Printable View
I gotta say that's a lot higher than I thought it would be. Might be worth throwing the Gamecube version at Dolphin and seeing how it performs. It might also help us figure out how accurate those PCSX2 numbers are if we can't see a noticeable difference in geometry.
You kinda read my thoughts... Double-checking is always good.
I had check "The Bouncer" before Burnout and the numbers were much lower though.
Supposedly, Alex Ward stated that the first Burnout game peaked at 13 million polygons per second but the only source linked to it is pretty dead and Internet archive doesn't have a copy of it.
http://www.sega-16.com/forum/attachm...4&d=1395905062Quote:
Originally Posted by Barone
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
Time to do the math:
Burnout (2002) (Criterion Games) (GC)
In game polygon count (by Dolphin 4.0): 92564 polygons per frame
About the scene: Single race at the Interstate track showing the usual traffic cars and some buildings, without the loads of special effects used in collisions and burnout/nitro mode.
Frame rate of the game: 60 fps.
Result: 60 x 92564 = 5,553,840 polygons per second ~ 5.6 million polygons per second.
I didn't notice any obvious increase in the geometry complexity of the cars and buildings in this version but I have played it very briefly and some reviews of the time claimed that the car models were improved. The draw distance is better but the 3D models are the same IMO.
So I guess it's safe to say that the PCSX2 numbers are fairly reliableand maybe they were also lowered a bit more due to the emulation errors - since it's easy to notice that it was not drawing all the polygons that it should be drawing(it seems that even when it doesn't texture it right the polygons are still there, even in games with far more graphical errors).
Anyway, what did sheath say about my posts creating errors in this thread? lol.
http://i.imgur.com/a6kmM7Q.gif
So basically if the Dreamcast wanted to run Burnout at the very least it would either need to drop the frame rate to 30fps or lower the geometric detail.
I'm interested to look at Resident Evil 4 now, to see if sheath's 900k Polygons per second claim for the PS2 and 3 Million Polygons per second for the Gamecube holds any water.
To not talk about the AI cars which deal with heavy traffic and the collision model which probably would eat some extra CPU time when compared to most of the DC racers.Quote:
Originally Posted by TrekkiesUnite118
I'll try Resident Evil 4 ASAP but, you know, the success rate of this PS2 emulator is very low. Several games already failed with me.Quote:
Originally Posted by TrekkiesUnite118
Dolphin seems to be both much more compatible and way more optimized.
I always read those numbers as being merely theoretical, now I grant you that there are probably a few people out there who were suckered in to believing that the 75M Poly/sec as being real in game performance and along with the whole "Emotion Engine" and "Toy Story level graphics in realtime" stuff Sony really did hype the PS2 to an extremely high degree, but I certainly didn't and I don't think anyone else on this forum does either, I also don't recall any magazines etc promoting that figure as real in game performance. However as far as outrageous hype goes it really is no different to what Nintendo did with the N64 (they codenamed the thing Project Reality ffs, not sure what kind of reality Nintendo was referring to unless it was the kind that is washed out, blurry and full of fog - yes I do know that it was called Reality after the SGI Reality Engine based chipset btw). Then there was 3DO/Panasonic with the M2 I still vividly recall the pre-rendered videos and screen grabs of supposed M2 games that 3DO claimed were running in realtime, it got even more absurd when Panasonic bought out the M2 system, upgraded it and then claimed that it was now as powerful as Sega Model 3 arcade system. There probably tons more promises of empty hype out there in the industry but Sony's PS2 hype is the one that gets singled out because the system was a success, imo the PS2 was a success more because the PS1 was a success not so much because of Sony's PR.Quote:
Originally Posted by sheath
Quote:
Originally Posted by sheath
Yes I am aware that the PS2 does not cull hidden polygons etc, I would hardly call this a loop hole and therefore presumably an advantage for the PS2, in fact I would say that compared to the HSR/tiling architecture of the PVR2DC that it is actually a disadvantage for the PS2 since it is having to draw all these extra primitives and I'm assuming that the only reason it works on the PS2 is because of the huge fill rate and memory bandwidth of the GS in the PS2, in pretty much any other competing systems (eg traditional PC based GPUs's etc) such a methodology would kill the GPU's performance pretty quickly. And TrekkiesUnite makes a good point when he asked whether in order to do the kind of effects that the PS2 is doing in games like the GT games that the DC would also have to allocate extra primitives to replicate the same effects. I also recall that rusty mentioned that on the DC with certain multipass effects that it requires TWO passes to apply the same effect when the PS2 doesn't. That would presumably mean that in certain multipass effect that the DC would be slower in generating these visuals compared to the PS2?
Quote:
Originally Posted by TrekkiesUnite118
Quote:
Originally Posted by Barone
@TrekkiesUnite and Barone
Great job at getting some actual real world in-game performance figures for these systems guys. I logged in last night and was going to try doing some Dolphin benchmarks myself but I don't think my ancient PC is really up to the task of running some of these games in order to get reliable figures, when I tried F-Zero I was getting less than 30 fps in most cases! :lol:
Looking forward to seeing more benchmarks for both Gamecube and PS2.
Well, the newer versions of PCSX2 are really good, it's just that's the last version to give polygon counts to my knowledge.Quote:
Originally Posted by Barone
EDIT:
Also here's some Rogue Squadron 2 shots:
http://i102.photobucket.com/albums/m...psd03e8839.png
This is from the Lucas Arts intro. The game runs at 60fps so this is 7,339,560 polygons per second.
http://i102.photobucket.com/albums/m...ps816b8c04.png
This is standing in the Yavin Hangar with only the X-Wings and Y-Wings available. Again at 60fps this is 12,203,100 polygons per second. I can only imagine what it would be with the Millennium Falcon and all the other unlockable ships visible in the background.
In game it was very buggy. The Death star wasn't even visible so I kept crashing. However it was still giving what appeared to be accurate polygon per frame counts in the range of 85k-110k per frame. So that first level is pushing over 6 million polygons per second. And that's not even one of the more impressive levels. So yeah, I think ABF is redeemed on this one. Rogue Squadron 2 is a polygon pushing beast.
I am getting the impression that every game on this planet runs at 60 FPS.
If Burnout runs at 60 FPS, then why doesn't Burnout 3? They decided its not necessary for sequels?
And it'd probably have texture mapping like in GTA III.Quote:
Originally Posted by TrekkiesUnite118
Great job guys. Pretty much all I wanted this thread to produce was actual figures from actual games that actual gamers can reproduce on their own computers. I expected some of you to do it out of school yard quality immaturity, but I didn't expect others to lose their personal integrity and honesty in the process. I do have to say seeing the form of this presentation is pretty pathetic.
I really don't care if I were made the fool in all of this, which would require my quotable words to actually say what some have claimed. I already knew some games surpassed 1 million polygons per second that generation. I already posted about those games too. I also not once denied Crazyace's assertion that the Sony 2003 doc's figures were averages rather than peaks. After I asked a few questions, I actually posted the author's own defense of those figures as averages not peaks. None of that matters though, I know, because of some twisted version of objectivity. But by all means continue to prove "me" wrong by posting more games that beat "me." Did any of you bother to post the process so others can verify?
I honestly do not intend to read some of your posts any longer but would love to perform the tests on my own. You aren't afraid of that are you?
I posted the process numerous times but here you go:
1) Load game in emulator.
2) Enable debug info.
3) Take note of displayed polygon counts.
4) ????
5) Profit!!!
And it's not just a few games going well beyond 1 million polygons per second, it's actually a lot of them. Hell, Rogue Squadron 2 is a launch game pushing 12 million or more polygons per second.
Sorry but you made these silly claims, deal with it. Don't accuse us of trolling you just because your arguments reek of bullshit and we noticed it. There's a very simple resolution to all of this. Stop regurgitating the same old disproven arguments over and over again. Actually do what you've claimed to do and admit wholeheartedly that you were wrong and move on.
And before you decide to accuse me of having to be right I remind you that I have admitted to being wrong and eaten plenty of crow over the years when I've been proven wrong with cold hard facts.
If Burnout runs at 60 FPS, then why doesn't Burnout 3? They decided its not necessary for sequels?
BF3 runs at 60fps (on NTSC). The physics runs in lock-step with the rendering, so it's ultra important that the update has not just a suitably high frequency but also has a constant frequency
If it's running lower for you, then it could be that your DVD drive has some issues because the stream processing happened in the spare time at the end of the game/scene update. And if that was delayed or slower than expected, it could have a knock on effect.
Is that actual triangles drawn or is it poly passed to Flipper before culling happens? I suspect it's the latter, so you're effective number of triangles actually rendered is roughly half that. And also, I suspect that those ships don't have suitable LOD's because, as you say, they're unlockable ships so they might not have aggressive LOD's - so they could consist of lots of small trianglesQuote:
Originally Posted by TrekkiesUnite118
Or we could quit making excuses, and accept that it doesn't run at 60 FPS. I was using my August 2008 manufactured PS2.Quote:
Originally Posted by rusty
I swear its a trend on this forum to label every game as 60 FPS, when they don't friggen run at 60 FPS.
I'm pretty sure it's after culling. For one the same number was used for Barone's Burnout comparison and it was pretty close to the PS2 numbers. Also when I was standing a bit to the left the count was reduced by half as the X-Wing was blocking most of the other ships. When I was in a corner looking at no ships the count plummeted to under 16k. So I'm pretty sure that is an accurate count of polygons per frame after culling.Quote:
Originally Posted by rusty
Also none of those ships are unlocked secret ships. I said I could only imagine what the count would be IF the unlocked ships were there instead of empty spots.
Well, since it's *paramount* that the physics runs at full-whack on that game, I'll have to just accept you're not seeing what you think you're seeing.Quote:
Originally Posted by azonicrider
Don't just take my word for it though
I think I might have mixed up the terminology - I meant back-face culling, not frustum culling. I'm willing to bet that the triangle counts are the number of tris passed to Flipper, and back faces are then culled at the triangle-setup stage. So the number of actually rendered triangles are half of that.Quote:
Originally Posted by TrekkiesUnite118
From what I can tell, the emulator just passes the triangles to whatever vertex shader has been created and lets the graphics hardware do the same thing - culling after vertex transform.
Of course...it's still impressive to see the through-put on the vertex transform side of things, especially when you consider that Flipper is transforming three vertices per triangle. If they had used a post-transform cache, I'm sure they could have tripled the actual triangle through put on that machine.
Awesome!Quote:
Originally Posted by TrekkiesUnite118
That is really impressive for a launch title.
I think we can also put claims like this to rest (pay attention to how he left the PS2 out of his comparison, hehe):
Quote:
Originally Posted by sheath
About RE4:
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
Time to do the math:
Resident Evil 4 (2005) (Capcom) (GC)
In game polygon count (by Dolphin 4.0): 204103 polygons per frame
About the scene: One of the highest polygon counts I could find at the very beginning of the game.
Frame rate of the game: 30 fps.
Result: 30 x 204103 = 6,123,090 polygons per second ~ 6.1 million polygons per second.
A bit further in the gameplay:
Sreeenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
In game polygon count (by Dolphin 4.0): 264194 polygons per frame
About the scene: One of the highest polygon counts I could find after playing the game for like 5-10 minutes.
Frame rate of the game: 30 fps.
Result: 30 x 264194 = 7,925,820 polygons per second ~ 7.9 million polygons per second.
Now on the fugly PS2 version (the poor colors hurt my eyes):
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
Time to do the math:
Resident Evil 4 (2005) (Capcom) (PS2)
In game polygon count (by PCSX2 0.9.6): 23104 polygons per frame
About the scene: One of the highest polygon counts I could find at the very beginning of the game.
Frame rate of the game: 30 fps.
Result: 30 x 23104 = 693,120 polygons per second ~ 0.69 million polygons per second.
Yeah, I'm not joking, those are the actual numbers. The PS2 version seems to be doing like 9 ou 10 times LESS polygons per frame all the time.
According to what I could see running the game on Dolphin, it seems that the trees (especially the branches) are boosting the GC numbers really high. IDK, it seems to be using something like that foliage technique of Transformers (PS2) to draw a much more detailed forest than the PS2 version (the screenshots don't do justice to what happen in th games, when you are running them you can see that the PS2's is a VERY watered down version - well, at least that's my opinion).
Some other tests:
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
Time to do the math:
Ridge Racer V (2000) (Namco) (PS2)
In game polygon count (by PCSX2 0.9.6): 34449 polygons per frame
About the scene: Just the very beginning of the game.
Frame rate of the game: 60 fps.
Result: 60 x 34449 = 2,066,940 polygons per second ~ 2.1 million polygons per second.
Right after the start:
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
In game polygon count (by PCSX2 0.9.6): 38980 polygons per frame
About the scene: Just a few meters after the starting line.
Frame rate of the game: 60 fps.
Result: 60 x 38980 = 2,338,800 polygons per second ~ 2.3 million polygons per second.
Yeah, this game is very buggy in that version of the emulator but the polygons of the cars seem to be rendered despite the lack of textures.
These numbers seem to make sense with this old info by IGN:
"By comparison, developers have told us that Namco's Ridge Racer V for PlayStation2 is realistically pushing approximately 35,000 polygons at 60 frames per second, or just over two million polygons per second. "
http://www.ign.com/articles/2000/07/...swers-you-want
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
Time to do the math:
Tekken Tag Tournament (2000) (Namco) (PS2)
In game polygon count (by PCSX2 0.9.6): 27531 polygons per frame
About the scene: Regular moment of the game.
Frame rate of the game: 60 fps.
Result: 60 x 27531 = 1,651,860 polygons per second ~ 1.7 million polygons per second.
Another stage:
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
In game polygon count (by PCSX2 0.9.6): 33231 polygons per frame
About the scene: Regular moment of the game in a more detailed stage.
Frame rate of the game: 60 fps.
Result: 60 x 33231 = 1,993,860 polygons per second ~ 2.0 million polygons per second.
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
Time to do the math:
Dynasty Warriors 2 (2000) (Omega Force) (PS2)
In game polygon count (by PCSX2 0.9.6): 42615 polygons per frame
About the scene: Regular moment of the game when you face a group of enemies.
Frame rate of the game: 60 fps.
Result: 60 x 42615 = 2,556,900 polygons per second ~ 2.6 million polygons per second.
With enemies and a mounted horse (seems to use a lot of polygons) in the background:
Screenshot: https://drive.google.com/file/d/0B0c...it?usp=sharing
In game polygon count (by PCSX2 0.9.6): 47716 polygons per frame
About the scene: Still in the very beginning of the game.
Frame rate of the game: 60 fps.
Result: 60 x 47716 = 2,862,960 polygons per second ~ 2.9 million polygons per second.
Launch PS2 "some games" pushing more than 1 million polygons per second???
http://31.media.tumblr.com/tumblr_ly...wo1_r1_500.gif
Integrity and honesty aren't good words for you to use here given that no one in this thread lied, false accused, offended and ignored facts like you.Quote:
Originally Posted by sheath
And this passive-aggressive attitude will not help you either.
Again, blaming the others will not help your predictions and biased statements to sound any less absurd.Quote:
Originally Posted by sheath
Please, stop ridiculing yourself.Quote:
Originally Posted by sheath
You awfully stretched the meaning of those numbers to make them fit in your preconceived shit...Quote:
Originally Posted by sheath
This is you some weeks ago:
Quote:
Originally Posted by sheath
You've said a lot of shit and after being proved wrong you acted like a spoiled kid who had his lollypop stolen. You've had many chances to admit the facts and stop rejecting sources but instead you chose to never get down off your high delusional horse while offending people for no good reason.Quote:
Originally Posted by sheath
Deal with it.
lolQuote:
Originally Posted by sheath
After years and years tossing around numbers based on distorted interpretation of selected sources you now want to put the burden of your laziness on us?
Gimme a break.
Like I said before, if you had spent 1/10 of the time that you've dedicated to your Sony rant and conspiracy theories in actually trying to discover new facts and sources you wouldn't be in this position now.
The problem is that your bias plagued both your research and your motivation; so all you can do now is to accuse and offend us for doing what you should have done prior triggering your infinite-loop of radical statements and deluded predictions.
The games being discussed here are well known and there frame rates are stated in several different sources. Feel free to point yourQuote:
Originally Posted by azonicrider
counterproofsources which claim otherwise.
And please, let us know where people are labeling "every game as 60 FPS, when they don't friggen run at 60 FPS."
I think that is the count after back-face culling though. If I go into first person camera looking at the same area the polygon count goes down significantly as portions of the ships are no longer visible. Remember, that part of the game is very limited. Only the pilot moves, everything else is stationary, and looking at the models it's clear that they are very detailed. I wouldn't be surprised if those models weren't the ones used in game, or if they may only be swapped in for cinematic or parts where a ship is very close to the screen. In game the polygon counts were 50-75% of what I was getting in the hangar.Quote:
Originally Posted by rusty
I wonder how this discussion would differ if the DC had kept its original specs? 32mb of RAM alone would have helped it ten fold I would think. Why the hell SEGA DOWNGRADED the system instead of making it as good as possible.... just another stupid SEGA decision I guess. I still think that 1998 was waaaaay to early to real ease a new system anyway. Maybe if it had been late 99/early 2000 in Japan with most importantly, a DVD-ROM drive, it could have done better. The Saturn could have survived in Japan until then I think.
That's wrong . The original spec for Dural had just 8Mb of Ram . Anway even with 32Mb of RAM the NA@MI couldn't outperform the PS2 for polygonsQuote:
Originally Posted by SegaSonic91
NAOMI 2 was more inline with the PS2/GC/Xbox though wasn't it?
It could also be that the ships are segmented and then frustum culled, which is common for models with large amount of polys. Especially for things like the x-wing, that would have moving parts. And if you want to use more than one texture (which is likely on the GC), you'd split the model up into sub-models.Quote:
Originally Posted by TrekkiesUnite118
I looked at the source of Dolphin-Emu, specifically the rendering back-end. To my knowledge, D3D and OpenGL don't have a method to count the number of triangles that have actually been rendered after backface culling - even the occlusion query extensions just tell you if the depth buffer was modified or not. And looking at the source of the emulator vertex manager, it just renders a vertex buffer and increments the number of triangles and the number of primitive calls from there.
So what you could be seeing with the different camera, just various sub-parts of the ships being frustum culled.
Cause Bernie wanted a new console for Sega!!Quote:
Originally Posted by SegaSonic91
If they had a smarter CEO, then they could've got things done properly.
1. Get the Saturn on track. as in:
1. drop the manufacturing costs. Remove the cart slot and have the RAM upgrade built in, get rid of the battery RAM, get rid of other not-totally-necessary stuff.
2. apologize to retailers that weren't included in the surprise launch. Pretty easy, don't need to give them BJs or anything.
3. work day and night to develop a graphics library that makes the Saturn easier to program for.
Remember the PS3 and Saturn had the exact same problems, and looking at the PS3 now, its safe to say the Saturn could've been repaired.
THEN they could've launched the Dreamcast in late '99 in Japan, with a DVD drive and more RAM.
Well looking at Trekkie's link, its safe to say the Naomi 2 could blow the doors off the PS2 in terms of polygons.Quote:
Originally Posted by Team Andromeda
Wait never mind Trekkie didn't post a link, I found it on Google: http://www.system16.com/hardware.php?id=725
Well such a system might have been able to "blow the doors off" the PS2 as you put it, it would however also have been prohibitively expensive to manufacture as a game console. Naomi 2 had 2 SH4s, 2 PVR2DCs and a dedicated TnL chip (Elan) and 4 times the amount of RAM that Dreamcast had. After the expensive failure of the Saturn, producing such a system would of been financial suicide for Sega. Even going with the standard Dreamcast spec and adding additional memory and a DVD drive would have driven costs up considerably, something that Sega would have struggled to absorb. The only thing that wasn't right with the Dreamcast was Sega's poor financial position at the time.Quote:
Originally Posted by azonicrider
The number of sh4's isn't really an issue. What would have been really expensive, was providing some sort of bus or local store memory to allow them to touch memory at the same time, or in a fast enough manner. And as for the dual PVR's with TNL, the big blocker would have been fast enough memory and improved pixel parallelism.Quote:
Originally Posted by stu
But with that said, all of those would have been off-the-shelf parts and if they could put them on a SOC, the cost could have been fairly low.
But that's a coulda, woulda shoulda argument. I wouldn't have minded working on something like that though.
It definitely would of been a powerful system, no doubt about it. It would just seem to be counter to everything Sega was working towards with the Dreamcast imo (clean hardware design, straightforward development system etc, competitive retail pricepoint). If they had been able to combine all those chips on to one SOC that surely would have been one BIG chip imo (and probably expensive for the 1999/2000 timeline I would think since SOCs weren't that common back then). I guess they could of made 2 smaller chips with essentially a dual-core SH4 and dual core CLX chip with on board T&L as well, but with the DVD, memory and other components I would of thought it would of pushed the price up beyond that of the Dreamcast's. I could be wrong though. :)Quote:
Originally Posted by rusty
Yep, it is just too bad you are too busy being "right" that you cannot see the flaws in your own argument. What exactly is a polygon or vector according to these emulators? Can we compare one emulator's output to another? Since they are all undoubtedly high level emulators I already know they cannot produce absolutely infallible numbers for the original hardware. Does anybody know how far off they are? Does anybody know whether any version of any 3D system emulator actually has accomplished the feat of showing the world what each game is showing polygon wise, and texture wise, per second or per frame?Quote:
Originally Posted by TrekkiesUnite118
I will absolutely enjoy the day that this happens. Odds are though, emulator authors will stop when the games work "well enough" for their target audience. Their target audience will_not care whether the source material is accurately represented for the original SD output in this case. I sincerely hope that these authors come up with a way to compare more than what we can on 16-bit consoles over emulators one day.
Claiming that loading emulators and looking at the outputs for an absolute figure on what the systems, or games, were actually producing is the worst disservice to this group and this thread that I have seen yet.
I think in terms of polys being sent to the GPU, an emulator can be pretty spot on. Since were MOSTLY dealing with triangles, it's a good measurement of what a game is pushing through the graphics pipeline. However, that's where it ends. In determining the actual number of triangles that are drawn, all an emulator can do is provide us with a baseline figure to do our estimates, which is why there's this disagreement of the numbers being actual polys rendered or number of polys submitted before culling and rendering happens.Quote:
Originally Posted by sheath
After looking at the emulators source, I'm pretty certain that it's the numbers of polys submitted before any guard band and back-face culling occurs, so the number of polys actually drawn is around half of that number, perhaps less.
If you want anything more fine grain than that then an emulator isn't going to help. I think one of the issues here is that Trekkie wants a certain result, and won't look or lacks the knowledge to be able to look any deeper to verify it. Which is ok, because somebody else can always do that digging and propose a counter-argument.
NA@MI also had 32Mb of main ramQuote:
Originally Posted by TrekkiesUnite118
You can't put all the blame on Bernie , because most of the damage was done already before he took over . Also you points don't all work imoQuote:
Cause Bernie wanted a new console for Sega!!
If they had a smarter CEO, then they could've got things done properly.
1. Get the Saturn on track. as in:
1. drop the manufacturing costs. Remove the cart slot and have the RAM upgrade built in, get rid of the battery RAM, get rid of other not-totally-necessary stuff.
2. apologize to retailers that weren't included in the surprise launch. Pretty easy, don't need to give them BJs or anything.
3. work day and night to develop a graphics library that makes the Saturn easier to program for.
1) SEGA already did that and with the Model 2 Saturn dropped the price . Also one couldn't really get rid of the Cart slot as that would have added to the cost of the Saturn and not take away from it .
2). I doubt that would have made any really difference myself
3). SEGA did with the SGL 1 and then the SGL 2 .
Why do people keep saying this ? The development of the Saturn lost SEGA not a penny . SEGA America and Europe blowing away SEGA marketshare it was cost it millions .Quote:
After the expensive failure of the Saturn, producing such a system would of been financial suicide for Sega.
I think it depends how you look at it. People blame the Saturn itself for the failure, when obviously that's not entirely true. Had the system been more in line with the PSX (strong 3d capabilities, easier to program for) it certainly may have been more appealing... but the hardware itself was more of an "and also" issue. Sega's issues and behavior is what caused the problems.. The Sega CD, for instance, wouldn't be seen in such a bad light if it wasn't for the 32X and Saturn that followed, making it a sting of "bad ideas".Quote:
Originally Posted by Team Andromeda
There is no reason to think things wouldn't have gone the same way if the PSX and Saturn hardware was reversed.
Liked that helped the XBox, DC, Cube against the PS2 ? . Fact of the matter is being easier to program for was an excuse used by developers - Its market share that decides support and the Saturn had strong 3D graphics when used right , you just had to use the system.Quote:
Had the system been more in line with the PSX (strong 3d capabilities, easier to program for)
What cost the Saturn was the lack of any real Sonic game on the system (nice one SOJ) and SOA/SOE focusing on the 32X at gifting SONY the 32 bit battle in the west imo
You appear to be missing my point, its not so much the money that Sega spent on developing the Saturn itself that cost Sega, but the design decisions that they made while designing the system, the total breakdown in communications within Sega and it's subsequent marketing of the system and the cumulative costs of all these mistakes on Sega's profit margin.Quote:
Originally Posted by Team Andromeda
Examples are the choice to go with the complex and convoluted hardware design that made the system more expensive to manufacture throughout the Saturn's time on the market and also harder for developers to exploit effectively and the costly breakdown in unity between the Sega divisions that meant SoJ were making design decisions without fully consulting SOA and SOE, the MD/Genesis was making Sega a buttload of money and SOJ should of been more conscious of keeping that going with the Saturn (sorry imo 32X was not the answer - but thats for another discussion) . Every price cut of Sony's that Sega matched was far more costly for Sega to implement than for Sony and that's not just because Sony "had mountains of cash" to piss away on the PS1 like some people claim, it was also because the PS1 design was more integrated and thus cheaper to manufacture. One rumored cost estimate that I read in an issue of EGM (iirc) for the Saturn was in the region of $370, that means they were losing at least $70 per system at $299, I'm sure that made the prospect of dropping the price further pretty unpalatable to Sega.
To sum up, if Sega had repeated some of those same design mistakes with Dreamcast (eg a system like Naomi 2 manufactured as a console) would of been suicide for them imo.
Quote:
Originally Posted by EclecticGroove
I'm not really so sure on that last point, I think Sega would have been in a vastly better position to combat Sony had they gone with a cleaner design like the Dreamcast (not the exact specs as the DC obviously but the cleaner approach/better development system) Sure Sony did have the tons of cash to back the system and they had hired a lot of people away from both Sega and Nintendo, but having "tons of cash" didn't exactly help Panasonic or Phillips with either of their console endeavors. I think if Sony had gone with a Saturn-like design for PS1 (and poorer development tools as well) it would of made the job of convincing 3rd parties to support the system harder (especially since this would have been Sony's 1st foray in to building a console with no previous success) and the machine more expensive for Sony to drop in price so easily. I'm not saying that it would of reversed Sega's fortunes completely but I find it hard to see how it could of been as bad as it actually was.
Well since SEGA needs to buy-in off the shelf parts for most its Hardware - It was always going to be slightly more expensive than a SONY machine which could produce a lot of its compainits it's self . And Saturn was only more expensive at the start , its quicky was able to drop its price , so Price wasn't so much a factor. In Japan it was SEGA that was taking the lead on price cutsQuote:
Examples are the choice to go with the complex and convoluted hardware design that made the system more expensive to manufacture throughout the Saturn's time on the market and also harder for developers to exploit effectively and the costly breakdown in unity between the Sega divisions that meant SoJ were making design decisions without fully consulting SOA and SOE
Well I remember a interview with SOJ saying that the Saturn (like the Mega CD) was being sold at cost price and that's why it had a high price . And like with SONY the price of CD-Roms and RAM would fall , allowing ever more price cutsQuote:
One rumored cost estimate that I read in an issue of EGM (iirc) for the Saturn was in the region of $370, that means they were losing at least $70 per system at $299
That's a bit of myth . The Saturn wasn't that hard to developer really , I've seen interviews with Treasure, Warp, even RARE who said the N64 was the hardest machine to code for back in those days ; It was just more of the case that PS developers couldn't share the C code over and that cost many Saturn ports . To use the Saturn you had to use assembly . Also lets stop the myth that SOA were kept in the dark .Quote:
Saturn's time on the market and also harder for developers to exploit effectively and the costly breakdown in unity between the Sega divisions that meant SoJ were making design decisions without fully consulting SOA and SOE
SOA/SOE knew fell well about the Saturn and SOJ Japan plans for the system (just like most of the media ( - after all SOA were the 1st to show off the Saturn and its games to the world, SOA set up the Away Team division in 1993 . SOA/SOE just felt that the 32X was the better call, and it blew up in their faces. That was the trouble SOA/SOE felt they could have their cake and eat it with the Mega Drive , 32X and Saturn .
I'm not really sure that is is a myth if the head of Sega's most storied arcade development teams criticized the design of the system. Yu Suzuki was quoted in Edge as saying "One very fast central processor would be preferable. I don't think all programmers have the ability to program two CPUs—most can only get about one-and-a-half times the speed you can get from one SH-2. I think that only 1 in 100 programmers are good enough to get this kind of speed [nearly double] out of the Saturn" In a recent interview with Eurogamer.net Ezra Dreisbach (formerly of Lobotomy Software - developer of Saturn games Powerslave, Duke Nukem 3D and Quake) had this to say about the Saturn It also addresses your points about the Saturn not being as expensive to manufactureQuote:
Originally Posted by Team Andromeda
Quote:
Eurogamer: Saturn versus PlayStation, let's put this to rest once and for all. You helped oversee two of the most technically accomplished games on those systems, so lay it out for us, which machine did you prefer from a technical perspective and why? What were the most rewarding and annoying aspects of both systems?
Ezra Dreisbach: I didn't exactly oversee anything. Jeff Blazier programmed the entirety of Lobotomy's PlayStation output. But it's not hard to tell the PlayStation is better.
The Saturn was really an insane abortion. The graphics hardware was made by guys that obviously wanted to just keep developing 2D hardware and tried to avoid learning anything about 3D. So they made this thing that was totally different from what everyone else in the 3D community was doing and missed some real key ideas, making some things (clipping) impossible.
And then the rest of the system had a whole other batch of warts caused (according to the internet) by a hasty pre-launch upgrade to match the PlayStation. They threw a whole bunch more parts in the box, and none of them worked out that great. The second processor in particular made it both more difficult to program, and impossible to fully utilise.
This probably ultimately doomed the Saturn. With so much different crap jammed in the box, it never got cheap for them to produce it.
I'm not about to post the entire interview right here in this thread but there are a number of interesting things he had to say regarding the Saturn and its faults, I guess its easier to speak freely once there is no Sega PR person sitting next to you.
The full interview can be accessed here, for those who would like to read it:> http://www.eurogamer.net/articles/de...bach-interview
BTW before anyone can start claiming that the interviewer was "biased against the Saturn" I would like to just point out that the interviewer was Richard Leadbetter who I beleive was formerly the editor of both the Official Sega Saturn Magazine and the Office Dreamcast Magazine in the UK.
I'm not so sure that it right either, I doubt SOA or SOE got to have a say in what was going on with the Saturn, I seem to recall that SOA told SOJ that the Saturn was going to be too expensive for the US market and SOJ essentially fobbed them off with the 32X design as a way of somehow negating that (instead of, you know, actually working to make the Saturn cheaper to manufacture which I would of thought would be classed as a "Good Idea".Quote:
Originally Posted by Team Andromeda
Well we can all do selective quoting . I seem to remember the head of Square Soft saying that no more than 13 developers will ever be able fully use the PS2 . Also Yu Suzuki knows full well about using multi CPU's ; I mean System X, Y boards had 3 CPU's for starters and Model 1 had 4 DSP, another one for the backgrounds and a NEC CPU . Also you should trying quoting what Yu Suzuki had to say over the Saturn in March 1995 when showing off AM#2 new Graphics library :)Quote:
Yu Suzuki was quoted in Edge as saying "One very fast central processor would be preferable. I don't think all programmers have the ability to program two CPUs—most can only get about one-and-a-half times the speed you can get from one SH-2. I think that only 1 in 100 programmers are good enough to get this kind of speed [nearly double] out of the Saturn"
I've seen it, But then again I could post interviews with Core (who all said the Saturn was more powerful), Travelers tails, Vidi Image, Silicon Dreams, NMS who had good things to say about the Saturn.Quote:
I'm not about to post the entire interview right here in this thread but there are a number of interesting things he had to say regarding the Saturn and its faults, I guess its easier to speak freely once there is no Sega PR person sitting next to you.
Not at all in one of the 1st interviews with the SEGA Saturn mag , Lobotomy listed some of the Saturns short comings , but there again the PS also had plenty of it's own .Quote:
BTW before anyone can start claiming that the interviewer was "biased against the Saturn"
No that's no correct at all . 32X started off as SEGA answer to the threat of the Jaguar and the 3DO and SEGA America pushed on with it . SEGA Japan did look at a design for the Saturn minus the CD Drive and half the systems ram but that was quickly dropped .Quote:
I seem to recall that SOA told SOJ that the Saturn was going to be too expensive for the US market and SOJ essentially fobbed them off with the 32X design as a way of somehow negating
Yes, we certainly can, you can selectively quote to make your point and I can to make mine. :DQuote:
Originally Posted by Team Andromeda
I'm sure that you can post all these positive interviews, just as I can find more negative ones if I so wished. I was merely pointing out to you that there were developers that disliked the Saturn and it had a reputation for being a difficult system to work with, it counters your claim that the reputation was "a myth" and therefore untrue, you can post as many accounts of the Saturn as being "rosy and wonderful" as you like but it doesn't change the fact that some developers disliked working on the system. Both the PS2 and the Saturn had a reputation of being somwhat of a "marmite" type system with regard to developers, they either loved it or loathed it.Quote:
Originally Posted by Team Andromeda
I wasn't aiming that you per se, just as a general disclaimer to everyone reading my post.Quote:
Originally Posted by Team Andromeda
Thats kinda funny with regard to the 3DO as Tom Kalinske went on record in an interview with MEGA magazine in the UK circa May 1993 that they were unconcerned about the threat of the 3DO and that they (Sega) felt that the combo of the MegaDrive and MegaCD represented "better value for money" since it offered a lot of the same functionality as the 3DO while being cheaper. As for Atari I can't really see how anyone could see them as a threat considering the desparately poor financial condition they were in circa 1993/94.Quote:
Originally Posted by Team Andromeda
Well like I said the N64, PS2, PS3 were seen as hard to work with it never held them back from impressive sales and all those systems.Quote:
I'm sure that you can post all these positive interviews, just as I can find more negative ones if I so wished. I was merely pointing out to you that there were developers that disliked the Saturn and it had a reputation for being a difficult system to work with
I'm not saying the Saturn was perfect, in fact far from it.. I know it couldn't handle the amount of polygons to that of the PS and the PS had better lighting and transparent effects for polygons , the launch tools were poor . All I will say its that's its a myth to make out the PS won the day thanks to being easier to program for . If that were the case the DC, Cube and XBox would be the best selling consoles of all time, the fact that all those machines were single handedly outsold by a single console (The PS2) and one that was meant to be one of the harder consoles to program for, just shows its a myth to make out the Saturn lost out thanks to being harder to program for .Quote:
counters your claim that the reputation was "a myth" and therefore untrue, you can post as many accounts of the Saturn as being "rosy and wonderful"
No worries, but in one of the 1st interviews with Lobotomy they said they had some issues with the Hardware .Quote:
I wasn't aiming that you per se, just as a general disclaimer to everyone reading my post
Look I've already proved Tom is bare face liar and will say anything to make him look good . You talk about Mega ? I remember Tom saying that SEGA could release a 32 Bit console tomorrow if they wanted to (in another dig at the 3DO) . So he knew full well of the Saturn plans, very early in . Fact are you don't bash or find faults with other corps you don''t see as a thread and any machine that was getting some serious backing from EA, Panasonic and Trip would get all the top CEO looking and asking questions at their R&D labs .Quote:
Thats kinda funny with regard to the 3DO as Tom Kalinske went on record in an interview with MEGA magazine in the UK circa May 1993
I mean if all listen to Tom the Saturn and PS were not a mass market price point and the 32X was the only mass Market 32 Bit system and would outsell the PS and Saturn . Yeah... nice one Tom
Mega CD is 1991 Tech and was made to counter the PC Eng CD-Rom and Super Famicom. Also the Mega CD is noted as being a bitch to work on and a difficulty system to get the best out of. Its also a system that liked the Saturn, used 2 CPU's , the MD GPU and then the Mega CD ASIC chip to make it sing . So again like I said , SEGA in both the Arcades and home used multi CPU and GPU's to get power . Lets not also forget the Mega CD cost more than the Saturn on its releaseQuote:
felt that the combo of the MegaDrive and MegaCD represented "better value for money" since it offered a lot of the same functionality as the 3DO while being cheaper
It was new Hardware vastly more powerful than the Mega Drive and coming in a rather nice price point . Also you need to remember that SEGA had to pay Atari Ł44 million and buy a ton of shares in the corp and offer Atari some of its IP . So while not a major threat Atari were seen as threat none the lessQuote:
As for Atari I can't really see how anyone could see them as a threat considering the desparately poor financial condition they were in circa 1993/94.
I really don't think N64 can be placed in the same group as those from an "impressive sales" perspective. Both the N64 and the Saturn were outsold by the PS1 by a huge margin. Yes I agree that the N64 is another system that has been criticized as being a pain to work on and like the Saturn it struggled in that gen. I really can't say that it was successful though, more successful than Saturn yes but in the same league as PS2 or PS3? :oQuote:
Originally Posted by Team Andromeda
I don't believe I ever made the case that the SOLE reason the PS1 beat the Saturn was due to the Saturn "being harder to program for". It was way more complicated than that and I covered those other reasons above in a previous post. But I would list that reason as one of the historical reasons for failure to garner the same level of support as the PS1 was due to certain developer's difficulties with the system and Sega's inability to solve those issues relatively quickly.Quote:
Originally Posted by Team Andromeda
But you see, you claim that he's a liar because he says in one part of the interview that Sega is able to counter the 3DO with the MegaDrive/MegaCD combo, but then when he says that Sega is just as able to put together a 32 bit console you seemingly believe him and point to that as being proof he knew about Sega's Saturn plans? He may of just said that to allay any concerns that Sega fans might have had that Sega had no 32 bit plans - defensive PR or whatever. How do we know he had knowledge of the Saturn? Is he just a selective liar or a total liar, most liars will lie all the time not selectively.Quote:
Originally Posted by Team Andromeda
Isn't this one of the most wildly misquoted quotes about the Saturn's abilities? If I remember correctly when this was brought up before it was pointed out that he was referring to some specific thing with Virtua Fighter's development that even the PS1 would struggle with.Quote:
Originally Posted by stu
It's really funny how this guy has changed his tune after his company died. Back when he was actually developing for the system he had nothing but praise to give the system as well as calling out other developers who called the system weak at 3D.Quote:
Originally Posted by stu