Consoles that provide(d) the best performance increase from 1stgen to last gen titles

[Hazuki Ryu]

Ofcourse there will be always improvements, not huge ofcourse i think thats an idea that playstation fanboys always have in their minds, always hoping that sony hardware will have some kind of hidden power that will be unleashed when the developers find the "secret ways", lol its funny theres always room for improvement though, devs cant always get everything right at the first time some times not even at second time, they learn and they improve in sequels its a natural cycle.

Just look at halo and halo 2 on the xbox i think thats about as big as improvement can get.

 

[ninzel]

Agree with PS2. Just look at Jak1 vs Jak 3.

 

[function]

The console which showed the biggest improvement within a generation was the Saturn: the progression from the first Virtua Fighter to Shenmue Saturn or NiGHTS, or from the first Panzer Dragoon to the later iterations.

I'm with you on this. The improvement the Saturn showed in the first 18 months dwarfs what the PS2 has seen in its entire lifetime. It wasn't enough to see the Saturn cling onto life though, sadly. Powerful multicore CPU arrangement with a GPU (or two) that you had to coax competitive performance out of - rings a bell even today for some reason ...

Saturn Shenmue was aborted in 1996 (iirc). While impressive, it's not even fully representative of what the machine was ultimately capable of. PDS probably has that accolade for my money - very detailed, amazing lighting, and a cool real-time morphing dragon. Amazing stuff.

 

[swaaye]

Having been playing it regularly recently, I think it's a tie between the N64 and the PS2. On N64, compare Turok: Shadow of Oblivion and 007: The World is Not Enough with Turok 1 and Goldeneye. Compare Battle for Naboo with Star Fox 64, or look at Jet Force Gemini vs Shadows of the Empire. Or look at World Driver Championship vs Crusin' USA. I think the gap is as big as anything that happened on PS2.

Waverace 64 looked really good, but it also was very little other than special effects. There wasn't much geometry in the levels and very little in the way of textures. It had great water effects and lighting, but that was it. That sort of approach really doesn't work for very many games.

I'd look at Top Gear Rally or Automobili Lamborghini vs. World Driver instead of Cruisin'. N64's biggest prob, IMO, was a lack of developers. Cruisin' USA wasn't really an attempt at a racing sim. It wasn't even a remotely good port and I'm not sure what to blame that on. Hell, the arcade game is mostly based on 2D texturing instead of 3D models.

Naboo was decent though, and its constant 640x480 was nice. I picked it up about a year ago actually. BTW, it's amazing how awful the sound is in most N64 games. Naboo has probably the best sound system of any N64 game, but it's also still painfully bad at times.

 

[fearsomepirate]

IMO, Turok 2 has the best sound of any N64 game I've played. Mario 64 isn't doing anything special, but it works very well for what it is. Goldeneye's sound, on the other hand, is beyond atrocious. In general, voices on N64 sound really, really bad.

But then, I'm also not someone who notices sound very much, whether good or bad.

 

[swaaye]

Well I meant that Naboo has (I think) the most technically impressive audio. It uses Factor 5's Musyx sound system and does Dolby Pro Logic encoding on the fly along with some environmental effects. Indiana Jones used it too, I guess.

 

[Mobius1aic]

Comparing the first PS1 games to the last good ones is quite a trip. Just compare Ace/Air Combat to Ace Combat 3: Electrosphere............so much difference. Oh and let's not forget Omega Boost

 

[Mr Deap]

For me, the almost all the top game all look the same in performance. There's usually a performance trade off in the graphic & how the game is designed for the first & last title.

It may look better in people eyes, but the game is just designed in a different way to fit the limitation of the hardware.

 

[fearsomepirate]

Oh. I guess I just notice how "good" or "bad" it sounds and can't pick apart the technical details. That's probably what designers want anyway.

 

[JacksBleedingEyes]

Best performance increase?

By your statement we seem to be putting the PS2 on some sort of pedestal due to bad architecture design and bad development tools.

Maybe you should change your thread title to
"Which console managed to get its act together and provide better tools for it's devs to work with"

Look at the cube for example. Rouge Squadron which was one of the first games to come out is still one of the most beautiful.

I think you need to find a balance. You don't want to make your development tools to difficult, but you don't want to make them to easy to use either (lazy developers). You want to give as much freedom as possible yet still work within technical limitations.

 

[swaaye]

Yeah I would have to say that the Playstations have seen major changes in visual quality throughout their lives. IMO, this is because they have such incredibly long lifespans. Game design techniques change a lot over such long periods of time.

PS1 saw the most gains though because it arrived when 3D was just getting started and the end of its life was after years of huge refinement in 3D tech. Hell, I think PS1 was ironically almost designed as a FMV machine (as in movie games) because they were all the rage when PS1 was hitting the market.

I think it would be fascinating if the best and brightest made a PS1 or N64 game today, knowing what they know today.

 

[fearsomepirate]

By your statement we seem to be putting the PS2 on some sort of pedestal due to bad architecture design and bad development tools.

What you call "bad" others might call "forward-looking." There's always a trade-off between making something familiar and easy to use and making something with as much raw power as current fab processes allow. There really wasn't anything like the PS2 when it came out. Heck, fixed-function T&L was all the rage when it was in development.

You're probably right about the tools, though. By all accounts, it was poorly documented at the beginning.

 

[archangelmorph]

I think you need to find a balance. You don't want to make your development tools to difficult, but you don't want to make them to easy to use either (lazy developers).

That's the worst thing i've ever heard..

For starters, WHAT LAZY DEVELOPERS..?

Secondly, You DO want to make your tools as easy to use a possible.. The only benefit you as a platform holder can provide will be efficient tools, libraries and documentation for writing code painlessly on the platform..

The easier it is the more productive your developers can be and thus, the more resources they have available to plough into weening power out of a system at their leisure..

At the end of the day as a platform you have to cater to the widest possible demographic of developments from the $20 million budget with more than enough resources to practically re-write firmware as well as build a stinkingly fast engine on top, to the small-scale developer catering to a niche (whilst trying to stay afloat in an industry where development budgets proceed to escalate faster than they can keep up), to the licensed IP dev with very strict deadlines, tight budget and high expectations for product quality..

The easier you make your platform in terms of accessibility with respect to getting the ball rolling, the more developers (of every size & shape) will consider investing in your platform & the greater depth and breadth your software library will be (which can have a positively solid effect on hardware sales..)

 

[Dr. Nick]

That's the worst thing i've ever heard..

For starters, WHAT LAZY DEVELOPERS..?

Yeah, even some of the worst games people put a lot of work into.

I gather that when he said lazy he was talking about graphics. While I would say graphics does contribute alot to the sales of games, almost all games make their debut with screenshots or video, so many of us rely on what we see in screenshots for our first impression of the game so the graphics better look good at that point. This is even more important when the game is in a genre with heavy competition. Once the game hits the market or is about to though the importance of the graphics go down because at that point everyone knows what it looks like and now want to know how it plays and that is the reason why devs need easier tools. The most important thing is to make sure the game is fun. They can't waste too much time getting stuff up and running and stable, they need to focus on making sure that the product they release is actually good or risk doing all that work for nothing or next to nothing. So gamers should want dev tools to be as easy as possible. Hardware manucfactures are the ones who should be the most worried about getting better performance out of the machines and passing that info on to developers.

 

[Lazy8s]

Getting greatly outperformed by a system with two-thirds its silicon doesn't demonstrate that Sony's PS2 architectural choices created high performance ceilings. Developers were simply started further back on the PS2 learning curve, which led to less games overall that got to take advantage of the PS2's hardware.

Those choices really were bad, not forward looking: coupled vertex and pixel processing; a lot of die space not used on logic for calculation; silicon wasted on high theoretical peak capabilities, like in triangle set-up; broken capabilities, like mipmapping.

 

[Jov]

snip...

Following your logic, doesn't this also applies to the Saturn for its 2nd graphics chip/core, and N64 sticking with cartridge?

If we want to nitpick there are trade-offs in all consoles, but we have to consider the competition and technology during that era, and as many has stated there was nothing like the PS2 at the time.

 

[Lazy8s]

PS2 ran its fully custom VF4 build (a game which would test with well-above-average hardware utilization in the Performance Analyzer) at less than half the performance of NAOMI2's, a system which was also released in H1 2000, version, and the PS2 was around one-third more costly in silicon. PS2's performance wasn't good for 500+ mm^2 of die size; it wasn't even good for 300 mm^2.

Not only was there something like the PS2 during its era, there was something a lot better. IMRs couldn't keep up with PowerVR's tilers, especially IMRs as inefficient as Sony/Toshiba's graphics system.

 

[Phil]

OMG! Lazy8s and the NAOMI2 discussions again! Wow, I'm sure if we go back to 2002, we'll find more than enough of the topic which has been done to death time and time again. Didn't we already establish years ago that such a comparison isn't as cut&dry as you like to put it?

 

[Crazyace]

Naomi2 was 3 custom chips + 96MB of video ram ( 64MB framebuffer/texture + 32MB geometry memory ) so I'm not sure how well that compared to the PS2 GS

At least compare like for like - DC having trouble with VF3 against PS2 running VF4

 

[Lazy8s]

Several specs of VF3, like resolution, were higher in the DC port, and the developer, Genki, reported that geometry counts were similar but had to be modeled differently.

VF3tb was a launch game anyway and not an above-average DC effort overall.

The combined expense of the silicon for the PS2's vertex processing vector units, the cost premium Rambus brand and DDR type -- necessitated by the system's bandwidth inefficiencies -- unified memory used for part of the graphics processing, and the GS bloated far beyond that of NAOMI2's graphics system, two-thirds of which was just old Dreamcast parts.