PS3 sales

[Shifty Geezer]

...and insufficient performance is waving back at you.

Though that's an argument, it's one that to date doesn't appear to be a differentiating factor, and won't be until PS3 shows that extra performance is applied in game-differntiating ways. From Joshua's example, it's likely the cited game isn't doing anything on PS3 that XB360 hasn't the potential to do itself, so citing that game as an example of where performance is a differentiating factor. That is, from One's inital statement, 'if Motorstorm could be realized on other platforms', chances are very high that Motorstrom could be realized on other platforms a it not doing anything technically beyond what's expected from XB360. The big difference here is that Ps3 is doing that now, and the software on XB360 isn't, but that seems to be more the software choices made. Future games are set to show similar terrain deformation, for example.

I'm unsure how One's later explanation ties in. If the difference between these technologies appearing on different platforms is all of six months, I doubt that'll make for substantial total platform sales on account of one console showing tech the other hasn't got yet. A console is bought for multiple reasons, not just cutting edge tech in its current lineup. Very few people will buy a PS3 just on the grounds one game shows better racing physics than that shown in any XB360 games.

 

[Nesh]

" We have built up a certain brand equity over time since the launch of PlayStation in 1995 and PS2 in 2000 that the first five million are going to buy it, whatever it is, even if it didn't have games."~Sony Computer Entertainment Europe CEO David Reeves

http://www.1up.com/do/newsStory?cId=3150935

Perhaps its too early

There is still some truth to it though. And notice the title. "if we make it, thwy will buy it." They didnt make it exactly (See shortages)

He was pointing out there that they have such a huge Playstation base that sales could easilly reach sales of millions. I doubt 5 million. But I am sure they could have sold some millions under the right circumstances. And it was a good bet back then

Ofcourse they havent projected the production issues. That was a major f*ck up. Also back then they couldnt see how GoW would have affected their competitor. Oh they forgot about Wii too. Both MS and Sony ignored Nintendo.

PS3 didnt come with the problems PS2 came. PS2 had DVD playability issues, many numbers of PS2 consoles suffered from overheating problems and disk read errors, and memory cards had the tendency of erasing theirselves. It was also a HUGE headache for devs to make games on. Despite that it was selling extremely fast for a very very long period.

PS3 major issues was the price and the extreme shortages. It also launched directly with another diversified console.

When the PS3 was available in extreme small numbers during launch, it unavoidably sold too few consoles, unlike PS2 which despite the shortages it had enough available to reach millions in a matter of weeks. Also because of this an even greater ratio of total consoles available make it to ebay at inflated prices. When PS3 is not available, when the distribution doesnt seem right and when there is a super cheap new console available during Xmass with games, ofcourse people will choose the alternative.

All these started to give more weight to the price and its impact on consumers.

I am sure that if Sony had enough PS3's shipped worldwide during launch, (and perhaps if Wii was released later), they would have made another record despite the price. If I am not mistaked PS2 surpassed PS1 sales during the same duration of their "life". Easilly 5% of more than 100 million PS2 users should want a PS3 dreadfully. Just as millions of PS1 owners wanted a PS2 dreadfully back then.

But in general this is the most important part and the others are mere details:

There are many of those who would rush in the first days to buy a console lead only by excitement.

The more consoles are available the more of these people will get it before they notice issues of a newly launched console. So these issues are outshadowed by the huge sales. Having huge availablility during launch is probably the most important strategy to make a great start.

Sony lost this very valuable strategy, PS3 was left with a smaller userbase than natural, and those that would have otherwise bought it started noticing things like price, games and features which they would have ignored.

When sales are huge people dont care as much about these. They see the word "success" next to the console notice the launch demand and they are driven by future expectations to get it now. Now people are waiting instead to get it later

This is what driven PS2 to a success and of a continuous high demand despite its problems.

So I dont think his assumption back then was completely illogical..

 

[Jesus2006]

Very few people will buy a PS3 just on the grounds one game shows better racing physics than that shown in any XB360 games.

well, that's exactly the reason why i want one so bad. I don't want to wait until the 360 shows similar effects (if possible at all), because until then im sure the PS3 has gone at least one step further.

And these are launch games - never forget!

 

[one]

I'm unsure how One's later explanation ties in. If the difference between these technologies appearing on different platforms is all of six months, I doubt that'll make for substantial total platform sales on account of one console showing tech the other hasn't got yet. A console is bought for multiple reasons, not just cutting edge tech in its current lineup. Very few people will buy a PS3 just on the grounds one game shows better racing physics than that shown in any XB360 games.

For that matter I assume Eye-toy tech leveraged by Cell has more potential. It can be applicable to non-game softwares too. These tech widen the variety of games.

Anyway. It's too early to write off Motorstorm only by screenshots, that's for sure
http://www.youtube.com/watch?v=4gcH4o4MP5M
http://www.youtube.com/watch?v=GBWzxcbMmqo

 

[Shifty Geezer]

For that matter I assume Eye-toy tech leveraged by Cell has more potential. It can be applicable to non-game softwares too.

EyeToy tech may be a strong-point, as PS3 has a SPE or 2 to spare on image processing which XB360's camera can't benefit from. However, actual developments in that capacity are few and far between. We've seen nor heard zip about Eye of Judgement for ages, despite a looming launch date. And what ever happened to the AR demo's from E3 '05? We saw there virtual water manipulation. How come there's nothing as advanced in development to wow the masses? If I were Sony, I'd have at least 3 big EyeToy properties in development and be sure to make noises about them to attract interest. At the moment, the best PS3 can offer is some potential, and I think people prefer to buy what they can see here and now.

 

[one]

We've seen nor heard zip about Eye of Judgement for ages, despite a looming launch date.

Frankly I think Eye of Judgment stays niche, I assume that particular game is almost a game design demo to give inspiration for third-party developers. Also the PS3 price has to be lower for these types of games.

From the marketing POV Eye-toy's biggest market is Europe where it sold million and obviously they should show something for the EU launch along with Heavenly Sword and other games.

Another thing is, the web browser. PS3 games can spawn it in-game.

 

[Mintmaster]

Everything can be done on any console. The question is just at what cost and performance. And ive yet to see any 360 game with as interactive environment as in MS.

What are you talking about, that mud effect?

First of all, I think it looks very weird visually. Mud doesn't look like that or reflect like that. Hence Tagrineth's "driving on plastic" remark.

Second of all, I've done a bumped-tracks-through-mud effect way back on my original Radeon. All you need is render targets and bump mapping. The CPU barely does a thing.


You think that CELL is going to revolutionize what can be done on a console. You're right when you compare it to PS2, but not when you compare it to XB360. Even if you can manage a 2x performance leap, that's barely enough to see a difference with physics. I've written a physics engine, I've read many papers on cloth simulation, etc. Things don't scale like you think they do. You're not as dominated by FPU calculations as those IBM benchmarks.

It's ALL about software and art. If Sony disabled 5 SPEs on PS3 and asked the Motorstorm devs to deal with it, they'd make the game just as enjoyable. The physics would run through fewer iterations and yield a less accurate solution, but you couldn't tell the difference. The graphics would be just as pleasing.

That's why ease of development is so important. You can call devs lazy for complaining about CELL and Sony's tools, but any man-hours you spend on fine-tuning for PS3 could instead be spent on making the graphics look better, adding more effects, improving gameplay, etc. There are SO many techniques that are years old and have not been used by games simply because the devs don't have the time or budget.

 

[Arwin]

Even if you can manage a 2x performance leap, that's barely enough to see a difference with physics.

And if the performance leap is 4x or more? When do you think it starts counting? Because personally, I think that while general game code may be easier to scale on the Xenon, physics code is going to scale a lot better on Cell. Whose SPEs, by the way, aren't too shabby at integer math either.

I think there are some great possibilities for combining the physics with the motion sensing, and I can't wait until someone is going to show that. Certainly, if they ever get drivers up for the sixaxis under Linux that expose the tilt and accellerometers, I'll play around with that myself. From the first work done on this and other forums, it's becoming increasingly clear to me that such things aren't as hard as they may first seem.

There's a lot of potential in Wii games for motion sensing stuff, but in combination with the sixaxis, that's going to be much more interesting. I could think of a great deal of games that you could very interestingly simulate. Of course, I don't know yet how precise the data is that you can read from the sixaxis, but certainly an SPE is going to help a great deal in detecting and translating complex movement from it. But I'd really, really like to know how precise you can measure translational movement with the sixaxis. I'm sure the tilt features will be precise enough, so there's definitely stuff you can do with that. But how many different speeds can the translational sensor detect?

 

[Fu3lFr3nzy]

Well, if you are in the states you should be able to get one, according to the e-mail I got from GameStop they are available in all their stores and online.

A funny story occured today somewhat related to that. My dad and I went to Sears today because he wasnted to buy a GPS system for his SUV, and I went over to the video games section looking for a PS3 section and surprisingly, then didnt have one

There was a disclaimer on the top of the glass shelf saying something about PS3's "high demand" and sony not shipping enough units, so they'll most likely get some sometime in 2007

I asked my sister's bf who works there if they have any PS3's in stock, and he said they dont, and he's not sure if they ever had some

 

[Mintmaster]

And if the performance leap is 4x or more? When do you think it starts counting? Because personally, I think that while general game code may be easier to scale on the Xenon, physics code is going to scale a lot better on Cell. Whose SPEs, by the way, aren't too shabby at integer math either.

I don't know because it's a grey area, but I think somewhere between 5-10x is where it notably impacts game quality for more than a handful of scenarios. When you talk about "scaling" physics code, what in god's name are you talking about? I mentioned scaling from PS2 to PS3, which will be very big. However, the phrase "scale a lot better on Cell" just doesn't make sense. Define it for me. And "integer math" is not the problem either. It's all the other stuff like data flow, branches, in-order execution, stalls, etc. I know Xenon suffers the same problems, but the point is that FPU prowess in simple streaming applications is very overrated even for physics.

I think there are some great possibilities for combining the physics with the motion sensing, and I can't wait until someone is going to show that.

You don't need massive computation for that. These possibilities are no greater on PS3 than on the Wii, and in fact the Wii probably offers more opportunities here simply due to the way the controller is set up. Of course Cell can handle a more complicated physics setting in a game, but that's true whether there's a motion sensing controller or not.

but certainly an SPE is going to help a great deal in detecting and translating complex movement from it.

Not really. We're not compute bound in gesture recognition, we're limited by what people have done with software.

I've been saying this again and again. Software and art is key. The variation in what different devs are able to do with the same hardware is so much bigger than speed differences in hardware this gen that the latter gets lost in the former. If someone shows you a new game without telling which system it's running on, there's no way you can tell whether it's PS3 or XB360. Their capabilities are just too similar.

 

[Laa-Yosh]

If someone shows you a new game without telling which system it's running on, there's no way you can tell whether it's PS3 or XB360. Their capabilities are just too similar.

This is a signature worth line, thanks!

 

[Cheezdoodles]

And if the performance leap is 4x or more? When do you think it starts counting? Because personally, I think that while general game code may be easier to scale on the Xenon, physics code is going to scale a lot better on Cell. Whose SPEs, by the way, aren't too shabby at integer math either.

The main problem with physics is that in order to get an result where the average end user would go "Wow, thats a lot better physics" you need to do at least 10 times the amount of math. (this is probably an understatement)

The Cell is a monster for doing physics, but the problem is that the physics engine must be extremely more advanced than its competitor for you to actually notice a difference, simply because you can make the calculations less accurate and you cannot really tell the difference, (meaning less calculations for everything thats happening), the problem here is that nobody here has studied how i.e. a body reacts to a pistol shot, so we cannot really tell the difference, we dont know how it should feel or look.. When calculating how a wall should blow up, none of us know how the bricks should be affected. One of them is less realistic, but since we actually dont have enough real life experience with what we are seeing we cant really tell whats better or not.

Just look at different Ragdoll technology, from the way dead bodies react in Hitman 1 to how Havoc 2\3 engines handle it, the difference is minimal, even tho the amount of processing in the havoc based engines is much much much much more taxing and complex.

Hitman being so old, we do see a difference, the bodies seem to be abit to "light" (this ofcourse could be tweaked quite easly) when you drag them around, but the amount of processing that is needed for you to tell a difference (if we are using the same input data) is incredible.

 

[jonabbey]

The main problem with physics is that in order to get an result where the average end user would go "Wow, thats a lot better physics" you need to do at least 10 times the amount of math. (this is probably an understatement)

For a fixed set of objects, you mean? Being able to do a constant level of simple physics on ten times the number of objects in a game would be immediately apparent, I would think.

 

[antlers]

For a fixed set of objects, you mean? Being able to do a constant level of simple physics on ten times the number of objects in a game would be immediately apparent, I would think.

But multiplying objects can introduce bottlenecks other than physics where the Cell might not be able to help you. Again, it'll depend on the specific software.

 

[Acert93]

Though that's an argument, it's one that to date doesn't appear to be a differentiating factor, and won't be until PS3 shows that extra performance is applied in game-differntiating ways. From Joshua's example, it's likely the cited game isn't doing anything on PS3 that XB360 hasn't the potential to do itself, so citing that game as an example of where performance is a differentiating factor. That is, from One's inital statement, 'if Motorstorm could be realized on other platforms', chances are very high that Motorstrom could be realized on other platforms a it not doing anything technically beyond what's expected from XB360.

Exactly. One made the flippant statement about Motorstorm's inability, in his opinion, to be realized on the 360. Yet the very feature he showed off in his screenshot examples is something explicitly mentioned to be coming to a future 360 game.

The big difference here is that Ps3 is doing that now, and the software on XB360 isn't, but that seems to be more the software choices made. Future games are set to show similar terrain deformation, for example.

Exactly. Again, we could go tit for tat and name many features and software implimentations in current games that we could use as such arguements. Take a launch game like Kameo, which to this day few games have done many of the neat things it does. Does it mean the 360 is no longer capable of doing these things or the PS3 cannot do them? Or how about the Lost Planet explosions?

Back to Motorstorm, besides the plastic look, the bumps are not even realtime deformed terrain. They are textures (parallax mapping), which if the R500 series and G70 series is any indicator, will be something the 360 does very well. The ruts as variable terrain is an old concept in games (see: driving in watery sand or going off the road into the dirt waaaay back on the SNES); Motorstorm does it to a much higher degree of accuracy and scale but this is a feature specifically planned for the 360 by 3rd parties (i.e. not even an exclusive).

The point on the middleware was simple: Middleware is rarely the most performant solution to show off how "powerful" a platform is. If Motorstom doesn't even require a ground-up solution for Motorstorm on the PS3 to extract the necessary performance to make all this possible, and general PS3 middleware will suffice, then I have a hard time believing that this is some great "proof" the 360 cannot realize similar results.

Besides the terrain, the other major physics driven aspect of the game is the damage model. Of course a number of 360 games have shown such technology, and one need look no further than a PS2/Xbox1 port to see that the 360 is more than sufficient to do visually impressive car collissions and deformation. I give you Burnout Revenge:

http://media.xbox360.ign.com/media/760/760830/img_3383900.html
http://media.xbox360.ign.com/media/760/760830/img_3391886.html
http://media.xbox360.ign.com/media/760/760830/img_3362138.html
http://media.xbox360.ign.com/media/760/760830/img_3327494.html
http://media.xbox360.ign.com/media/760/760830/img_3306778.html
http://media.xbox360.ign.com/media/760/760830/img_3383837.html
http://media.xbox360.ign.com/media/760/760830/img_3391893.html
http://media.xbox360.ign.com/media/760/760830/img_3383889.html
http://media.xbox360.ign.com/media/760/760830/img_3383887.html
http://media.xbox360.ign.com/media/760/760830/img_3383870.html
http://media.xbox360.ign.com/media/760/760830/img_3383904.html
http://media.xbox360.ign.com/media/760/760830/img_3330037.html
http://media.xbox360.ign.com/media/760/760830/img_3306780.html

And if the performance leap is 4x or more?

What if?

That is the problem with the post (edit: not yours Arwin, but in general), the implied assertion as fact. A multiplatform developer came over and said, "Ok guys, for our Game -- Game XYZ -- the overall performance differential is 4x on Cell" than I would have no problem accepting that as fact for his design. If he said, "In situations A, B, and C we see a 10x speed up, in D, E, and F a 50% slow down, and a net gain for 4x" again I would defer. He knows his game engine a LOT better than I do.

But if some casual forum poster is asserting that a certain technology gives an overall 4x speed up, or that a game isn't even possible on another platform by implied power, well, we should question that. And when certain techniques and effects are given as a reason a game cannot be realized on another platform -- when the other platform is more than capable of such because games already demonstrate the same general ideas and techniques -- then we should start asking questions.

I agree, there are situations where Cell is a big win in performance over Xenon, especially in SIMD situations and code that works nicely with the Local Store (size; system memory access patterns). I am not saying that the PS3 and Cell won't do some amazing stuff.

I just don't see Motorstorm as proof of this. And I really scoff at the idea at talking Motorstorm up as a game that cannot even be realized on the 360 when it is either already clearly demonstrating the same general technologies or they are already announced. Where is the proof to back up such statements? Opinions are not facts.

 

[inefficient]

You could look at a PS3 game that implements features X, Y, Z then point out that Xbox360 game a does X, game b did Y, and game c will do Z.

In general 360 should be able to match graphical features with the PS3. It's just the potential for multiple combinations of those features in frame at the same time on the 360 will be limited compared to the PS3.

Motorstorm is probably not the best example since it is a launch window game and uses launch window middleware. The Havoc engine probably still exclusively runs on N number of SPUs and no other code can run on them. A decent SPE job rescheduler like that HS guys have that keeps the idle time down would show more amazing results and let you have more effect combinations in frame at the same time.

Motorstorm is still a very impressive game for what it does do. 16 cars or whatever all bumping an crashing into each other, models deforming, pieces flying off, cars flying into obstacles and exploding and still maintaining 30 fps is good stuff.

 

[blakjedi]

Motorstorm is still a very impressive game for what it does do. 16 cars or whatever all bumping an crashing into each other, models deforming, pieces flying off, cars flying into obstacles and exploding and still maintaining 30 fps is good stuff.

Try levels 8 and 9 of Lost planet with 30 people, 10 mechs, and 30 or so AKrid of small to outrageous size along with exploding platforms, fire, smoke, plasma, missles, hundreds of cinders and debris flying at the same with no hitches, or slowdown.

on topic.. my local bestbuy has 32 60 GB PS3s in stock which they say have been there for at least three days. The clerk said price and lack of high quality games are slowing sales to a crawl. Maybe i'll pick mine up when Motorstorm and Heavenly sword drop...

 

[Tap In]

But if some casual forum poster is asserting that a certain technology gives an overall 4x speed up, or that a game isn't even possible on another platform by implied power, well, we should question that. And when certain techniques and effects are given as a reason a game cannot be realized on another platform -- when the other platform is more than capable of such because games already demonstrate the same general ideas and techniques -- then we should start asking questions.

exactly

some of the assertions in some of the threads about what the 360 will and won't be capable of compared to PS3 are ludicrous considering they are being presented as facts when they are, as you said, opinions.

I for one thank you for continuing to question those opinions with facts, Joshua.

 

[gokickrocks]

well i think ps3 is fairly available now, for those that cant get it in their area, gamestop/eb games' website is now offering them without the bundle and ships within 3 days...i just checked the availability nearby of the 60gb model and it seems as though most stores have 1-3, some none, some with 4+

 

[Inane_Dork]

Surely it's too early in the generation to determine if different physics effects cannot be done on the different consoles. It seems a little premature. Maybe we should wait for an answer to show itself.

'Course, by the time an answer can show itself, the PC will have totally passed consoles over and it won't matter. And there's probably a downside here too...