Sega_Model_4
Newcomer
Could Dreamcast et al handle this/that game/effect? *DC tech retrospective *spawn
- Thread starter xaeroxcore
- Start date
-
- Tags
- sega
xaeroxcore
Newcomer
Suddenly the DC looks like the most underutilized and underrated console.
Just makes you wonder how DC and PS2 would look like if both could be utilized to their fullest with the knowledge and tools available today. The DC and the Saturn get crazy homebrew projects. PS1, PS2, GameCube, XBOX have barely anything.
Just makes you wonder how DC and PS2 would look like if both could be utilized to their fullest with the knowledge and tools available today. The DC and the Saturn get crazy homebrew projects. PS1, PS2, GameCube, XBOX have barely anything.
Sega_Model_4
Newcomer
These other consoles were properly used in their lifetimes. DC and Saturn had short lives and much less attention from developers.
Surely we can say the other consoles were better utilized. Considering when the DC was released and how much more straightforward it's HW was compared to PS2, you would have thought it that we knew it's limitations better.
The recent homebrew show a completely different picture. Unexpected one. We don't know how much of that is due to a)better modern tools and technical knowledge and b) how much of that is due to the untapped power considering the available knowledge of that time
Would developers have managed to create such impressive feats if the DC lived fully with the knowledge of the time?
We see handful of people porting and creating impressive tech demos on their spare time that surpass whatever whole software houses and Sega themselves could pull out back then.
So there is also the question how much could the PS2 be pushed with the knowledge and technology of today?
Absolutely. There is knowledge and paradigms that could be brought to bear on PS2 for different results. It'd be wrong to think the best PS2 achieved in its lifetime is the best it could achieve for all time. And TBH the only people wanting to use current trends as 'this proves DC was better' are fanboys warring. The actual interest here is 1) DC was far more capable than it appeared at the time, and the hardware design had some great features. 2) All hardware had room to grow and still does; the limits of these retro consoles haven't been reached yet.
I think the short lifespan of the DC combined with it's very advanced feature set make the differences between "then" and "now" particularly stark. Another factor is that game development changed particularly rapidly over the 3~5 years following the DC's early death.
Streaming for storage, which the DC was actually pretty well equipped for, became more and more of a factor and the tools and techniques for developing assets improved radically. 3DS Max exploded in popularity and capability over this period, and Z-Brush made creating amazing looking normal mapped low poly assets very attainable.
What's so great about DC GTA3 that @EsppiraK is posting is that it's using assets and streaming requirements from a contemporary game (actually even a little more detailed than the PS2 version) and it's running them using an SDK that is still, afaik, missing a couple of features that official Sega SDKs had around 2001. It's a handful of passionate and really talented indie developers applying what was known about the hardware by talented developers back in the day to a software problem (GTA 3) that only came around after the DC was dead.
This is not to say that the PS2 didn't also have more gas in the tank to accommodate further evolution and growth. For example, with it's insane fillrate and massive vector processing power it could probably have begun to move into some "pixel shader" like areas. There's always the chance that some new way of using old hardware can dreamt up!
This Megadrive demo was by a team that found a couple of new tricks in old machines toolbox. All without a framebuffer! And this was 29 years after the Megadrive came out!
Streaming for storage, which the DC was actually pretty well equipped for, became more and more of a factor and the tools and techniques for developing assets improved radically. 3DS Max exploded in popularity and capability over this period, and Z-Brush made creating amazing looking normal mapped low poly assets very attainable.
What's so great about DC GTA3 that @EsppiraK is posting is that it's using assets and streaming requirements from a contemporary game (actually even a little more detailed than the PS2 version) and it's running them using an SDK that is still, afaik, missing a couple of features that official Sega SDKs had around 2001. It's a handful of passionate and really talented indie developers applying what was known about the hardware by talented developers back in the day to a software problem (GTA 3) that only came around after the DC was dead.
This is not to say that the PS2 didn't also have more gas in the tank to accommodate further evolution and growth. For example, with it's insane fillrate and massive vector processing power it could probably have begun to move into some "pixel shader" like areas. There's always the chance that some new way of using old hardware can dreamt up!
This Megadrive demo was by a team that found a couple of new tricks in old machines toolbox. All without a framebuffer! And this was 29 years after the Megadrive came out!
The games that were being developed at the time mattered as well. Most of the Dreamcast library, outside of the first party stuff of course, are PS1 ports. So they were somewhat limited by the lead platform as well. And I think that's why better looking titles for the system were exclusives. And maybe why some of the earlier titles are some of the nicer looking ones, because there were less exclusives and more ports.
I believe the DC wasnt exactly getting PS1 ports but rather the PC ports that were also designed for PS1. It was getting in other words the best versions of multiplatform games. But yes it is true that a lot of those werent pushing the console. A lot of these games were designed around mainstream hardware, whereas the DC could punch above mid specs
DC had some exclusive games that trully shined also from third parties. Sega exclusives were a mix of jaw dropping visuals and mediocre. Blue Stinger sucked. But Sonic Adventure in my eyes back then was like Toy Story in real time. Blazing 60fps in fully realized environments with high quality textures was something else. But curiously it lacked killer racing and fighting games developed by Sega that could impress as much as Sega Rally and Virtua Fighter 2 did on the Saturn. Virtua Fighter 3 was outdated next to DOA2. Sega Rally 2 was no where as good looking as its arcade counterpart as it was handled by a third party that did a mediocre job at it. Other ganes like Headhunter wasnt particularly impressive either.
The PS2 despite being a pain to make games on, it was getting a lot more jaw dropping games much sooner in its lifespan. Tekken Tag and Ridge Racer V alone obliterated visually anything that came on the DC in the whole duration it lasted. A lot of games exhibited state of the art visual effects (like dynamic water and dynamic shadows) that were barely seen on the DC.
Putting aside the real capabilities of the hardware, the perceptual difference in quality was huge between the two consoles very very early into the PS2's lifespan with games like SSX, The Bouncer, Tekken Tag etc.
Last edited:
Sega_Model_4
Newcomer
If you read interviews with PS2 game developers, you'll see that everything was done to get the most out of the console. Kojima's team even programmed parts of MGS2 and 3 in assembly language to get the most out of the visuals.
Team ICO managed to emulate the fur effect in Shadow of The Colossus which the console's GPU was not capable of doing (like GameCube GPU).
Developers had much more time and money to work on PS2 versions.
The Dreamcast, in addition to having a weak base from the Saturn, which had already driven away some developers, didn't last long, games sold less on the platform and even development tools weren't very good, considering that Sega had a habit of keeping the best for itself and letting third parties fend for themselves (the partnership with Microsoft and Windows CE was an attempt to get around this). Sony used to help develop games from partner companies in exchange for exclusivity, so many studios had access to the best that Sony had to offer in development.
Your argument is that PS2 was maxed out, has nowhere else to go, and now DC finally proves its superiority?
PS2 libraries were non existent for an extended period when it launched and the documentation was lackluster and in Japanese. Kojima was probably forced to rely on Assembly because they had no proper libraries to work with when they received the console.
Developers were struggling as much as they did with the Sega Saturn on PS2
To the contrary, back then according to interviews from Sega themselves with the DC they aimed to fix the mistakes they made with the Saturn and that included a friendlier game development environment.
Sony on the other hard repeated Sega's mistakes with the hardware. Developers were complaining about PS2 and how it could kill smaller dev houses but interviews appeared highly positive for the DC.
Nothing about PS2 was straightforward.
It's true that the DC died too soon and that Sega wasn't sharing all of its tools But Sony wasn't sharing much either because they didn't have much they could share since the hardware was a pain by default. Even with Sega's best efforts near it's death, the DC was lackluster in comparison to PS2's first games.
The PS2 had an awful start with developers. Surely at the end the PS2 pulled off some neat tricks as it matured. But it doesn't mean all possibilities were squeezed out.
So I am also trying to understand what's the point that you are trying to make. That ultimately the DC was overall better than the PS2 in capabilities?
It also ignores the point it replies to, that thinking has changed and there are entirely new paradigms to consider. I dare say PS2 could do quite well with SDFs, something that didn't exist in its heyday.
It's a curious response to a post that said the only people who care which was ultimately maxed out are fanboy warriors. I can't interpret it as anything other than that, wanting to identify the DC as the best despite the differences in when and what was achieved. As mentioned elsewhere, every piece of hardware that ever had fanmade creations decades after its release has shown incredible, untapped potential, and it's (I dare say wilfully) myopic to think otherwise for PS2.
Last edited:
EsppiraK
Newcomer
Who is talking about superiority, you keep repeating it like a parrot, we are talking about maxing about the console not it being more powerful than the PS2, it isn't.
And lol about new paradigms, you have no idea how this ports are being developed.
The fact is that these impossible ports due to power, memory, storage etc etc, are running and are payable from Start to finish on a console it wasn't supposed to even boot them, and they are still in beta.
The console was very underutilised, it was easy to develop for that doesn't mean it was easy to squeeze all it had to offer, actually quite the opposite.
Again the PS2 is more powerful stop parroting that we say the Dreamcast was superior only a fanboy or delusional would say that....
It seems like Sega_Model_4 is. I don't understand why they responded as they did to my post if not that.
I'm not addressing 'we'. I'm addressing Sega_Model_4. If Sega_Model_4 isn't saying that, they can explain how I've misread their post.
It's nothing to do with these ports. My statement is following on from Nesh's about the homebrew scene and how lacking that is on PS2, and how much might be possible on PS2. The current arc of discussion is, "DC was so much more capable than appeared in its lifetime thanks to homebrew efforts realising its potential; what could PS2 achieve if it also had that level of homebrew efforts?"
What relevance has this to anything said? No-one's disputed that. It's tangential to the comment about how old hardware can be stretched even further decades after release when talented devs continue to work on them.
I don't think that's even a valid statement. 'Most powerful' is by and large a ridiculous measure as each machine had its strengths and weaknesses and there's no standard for averaging these into some 'more powerful' metric. Discussion can ignore trying to decide 'what was the most powerful' and just focus on the specifics and the accomplishments. Look how great these ports are. Wonder what else is possible on different machines.
Last edited:
I'm taking @Sega_Model_4's comments to mean that within its lifetime, the PS2 was ultimately well utilised due to what was learned about the machine and games often being designed specifically around what was learned. Early on PS2 development was extremely difficult, but with the resources and the talent applied to the machine it ended up being well very well. Huge amounts were learned about how to use it, and a massive base of experience was developed.
I'd agree with this, and add that due to the incredible commercial success of the machine and it's long life, developers got to make multiple games on the platform and apply that learning to heavily reworking their engines and code base and production pipeline. This is not to say that the PS2 was not capable of more that it achieved over its commercial lifespan.
The DC never really saw anything like this. PS1 ports, or speedy conversions of hastily made arcade games, produced over a very narrow window of time, can't compare. Shenmue 1 and 2 were huge achievements for the time, but the second game wasn't a made from the ground up with everything learned from the first - the games were in development simultaneously with most of the assets from Shenmue 2 pre-dating the release of Shenmue 1 and in many cases pre-dating the launch of the console itself. Probably the closest we saw to the progression the console was really capable of was Sonic Adventure 1 to Sonic Adventure 2, but even then the latter part of development was done after the console had been cancelled. Even the WinCE versions the machine got turned out to be terribly inefficient, and only slowly getting less bad over time.
The scale of what the homebrew community on the DC has done and demonstrated would not be possible on the PS2 simply because PS2 actually got a fair crack of the whip. Again, this is not to say we saw everything the PS2 could have done, but enthusiasts on the DC are filling in some of the gaps in software progression that were never there on the PS2 because it had a full and happy life.
For a long time the narrative in the gaming community was that it wouldn't have mattered if the DC lived longer: it was already maxed out because it was so easy to develop for and it would only have fallen further and further behind. And that had to be true because there was no software that proved otherwise. GTA3 - being the defining game of the first half of the generation - was often the poster child for this. "It couldn't have handled GTA3, it couldn't have stayed in the generation, it was a mid gen stop gap" etc etc.
In an attempt to hold onto this (IMO mistaken) idea, there's been a lot goal post moving since GTA3 tech director said basically "yeah DC could have done GTA3, the move to PS2 was for commercial reasons". I won't go into that because it's not productive ... other than to say I had a good old laugh when the talk got to somewhere around "... but Vice City was more advanced, and it probably couldn't do that, so DC GTA3 doesn't change anything!". And then bam, DC Vice City alpha drops. LMAO!
I'd agree with this, and add that due to the incredible commercial success of the machine and it's long life, developers got to make multiple games on the platform and apply that learning to heavily reworking their engines and code base and production pipeline. This is not to say that the PS2 was not capable of more that it achieved over its commercial lifespan.
The DC never really saw anything like this. PS1 ports, or speedy conversions of hastily made arcade games, produced over a very narrow window of time, can't compare. Shenmue 1 and 2 were huge achievements for the time, but the second game wasn't a made from the ground up with everything learned from the first - the games were in development simultaneously with most of the assets from Shenmue 2 pre-dating the release of Shenmue 1 and in many cases pre-dating the launch of the console itself. Probably the closest we saw to the progression the console was really capable of was Sonic Adventure 1 to Sonic Adventure 2, but even then the latter part of development was done after the console had been cancelled. Even the WinCE versions the machine got turned out to be terribly inefficient, and only slowly getting less bad over time.
The scale of what the homebrew community on the DC has done and demonstrated would not be possible on the PS2 simply because PS2 actually got a fair crack of the whip. Again, this is not to say we saw everything the PS2 could have done, but enthusiasts on the DC are filling in some of the gaps in software progression that were never there on the PS2 because it had a full and happy life.
For a long time the narrative in the gaming community was that it wouldn't have mattered if the DC lived longer: it was already maxed out because it was so easy to develop for and it would only have fallen further and further behind. And that had to be true because there was no software that proved otherwise. GTA3 - being the defining game of the first half of the generation - was often the poster child for this. "It couldn't have handled GTA3, it couldn't have stayed in the generation, it was a mid gen stop gap" etc etc.
In an attempt to hold onto this (IMO mistaken) idea, there's been a lot goal post moving since GTA3 tech director said basically "yeah DC could have done GTA3, the move to PS2 was for commercial reasons". I won't go into that because it's not productive ... other than to say I had a good old laugh when the talk got to somewhere around "... but Vice City was more advanced, and it probably couldn't do that, so DC GTA3 doesn't change anything!". And then bam, DC Vice City alpha drops. LMAO!
Cloofoofoo
Regular
I think it's a misconception in that way of thinking. It only did well because of "today" way of thinking. The true boost of performance that made these port possible and edging close to 20+ to 30 fps actually came from something available day one. Sh4 special features ocindex / ocache to use as super fast scratch pad. These uses were even listed in the documentations both English / jp but I guess devs didn't go that far to optimize because from what I hear is extremely hard but can used for multiple things not just tnl.
Funny part that way it works mirrors how PS2 programmers could break models up to be cache Friendly to its co processor to send as quickly as possible( was it vu1?) . Well turns out due to the ocache stuff on the sh4 the Dreamcast could do exactly the same. But since this was really underused you wouldn't know till the gta3 ports on dc , it's actually the main reason they weren't working on emulators because this was underused it was unemulated / not done correctly. Didn't matter since no game used.
I hope falco comes around to explain it better and why it mattered. Also he realized it was the same crap done PS2 for speed.
What you describe is still impressive and makes me just as curious. Whole development houses avoided using it even though it was in the documentation because it was too difficult. Yet only handful of people managed the impossible port by taking advantage of it. What else is in documentations that devs barely used or not used at all? And how did a handful of people in their spare time manage such incredible feats when devs avoided such hardware features? There is a similar case also with PS2 where devs barely took advantage of one of the VUs
I know you weren't asking me, but I'll chime in and say that it would probably have got used more had games continued to be developed on the DC. More challenging parts of a system or architecture (or any tool or device really) tend to get used or tackled when you actually need to. SH4 was a new chip as it was and exploring uses of powerful but exotic capabilities wouldn't have been a priority at first.
PS1 ports didn't need it, WinCE for the DC was poor so it wouldn't have used it, arcade games and their ports had short development cycles and assets weren't particularly 'high poly' at first. The Japanese focused Ninja 2 dev kit had significantly improved T&L performance over the first version but it was never used due to the DC being terminated. It seems likely that by Ninja 2 Sega were delving deeper into the capabilities of the SH4, and possibly what Cloofoofoo is talking about.
If the DC hadn't died more of the CPU performance it's been proven to have would have been extracted simply because it would have to have been. Just like it had to be for these awesome homebrew ports of GTA3 and Vice City. That's my take anyway.
Absolutely. The first games you make you release to get something out, learn the rudiments of the platform, and make some money to keep the company going. Competition is against other devs doing exactly the same instead of taking 5 years to learn the hardware before their breathtaking first release - all those launch titles are a bit jank. Second game, you now know the basics and can consider something a bit more adventurous. Third game, you can start to consider best practices. Fourth game, maybe you get ambitious.
And I think that in part explains why homebrew devs do so well. 1) They aren't creating new games but are porting. They aren't balancing new engines on unknown new architectures with art assets being developed at the same time. "Hey, we've found out we can double our poly counts." "Well it's too late to redo the art now. Next game." And if there is no next game, it never happens.
2) Homebrew can spend as much time as they want exploring possibilities without any publishing deadline or commercial pressures. It ends up being an academic exercise that can indulge in pursuing the very best of the hardware. Plus motivation is pride, in their own abilities and in the platform they love and are dedicating all this effort to.
Putting it another way, look what modders accomplish on PC. How come these big studios can't produce the same quality? How come modders in their free time manage to make the visuals better, up framerates, add features, etc. improving on the efforts of several hundred highly paid, dedicated developers working for years?