with Shenmue because there wasnt any high profile RPG game developed for it..except one...the release of FF10 in July 2001 which showed tremendous difference in quality (thats just one tiny example). .

Talking about artistry Rez was just ok. Jet Set Radio looked outdated compared to 2000-2001 PS2 games even with my "retro perspective". Unlike examples like Zone of the Enders 2, GT and Rachet and Clank that were both technical and artistic showcases that could somewhat compete in some respects with GC and XBOX games of the same genre.

The most notable effect the PS2 could not produce compared to XBOX were the bumb maps and texture quality. Almost everything else was done or faked in one way or another.

Yeah even PS2 is getting releases of PS3 and 360 games but is judged as such. An old generation console that has passed its time. As for the Wii it primary attracts consumers for its unique experience. It is the main selling point. The DC was not PS1, nor Wii. The Wii can replicate physics adequately to maintain at least the gameplay elements of its ports. Which is where I saw the PS2 and XBOX providing a lot but DC wasnt. Even as such how many titles does Wii not get because of its performance? How many ports lack? A lot. But people wont care. They do care if their PS3 or 360 version though lack.

This arguement "if DC this and if DC that" doesnt say anything. Thats a huge conditional assumption. If DC was a super multimilion seller developers might have even ignored better technology altogether and made DC ports to everyone else that didnt take advantage their better technology. But since it wasnt this extreme case, it doesnt count for the DC. Other things mattered more than if things were different.

It just seems like you're arbitrarily drawing a line where performance is good enough for people to want to play the games, and you've put it in the relatively small performance gap between the PS2 and DC. If such a small gap could make such a big difference, the PS2 would have been annihilated by the Xbox, the Megadrive stomped by the SNES and the NES killed off by the Master System.

I don't see GT on PS2 competing with the phenomenal Rallisport Challenge 2 on Xbox any better than F355 competed with GT. But then again I'm not trying to draw a line of "can't support competitive gaming experiences" between the either the Xbox and PS2 or PS2 and DC.

It also feels like you're arbitrarily defining "dated" as "runs on the Dreamcast". There was nothing in 2000 that I can recall making JSR look "dated", but clearly there was a big gap in there somewhere for you. I doubt you would be willing to concede that Halo made everything on the PS2 look dated though.

On top of everything else, the Dreamcast also had a unique and compelling line-up of on-line games, which only began to show in the months before and after it was killed off. It came too late to help the platform make money. It's an entire area of gaming that seems to have no value, and be disregarded as an experience people might want in the rush to condemn a platform as dated for not having PS2 graphics.
 
@ function:

i will try to post my demo sometime this year as i do not have much time right now. In its current state it uses a single dynamic light and a prebaked ambient occlusion texture. It looks good but i want to add some extra layers of environment mapping, bump mapping and maybe cook some shadows as well.

80k verts @60 fps means i will have to dedicate another 1/60th of a second for game logic, so for such levels of geometry i think that 60 fps is not feasible. 30 fps are achievable though, and with the use of a good triangle stripping library like nvTriStrip in my exporter i can bring the geometry dataset down to half the size with no quality loss...so i would be pushing the equivalent of something like 150k verts @30 fps along with plenty of time of time for game logic to execute...

also someone said lemans pushes around 4.5MPolys/sec...not even close, just run nullDC and look at the number of vertices it sends to the gpu...

Thanks, and I look forward to seeing your demo. It sounds like you're getting pretty good performance relative to titles that were actually released for the machine.

It's interesting that you're going to be using a triangle stripping library to enhance performance - I wonder how well optimised mesh data would have been back in 1998 - 2000, when most of the DC games that were released were made.
 
It just seems like you're arbitrarily drawing a line where performance is good enough for people to want to play the games, and you've put it in the relatively small performance gap between the PS2 and DC. If such a small gap could make such a big difference, the PS2 would have been annihilated by the Xbox, the Megadrive stomped by the SNES and the NES killed off by the Master System.

I don't see GT on PS2 competing with the phenomenal Rallisport Challenge 2 on Xbox any better than F355 competed with GT. But then again I'm not trying to draw a line of "can't support competitive gaming experiences" between the either the Xbox and PS2 or PS2 and DC.

It also feels like you're arbitrarily defining "dated" as "runs on the Dreamcast". There was nothing in 2000 that I can recall making JSR look "dated", but clearly there was a big gap in there somewhere for you. I doubt you would be willing to concede that Halo made everything on the PS2 look dated though.

On top of everything else, the Dreamcast also had a unique and compelling line-up of on-line games, which only began to show in the months before and after it was killed off. It came too late to help the platform make money. It's an entire area of gaming that seems to have no value, and be disregarded as an experience people might want in the rush to condemn a platform as dated for not having PS2 graphics.

It is not a matter of how big the gap was, but where that line of its average performance lies, and where the performance of others continue from there.

DC games were good, but they started to feel outdated to me in the same manner 32/64 bit games felt outdated next to the DC ones (that said that doesnt mean I do not rate them high. I am still craving for many DC games). I am not referring solely to "static" visual quality, but to what could be done on screen, the amount of interactivity and physics on top of the visual expression. To me it looks like Sega was aiming to create a console that targeted mainly arcade quality gaming for home because they still had the same mentality they had with the Saturn. It was easy to get direct results fast, with the expected visual quality of the arcades (higher polygon models+480i/p resolution+smoothed textures). The DC was the perfect machine to play arcade style games, and the hardware did not leave a large enough room to do more things with other genres.

No really, I admit that XBOX had games that the PS2 could not dream of like Panzer Dragoon and Rallisport Challenge 2 (didnt like the original though at all) and Halo were awesome and there wasnt anything on the PS2 in respect to those genres that could compete directly in terms of visuals, but I didnt find many XBOX games that made me think like that, although on average it performed better and could do much more. The majority of popular/good multiplatform games performed similarly on both, and the PS2 exclusive games performed well enough to minimize the perceptual gap between the two consoles (still examples like DOA on XBOX looked better than Tekken 4 but some Tekken 4 stages like the mall minimized the perceptual gap). During the PS2, XBOX era I remember clearly when me and my friend we were bringing our consoles and comparing how the multiplatform games faired on each and the best each had to offer, just to see what was achieved on each, and how far PS2 games could go.

On average the XBOX clearly performed better. But the PS2 could in general reproduce or fake effects the XBOX could do, (although not always in the same quality and bump maps are out of the question). The DC persuaded me that it couldnt even "fake" effects that were considered standard. It looked like the biggest and main improvements Sega aimed for the console relied on increasing polygons, resolution and texture quality and almost stopped there. So it always left me with a sense of "something missing". Especially in physics and AI.

I think you have misunderstood me about how I view the DC. I am talking solely about performance here. I do value DC's games, and there are many many favorites of mine.


This figure sadly was for second gen titles, developers the first year were straggling to get 1,5MP/s...

I remember that the Le Mans programmer said that he achieved 4,5MP/s (he said 5MP/s in another interview at the time, good old round number (marketing...))

I don't know if this is true (4,5MP/s) but it was in 2001 and only one game...

I remember that claim. I think also that the media was comparing it with Gran Turismo's visuals. I tried with a friend that particular game to remember how it looked. Well, I dont know where they put all those polygons. MSR looked like it had far more polygons and looked better too to me.


Could you imagine the DC ever trying to pull off something even remotely close to this -

http://www.youtube.com/watch?v=iqCydukUqKY

It wouldnt have done much of a difference believe me. They would have came with a more expensive machine in 2000 that risked the company's existence more. Besides they needed to get some revenues back for the two years of absence. The Saturn was bleeding money so that was no solution. Would they have started developing games for competition and suddenly go back into developing soleley for their console?

Unless Sega could come up with a cheap solution in 1998 that could compete hardware that came two years earlier (Which would have been utter impossible) they had no choice.

edit: And not to mention the loyalty and name damage they caused to their selves unlike Sony and the Playstation that were massive.
 
Last edited by a moderator:
It is not a matter of how big the gap was, but where that line of its average performance lies, and where the performance of others continue from there.

DC games were good, but they started to feel outdated to me in the same manner 32/64 bit games felt outdated next to the DC ones (that said that doesnt mean I do not rate them high. I am still craving for many DC games). I am not referring solely to "static" visual quality, but to what could be done on screen, the amount of interactivity and physics on top of the visual expression. To me it looks like Sega was aiming to create a console that targeted mainly arcade quality gaming for home because they still had the same mentality they had with the Saturn. It was easy to get direct results fast, with the expected visual quality of the arcades (higher polygon models+480i/p resolution+smoothed textures). The DC was the perfect machine to play arcade style games, and the hardware did not leave a large enough room to do more things with other genres.

No really, I admit that XBOX had games that the PS2 could not dream of like Panzer Dragoon and Rallisport Challenge 2 (didnt like the original though at all) and Halo were awesome and there wasnt anything on the PS2 in respect to those genres that could compete directly in terms of visuals, but I didnt find many XBOX games that made me think like that, although on average it performed better and could do much more. The majority of popular/good multiplatform games performed similarly on both, and the PS2 exclusive games performed well enough to minimize the perceptual gap between the two consoles (still examples like DOA on XBOX looked better than Tekken 4 but some Tekken 4 stages like the mall minimized the perceptual gap). During the PS2, XBOX era I remember clearly when me and my friend we were bringing our consoles and comparing how the multiplatform games faired on each and the best each had to offer, just to see what was achieved on each, and how far PS2 games could go.

On average the XBOX clearly performed better. But the PS2 could in general reproduce or fake effects the XBOX could do, (although not always in the same quality and bump maps are out of the question). The DC persuaded me that it couldnt even "fake" effects that were considered standard. It looked like the biggest and main improvements Sega aimed for the console relied on increasing polygons, resolution and texture quality and almost stopped there. So it always left me with a sense of "something missing". Especially in physics and AI.

I think you have misunderstood me about how I view the DC. I am talking solely about performance here. I do value DC's games, and there are many many favorites of mine.

Fair enough. I guess a lot of this difference of opinion comes down to me thinking that the DC still had some way to go to show its best and also that the right games overcome (in terms of customer appeal) anything other than huge differences in platform capability. (If most people really cared about visuals, surely bad Pal conversions and composite cables would have died a long time ago!?)

The DC never really got beyond first generation software, and some of its coolest features were hardly touched on. I'd like to see what something like Shenmue would look like if they'd made most of it after they got familiar with final hardware. Some of its assets are horribly crude even by 2001 standards - the polygon counts showing on NullDC screenshots off the web don't seem very high.
 
MSR perhaps? (pre-release video):

I think the video shows the weakness of DC...it has vibrant textures but everything else look flat...there lies where i perceive the limitation of DC hardware wrt PS2 ..... watching the NFS video above, there are better shadows...smoke trails... dynamic lighting...interaction with other objects...PS2 games just looks lively..fuller...just my 2c
 
PVR2 was at 100MHz, capable for around 6 MP/s (100/16) FLAT shaded polygons...

If you add textures and effects that number is going down...

Sega said that the system could do something like 3 million polygons with textures and effects etc...

This figure sadly was for second gen titles, developers the first year were straggling to get 1,5MP/s...

3 Million Polygons realworld vs 20 Million Polygons realworld for the PS2.

That's a nearly 7 fold difference between systems that were part of the same generation.

Can you believe that?

That's how badly SEGA messed up by launching as early as they did.
 
3 Million Polygons realworld vs 20 Million Polygons realworld for the PS2.

That's a nearly 7 fold difference between systems that were part of the same generation.

Can you believe that?

No, actually.

http://www.technology.scee.net/files/presentations/PSP/HowFarHaveWeGot.pdf

And the DC wasn't getting 3 mpps either.

That's how badly SEGA messed up by launching as early as they did.

At least they didn't mess up as badly as if they'd tried to launch later with a more powerful machine.
 
Seems like the maximum they got in 2003 was around 7,5M polygons. I wonder how much they achieved until now.

I doubt they managed to reach anything nearly as much as 20 million

I seem to remember that some PS2 games managed about 20 million polygons. I think somebody even posted those informations here long time ago, that person had access to the performance analyzer, but I don't remember anymore who it was, or what games those were... I remember it wasn't any of the GT series, those had surprisingly low polygon amounts. Maybe it was nAo...
 
Just say the DC launched a year later with a DVD drive, dual SH4's doubled RAM, and a faster PVRDC (or two of them, since we're doubling everything else). There's no way the market would have accepted a $500+ console at the time, and the $400 Saturn-preceding launch price would be pushing it. But let's say they still managed to have record-breaking launch sales at $400. The system would cost more than twice as much in parts to manufacture, plus licensing fees for DVD playback (unless they went the Microsoft route and lumped that onto the remote control "add-on").

Anyway, Sega would be losing more money per unit sold than Sony. Money that they weren't making for a whole extra year with no console on the market. On top of higher R&D costs. On top of counter-marketing against the Sony launch. Sega was not healthy enough to cope with this scenario. They would have delivered an incredible system with some incredible launch titles, but it was still a transitional phase for the industry, so most of it still wouldn't have looked as good as Sony's second gen stuff. A strong company could have made the right marketing decisions to keep Sony at bay, and stayed in the game until the generation ran its course. But the Sega that existed in 2000 would have died. And it wouldn't have even taken a year of competition against the PS2 to kill them.

Most of those DC games we treasure wouldn't have existed. There'd be nothing left of the company to go on making software for other consoles. The Xbox would have probably had a weaker run without the hardware cues and software aid of Sega. Coming second last gen, Nintendo might not have drastically changed their strategy, and today we wouldn't have a Wii. Without Wii, there's no Natal, and without Natal, there's no serious push for motion controls on PS3.

That's not even touching on the effects of launching another console at $400. If it launched successfully at that price, what would that mean for the industry? If they had to drop it to $299, Sega would have lost more money and died faster.
 
Like I said, the 3 million polygons (with textures and effects etc...) figure, is what SEGA said about, what is the capability of DC (think 5-6 years lifetime with third generation engines...)

Did SEGA's figure was only marketing?

I don't know, but my estimation is that for 3rd gen engines (depending also on the priorities of the engine) this figure could be approached...

The problem for me, is that in the final years of the theoritical projected lifetime of the DC (5-6 years?) the code requirements for staff like physics etc.. will steal more CPU power than early titles, so the use of 3 million polygons would also depend on the type of game & engine also...


About the PS2, i think 20 millions is a high number.

In the first 2 years Namco programmers said that they achieved 3-4MP/s (i don't remember the game)

I saw the presentation (4,5MP/s highest figure, 145*30), maybe it was a Q1 2003 presentation, based on a 2002 Japanese or USA Performance Synth. Analysis.

My original estimation (2000) about the PS2 capability (3rd gen engines) was around 9MP/s (150/2=75 the theoritical highest, 75/8=9MP/s with textures/lighting/effects etc..., long talk...)

But I remember some claims of 15MP/s perf. of the Killzone programmer, but i don't know if this is true.

I think that the PS2 was at least 3X faster than DC in actual polygon capability (in game) and my estimation actually is 5-6X.

That's why i said that it will need at least 3X to approach PS2 polygon speed.
 
Seems like the maximum they got in 2003 was around 7,5M polygons. I wonder how much they achieved until now.

I doubt they managed to reach anything nearly as much as 20 million

Yeah, it's pretty amazing. This presentation goes into even more detail:

http://www.research.scea.com/research/pdfs/GDC2003_Intro_Performance_Analyzer_18Mar03.pdf

Look under "An optimal Example (cont. 3)". Optimal example is 7.2 mpps, so assuming it's the same game that's 50/60fps! Judging by some of the pictures used in the presentation, that might be from Naughty Dog.

It's all the more amazing because the average number of pps for all the games they sampled by 2003 was about 2.2 million (unless I screwed up my calculation: ((30 * 60 + 60 * 40)/100) * 52,000 = 2,184,000) ).

With regards to the DC, clearly it can't get near the 7.2 figure, but it could get much closer to 2.2 even back in 1999 - 2001. Heck, I'll bet most 2006 PS2 games weren't reaching the 7.2 million figure. The DC only had to last until about 2004...

I think it goes without saying (but I'll say it anyway) that careful use of bumps on flat looking, low poly surfaces would have helped add to the appearance of detail in DC games.
 
Maybe we were all Sony fanboys during the day (ahem), but quite honestly, the DC just came at the wrong time. When the DC came out, I was just getting into my PSone library with games such as Final Fantasy IX and others. And knowing that the PS2 would be around the corner eventually with successors to all the games I enjoyed playing was just good enough to keep me going.

Of course, the DC looked great when it launched - awesome visuals, low price etc. Didn't like the controller though and the games that really interested me were also ported to the PS (Tomb Raider IV). Also, while the DC had great graphics, they were nothing that exciting given that the PC was around the same ballpark at the time with Half Life taking a lot of its shine away.
 
I don't think you can blame the performance per se. Here's probably what killed it:

1. A bunch of features few people wanted, which drove up Sega's costs and left less money for marketing the machine and developing software, like those idiotic VMUs, VGA output, and a modem (yes, yes, online gaming is teh aweXom3z, but it about broke Sega's bank).

2. A legacy established by Sega CD, 32 X, and Saturn of releasing hardware, supporting it for about four days, then abandoning it. 3rd parties were understandably reluctant to jump on the Dreamcast train, even when initial sales looked promising. It became a self-fulfilling prophecy, as Sega supported the DC for all of about 2 years.

3. A lack of really compelling 1st party software. And by really compelling, I mean, "drives sales to the tune of several millions of units," not, "remembered nostalgically," or "got good reviews." Sonic Adventure broke 2m, but other than that, nada. I think Nintendo has shown with NES, Game Boy, DS, and Wii that you don't need cutting-edge hardware if you take the lead with software yourself and can deliver hit after hit.

4. The controller sucked; it's a simple step backward from the controllers of the 32-bit era. Yes, it's a step forward from Saturn's controller...but that was a relic of the 16-bit era to begin with. One thumb stick? Six buttons? Really? Either go with the standard, or take things in a new direction. Regressing is unacceptable. Also, it was uncomfortable to hold.

Basically, Sega didn't know what the hell it was doing or even trying to do. It had completely played out the "Console war! We are so much cooler than Nintendo!" schtick, and once they played that out, they were plain out of ideas. The Dreamcast shows that.
 
4. The controller sucked; it's a simple step backward from the controllers of the 32-bit era. Yes, it's a step forward from Saturn's controller...but that was a relic of the 16-bit era to begin with. One thumb stick? Six buttons? Really? Either go with the standard, or take things in a new direction. Regressing is unacceptable. Also, it was uncomfortable to hold.



I disagree with you in saying the DC's controller is a step forward from Saturn's controller (unless you're talking about the analog controller that came with NiGHTS). The Saturn controller was awesome for fighting games of all types and much better than Playstation's in that regard. Its d-pad was also superior in every single way. ABC and XYZ are great as well.
 
Basically, Sega didn't know what the hell it was doing or even trying to do. It had completely played out the "Console war! We are so much cooler than Nintendo!" schtick, and once they played that out, they were plain out of ideas. The Dreamcast shows that.

Online gaming, generating revenue through online services, VGA support, motion controllers, and less complex controllers to attract casual/none gamers don't strike me as things that would come from a company that's plain out of ideas. They might not have been able to make their ideas work to their advantage (everyone else did instead), but the last thing you can say about the DC is that it's from a company that's out of ideas.

Adding VGA support to a console with full height screen buffers that outputs RGB anyway, when you've already develop it for the arcade machine, is not what bankrupted Sega. Neither are VMUs - they might have been largely pointless but they sold for a profit and thanks to their limited storage you had to buy lots of them (too many).
 
I disagree with you in saying the DC's controller is a step forward from Saturn's controller...The Saturn controller was awesome for fighting games of all types.
I already said it was a relic of the 16-bit era; no need to repeat me. ;)
function said:
[Sega's stupid things] don't strike me as things that would come from a company that's plain out of ideas.
Quite to the contrary, throwing a bunch of random, useless, money-sucking features into a product are a key identifier of a company that lacks vision and direction. What do you do when you're out of ideas? Start throwing poo at the wall and hope that eventually, something sticks.

Also, point of correction: SegaNet did not generate revenue, it absorbed it. That was a source of negative cash flow.

Edit I might add that this company's insistence on making HD consoles the lead platforms for Sonic games shows that this company is run by lunatics. Let's see, Sonic games sell best on Nintendo platforms, the series obviously appeals more to kids...so, the obvious thing to do is make more HD Sonic games with throwaway Wii ports!
 
Last edited by a moderator:
Status
Not open for further replies.
Back
Top