RAGE : It Deserves its own thread now!

[assurdum]

No. You optimise to downplay the BR's read speed. You also optimise to downplay the DVD's limited space. They're 2 different problems that just HAPPEN to have the same solution. There is no connection. They are NOT using the same solution to appease the fanbois.

Forgive me but where JC mentioned bluray read speed problems? I haven't read anything about that. He talking only a little bit less of memory and slowest on ps3 compromise both versions ( how cpu for 360 etc..). However we talking of a little bit. Time ago I remember he complain about compromise of two dvd on 360 version (and only for microsoft rights reserved costs for a 3rd dvd). Now he said for both platforms developing on ps3 'porting' same contents of two dvd's. Isn't it a 'porting' reason ?

 

[Betanumerical]

Even if Sony brings down the OS memory use to zero it still doesn't change the fact you have 256mb MAX of ram for the game to use up which is lower than any of the other three platforms.

Do games really use up a lot of RAM? I thought the video card used up the most.

 

[Arwin]

Look, it's really much simpler than people make it now. JC's just saying that if they can bring down storage requirements for the 360 with a minimal impact on the quality of the graphics, this has load-time advantages for the other platform(s) as well. Therefore he doesn't want to exclude the possibility that they will then use these graphics on those platforms beforehand.

 

[Richard]

However - like TimothyFarrar explained in his very eloquent post, HDD equipped unit will tend to have a significant advantage in streaming scenarios, without requiring an installation. There are realworld examples of exactly what he talks about, out in the wild already.

Yep. He did mention the higher CPU requirements though. As the MT already requires CPU for (de)compression on the fly, raising the CPU load just for the cache system might not be optimal.

Are you saying this is a bottleneck? I've heard people cite the 256MB VRAM as a handicap for the PS3 even though main ram is usable by RSX. Never really hear the reverse complaint that CELL is hampered by the split RAM pool.

Carmack is saying that, but yeah. See next.

Do games really use up a lot of RAM? I thought the video card used up the most.

MT reduces your video-ram requirements tremendously AND keeps them constant no matter if you're looking at a dimly-lit corridor or incredibly bright open vista. What this means is that the ratio between game and graphics memory requirements change quite a bit. For the XBOX which has flexibility this is great so Rage can do new stuff in game memory that other games can't. For the PS3 it's not so great, and that's why id says they always hit a memory limit first on the PS3. For the PC, it's also split but it doesn't matter because even low-end computers already have more memory than both consoles put together.

Forgive me but where JC mentioned bluray read speed problems? I haven't read anything about that. He talking only a little bit less of memory and slowest on ps3 compromise both versions ( how cpu for 360 etc..). However we talking of a little bit. Time ago I remember he complain about compromise of two dvd on 360 version (and only for microsoft rights reserved costs for a 3rd dvd). Now he said for both platforms developing on ps3 'porting' same contents of two dvd's. Isn't it a 'porting' reason ?

Here:

We think that it’s going to cram onto two discs right now. We’ve been spending a lot of effort on an extra compression method for it, and profiling where we can cut on that. We haven’t made a final call yet about whether we would actually compress things differently for the PS3, which could have a little bit higher quality. But there’s a lot of picky issues on that where the smaller we make it on disc, the faster it loads and other benefits like that. So if we get it to look decent on the 360 on two discs, we’ll probably leave it with that data set on the PS3, but we won’t make that decision final for another six months or so.

 

[JardeL]

Very early on, id hired people from Naughty Dog and few other "PS darling studios" to work on the PS specifics.

I wonder who they are and what their position was at ND ... We know that ND are working on Uncharted 2 and they also helping Santa Monica Studio for making God of War III which is one of the flagship franchises of PS brand ...

And I dont get your point about this whole issue ... It seems like you're saying " even X360's disadvantages are have positive impact on other platforms " ...

What's the point of id Tech 5 if it's not gonna use best features of specific platforms ?.. They hint " We're not gonna use advantages of Blu-Ray because X360 does not have it , instead of using advantages of Blu-Ray we'll reduce the quality to fit on DVD9" , and you defend this ?..

If they are gonna go this way so why they are wasting their time , just port it to PS3 as other devs do ...

 

[Yemeth]

I wonder who they are and what their position was at ND ... We know that ND are working on Uncharted 2 and they also helping Santa Monica Studio for making God of War III which is one of the flagship franchises of PS brand ...

I remember hearing about Jon Olick ( http://en.wikipedia.org/wiki/Jon_Olick ), seems he was an architect/lead programmer at Naughty Dog... If you listened to certain EDGE presentations, he was there too.

Cheers,

Mijo

 

[Fafalada]

As the MT already requires CPU for (de)compression on the fly, raising the CPU load just for the cache system might not be optimal.

To be fair though, these are the types of CPU load consoles are best suited for. But yea, at the moment far too little is known to even hazard a guess at what their CPU usage might be outside of rendering stuff.

I wonder who they are and what their position was at ND ...

The one that stands out would probably be Jon Olick. He's probably one of the best engineers in industry period, not just NDD. That said, he stayed with ID for just over a year before leaving, but that's a whole other topic.

 

[JPT]

That said, he stayed with ID for just over a year before leaving, but that's a whole other topic.

Made me wonder what he is doing now and from his Linkedin profile he is an "Advisory Board Member at Sony" , http://www.linkedin.com/ppl/webprof...thType=name&trk=ppro_viewmore&lnk=vw_pprofile

And now I feel like a stalker......

 

[Richard]

And I dont get your point about this whole issue ... It seems like you're saying " even X360's disadvantages are have positive impact on other platforms " ...

That wasn't my intent. I'll try to clarify below.

What's the point of id Tech 5 if it's not gonna use best features of specific platforms ?.. They hint " We're not gonna use advantages of Blu-Ray because X360 does not have it , instead of using advantages of Blu-Ray we'll reduce the quality to fit on DVD9" , and you defend this ?..

I don't defend this because I don't believe this is happening. Which option do you prefer?

a) You get higher max quality textures but they take longer to load. Since they are MTs this doesn't increase the loading time but instead you get lower-rez textures on screen for a longer duration before the high quality tiles are streamed in.

b) You get slightly lower max quality textures (probably not noticeable in the vast majority of scenes) but you also get faster streaming of high quality tiles. So the lower-rez tiles stay on screen for shorter periods of time.

Pick one. You can't have both. As you see, they are using the best features of each platform. They're raising the run-time texture quality on the PS3 because they have extra video memory they can use. They are working around the BR's slower read speed to improve texture quality. Etc.

If they hadn't mentioned the solution to both the XBOX limitation and the PS3 limitation in the same paragraph we probably wouldn't be having this conversation. In the original article it may appear as though they are _porting_ the MTs from XBOX to PS3 but if you read a little closer you'll see the same solution *may* be the optimal path for both platforms. Again, this isn't set in stone - enter the whole, they'll only decide in 6 months time.

The one that stands out would probably be Jon Olick. He's probably one of the best engineers in industry period, not just NDD. That said, he stayed with ID for just over a year before leaving, but that's a whole other topic.

Hmm? Do go on. I was under the impression he was still at id, after his Siggraph presentation on id Tech future tech from last year.

 

[Love_In_Rio]

Made me wonder what he is doing now and from his Linkedin profile he is an "Advisory Board Member at Sony" , http://www.linkedin.com/ppl/webprof...thType=name&trk=ppro_viewmore&lnk=vw_pprofile

And now I feel like a stalker......

So he is advising about hardware in PS4...mmmm... It is very logical that people that knows well how to work with Cell knows also how to improve it...

 

[catisfit]

Richard, you've skewed option b by including your opinion "probably not noticeable in the vast majority of scenes". Whether it is noticable or not is the main factor behind whether they do it or not, and will only be determined once they've finalised exactly how much compression they need to fit the game into 2x6.8Gb. Like JC said, they're putting a lot of effort into that so it's clearly not an easy task and is still ongoing, so the final quality loss is unknown at this point.

As for loading times, well I think it's a jump from JC:"the smaller we make it on disc, the faster it loads" (which is true for any media) to "blu-ray read speed problem" which is what assurdum was asking for quotes of. There are a number of factors which don't seem to have been considered here: Seek time comparisons based on the amount of surface covered by 6.8Gb of data; layer change costs; simple performance of the drives in the consoles when reading SL blu-ray media and DL DVD media. There was a thread about this on this very forum, the numbers don't need to be replicated here to drag this thread off topic. I'll try to find the link.

I would expect id to fully explore this before making a decision for Rage, and if they are serious about idT5 becoming a thirdparty engine leader, I would also expect their tools to reflect that exploration so that licensees can make similar judgements for their own games with differently sized data sets.

The comment...

So if we get it to look decent on the 360 on two discs, we’ll probably leave it with that data set on the PS3, but we won’t make that decision final for another six months or so.

...seems to sit perfectly in line with this expectation - the continuation of the thought is "If we don't get it to look decent on the 360 on two discs, we'll probably generate a specific data set on the PS3", which a lot of people seem to have missed. Either way it's an if, and like he said, there's half a year until the results of that if can be evaulated.

 

[Dargakis]

"if we get it to look decent on the 360 on two discs"
they will use the same data for the ps3 version, otherwise it would be saying the game does not look decent on the 360 on two discs.

 

[warb]

Tim Willits also mentioned slower streaming from bluray in this old interview here: http://www.shacknews.com/featuredarticle.x?id=1001

"We haven't solved all the compression issues yet. We're close to it. The PS3 streams things slower than the 360, so by the time we ship, there may be the case where getting things off the disk faster may ultimately look better. Until we are really done, and until we can get a reporter to come to id for a 'first look,' it's really difficult to predict, because DVD streaming may play a factor."

 

[Richard]

Richard, you've skewed option b by including your opinion "probably not noticeable in the vast majority of scenes". Whether it is noticable or not is the main factor behind whether they do it or not, and will only be determined once they've finalised exactly how much compression they need to fit the game into 2x6.8Gb.

That's not my opinion. That's the opinion of id Software which they articulated last year at QCon when they mentioned the XBOX version might be getting extra compression to get smaller files:

Willits went on to explain that the engine's "reachability calculation" will determine whether the game needs to load resources based on the character's location. When pressed, he admitted that some differences will be noticeable: "If you know what you're looking for, you will notice it, yes."

However, the designer was adamant that the average user will not notice the compression, and that in the end, the whole confusion amounts to a very minor graphics compromise for the sake of a convenient place to split the story.

"Practically, it's not really gonna make a big difference. We're trying to make it look awesome on all the platforms," he said, while adding that Rage "would require a lot of design rework to turn it into a three-disc game."

Like JC said, they're putting a lot of effort into that so it's clearly not an easy task and is still ongoing, so the final quality loss is unknown at this point.

The PS3 decision is definitely not easy, nor did I ever make it out to be. It's simply a trade-off. The XBOX decision between 2 DVDs and more compression or 3 DVDs with higher quality which id states will not be noticeable unless you know what to look for, is definitely an easier decision to make: I don't think anyone wants to keep swapping discs in and out while they are playing the game.

As for loading times, well I think it's a jump from JC:"the smaller we make it on disc, the faster it loads" (which is true for any media) to "blu-ray read speed problem" which is what assurdum was asking for quotes of. There are a number of factors which don't seem to have been considered here: Seek time comparisons based on the amount of surface covered by 6.8Gb of data; layer change costs; simple performance of the drives in the consoles when reading SL blu-ray media and DL DVD media. There was a thread about this on this very forum, the numbers don't need to be replicated here to drag this thread off topic. I'll try to find the link.

You're still comparing BR transfer speeds/seek to DVD values. That's not what the decision is about. XBOX has smaller files because of the storage limit. PS3 benefits (according to id) of smaller files too because of BR speed. IF the xbox quality is good enough then _the same process_ it's good enough for the PS3. Once again, same solution for two different problems. They aren't putting smaller files on the PS3 "just to appease the xbox fans", or "just 'cuz", there's a reason that is also different from the xbox reason. And again, according to id.

I would expect id to fully explore this before making a decision for Rage, and if they are serious about idT5 becoming a thirdparty engine leader, I would also expect their tools to reflect that exploration so that licensees can make similar judgements for their own games with differently sized data sets.

If id Tech 5's MT tools are anything like the ones from ETQW you'll be able to generate MTs from uncompressed all the way to tiny mb sized (fugly) ones simply by adjusting three sliders.

...seems to sit perfectly in line with this expectation - the continuation of the thought is "If we don't get it to look decent on the 360 on two discs, we'll probably generate a specific data set on the PS3", which a lot of people seem to have missed. Either way it's an if, and like he said, there's half a year until the results of that if can be evaulated.

Which should tell everyone that thinks they are "porting" to the PS3 that id is in fact, looking out for the optimal path in performance/quality for all platforms, don't you agree?

 

[catisfit]

That's not my opinion. That's the opinion of id Software which they articulated last year at QCon when they mentioned the XBOX version might be getting extra compression to get smaller files:

They're talking about the decision to go 2-disc on 360 there - particularly to assuage the fears of 360 owners who may have been concerned that there was going to be an appreciable IQ loss due to the decision.

Yet that was QCon last year, and now we're hearing that they are still working on extra compression methods to try and get as much quality out of a 2-disc release as they possibly can (and laudably so).

So my point is simply this: They don't yet know qualitatively what that difference is going to be. The simple fact that they are working on extra compression methods right now suggests that it's not yet where they want it to be, and they are working on improving it. Therefore, the final difference between the 360 data set and a potential PS3 data set is unknown, and therefore the decision over whether to use the 360 data set or generate another one for PS3 is not going to be made until the final 360 data set quality has been determined, supposedly in six months or so.

What they said they want to do last year doesn't really have much bearing on what they manage to have done six months from now, other than to assure that they think it's important to work on it and are putting significant effort into doing so, so we can likely assume that whatever the situation is in six months, it's about the best they could do given the constrants. We'll have to wait and see where that is, though, and so will they.

You're still comparing BR transfer speeds/seek to DVD values. That's not what the decision is about. XBOX has smaller files because of the storage limit. PS3 benefits (according to id) of smaller files too because of BR speed. IF the xbox quality is good enough then _the same process_ it's good enough for the PS3. Once again, same solution for two different problems.

I think streaming times will be the crux of the decision. Like I said, any media benefits from smaller files.

The 360 version will be squeezed into 2x6.8Gb. That has particular ramifications on the "access time" to retrieve any particular piece of data as it means full surface coverage and DL media, and results in particular seek times, potential layer changes and transfer rates, and of course the extra decompression algorithm running once the data comes back from the optical drive.

The decision for the PS3 version is whether or not to use the same dataset or to generate a new one. In the first instance, the quality will be identical, but the data set will be smaller (no data replication as needed to minimise disc swapping), and the media will be SL and have a much smaller surface coverage, resulting in different characteristics for seek time and transfer rate, but with no possible layer change and of course the same extra decompression requirement, albeit running on a different architecture so that may vary in cost one way or the other.

In the second instance, they increase the data set quality but also it's size, resulting in increased seek times, transfer rates, but possibly avoiding the extra decompression step over the first instance.

So the decision as I see it is whether or not the quality difference is large enough with a similar average data "access time" for the data so that the gameplay is not affected. There are way too many variables in there to make a judgement on this at this point - not least of which what the two quality levels may be, but I see the characteristics of the data on the media and the drives themselves to be an important part of the decision.

If you don't take the media and drives themselves into account, then you're saying that yes, smaller files are a benefit, but you have no basis for comparison because a smaller file on one media may very well be slower to read than a larger file on another media.

HDD caching, recommended HDD installs/copies or even mandatory installs are possible factors too - but don't apply to all gamers unfortunately.

Which should tell everyone that thinks they are "porting" to the PS3 that id is in fact, looking out for the optimal path in performance/quality for all platforms, don't you agree?

Well yes, that's exactly my point, and exactly the reason I wrote it in the first place

 

[Richard]

They're talking about the decision to go 2-disc on 360 there - particularly to assuage the fears of 360 owners who may have been concerned that there was going to be an appreciable IQ loss due to the decision.

YES! EXACTLY! And since that problem, put through a process, generated a solution that also works to solve the PS3's different problem means the same quality that is judged "good enough" by id for xbox is good enough for PS3! So, of all that we know thus far, compressing the PS3 MTs more than disc space limitations would require does not necessarily equate to a "gimped port" <---- my whole point.

Therefore, the final difference between the 360 data set and a potential PS3 data set is unknown, and therefore the decision over whether to use the 360 data set or generate another one for PS3 is not going to be made until the final 360 data set quality has been determined, supposedly in six months or so.

Yes, but as per id's latest:

So if we get it to look decent on the 360 on two discs, we’ll probably leave it with that data set on the PS3

The XBOX "data set" quality will be the same for the PS3 if the same level of compression is used, thus, like the xbox, it will also "not be noticeable by the average user". As for what you said about only being decided in 6 months, that doesn't invalidate their affirmation back at QCon since back then the XBOX decision wasn't final either. In fact, there was also talk of id pressuring Microsoft to soften the duplication costs, etc.

They did an internal test and quality was deemed good enough. A final decision hasn't been reached as the game isn't finished. MT tech let's them easily drag this decision on up until the eve of sending the gold candidate to Sony/MS for approval.

In the second instance, they increase the data set quality but also it's size, resulting in increased seek times, transfer rates, but possibly avoiding the extra decompression step over the first instance.

The "extra" decompression still happens in this instance. We're talking degrees of lossy compression of the MT, i.e. throwing away data rather than using other, more sophisticated, algorithms using more memory with larger dictionaries, etc.

As for using the optical drive characteristics, I only mentioned: "You're still comparing BR transfer speeds/seek to DVD values." I.e. you don't need to directly compare DVD to BR to judge whether the PS3 should have more aggressive compression. Obviously you need to take into account the DVD drive caracteristics for the xbox MTs and the BR drive characteristics for the ps3 MTs.

 

[Richard]

Is it just me or did Todd announce Rage will be out this year?

 

[slapnutz]

Is it just me or did Todd announce Rage will be out this year?

Nope. If you listen carefully he says... "err... no, wont be out this year".

Also he is shaking his head which pretty much reinforces the body language.

 

[Rangers]

Wow, Id sure is slow..I expected RAGE this year.

Really hope we get a nice Doom 4 trailer at E3 though..

 

[the ignoramus]

The first realtime demonstration was back in summer of `07.
Are Rage and Doom 4 being intentionally pushed back towards the last few years of the current hardware cycle. id software fully intends to fight UE4 and CryENGINE3 with id tech 5 and if id tech 5 can flex its muscles and market itself in two full blown, hugely impressive, current gen multiplat games (Rage and Doom 4) around the same time Epic and Crytek start marketing UE4 and CryENGINE3, (neither of which will likely debut on current gen consoles) id will have a competitive edge.