Dead Space: Performance Analyzed

[nandopb]

Cleeve defense mode off) jejee......Great article, a little late but it was worth the wait.Keep them coming(COH if you can).I read a similar one(less detailed) a couple of months ago and you can tell this one took some time to write.

 

[cleeve]

[citation][nom]Curnel_D[/nom]Honestly, this review *is* Biased...

...Next Article: Nvidia's ambient occlusion VS ATI's Global Illumination. Please?[/citation]

Honestly, it *isn't* biased. This is a game performance analysis, not an Nvidia vs. Ati article. The purpose is to see what you need to run a specific title. As such, title-specific enhancements are fair game.

In the same vein, sooner or later I will review a Dx10.1 title, and in that review any app-specific enhancements will *also* be fair game.

Somehow a lot of you are assuming this is a hardware review.
It's not, look at the title: "DEAD SPACE performance analyzed", not "Nvidia vs. ATI performance analyzed"

 

[curnel_D]

[citation][nom]Cleeve[/nom]Honestly, it *isn't* biased. This is a game performance analysis, not an Nvidia vs. Ati article. The purpose is to see what you need to run a specific title. As such, title-specific enhancements are fair game.In the same vein, sooner or later I will review a Dx10.1 title, and in that review any app-specific enhancements will *also* be fair game.Somehow a lot of you are assuming this is a hardware review. It's not, look at the title: "DEAD SPACE performance analyzed", not "Nvidia vs. ATI performance analyzed"[/citation]
In reality, this comes down to be a review of Ambient occlusion technology, which is why everyone is on your ass about it. You're right, this is totally a dead space performance review, but you stated already you chose dead space simply because it's one of the games on Nvidia's AC list.

We're not going to beat you into submission to give us an edit to this article, since that's not the scope. But it'd be awesome if you compared the two competing technologies and other competing technologies in your next game or games review.

 

[cleeve]

[citation][nom]Curnel_D[/nom]But it'd be awesome if you compared the two competing technologies and other competing technologies in your next game or games review.[/citation]

That sounds like a whole separate kettle of fish, and I'd like to accommodate you folks, but how exactly do I compare Dx 10.1 enhancements vs. AO enhancements?

It's not exactly apples to apples, and pretty darn subjective. Even if I could find a game that supported both, it would be an subjective - not objective - comparison, because both features would be supplying different output and performance.

The usefulness of these features comes down to the games that support it, and how much it enhances those particular games. As such, it's almost impossible to look at a couple and say definitively that one is good and the other is bad, because one might be great in one game but suck in another.

We're planning to do a game performance analysis once a month, so sooner or later we're going to have a large body of work that can be referenced, and as we test more Dx10.1 titles and more AO-supporting titles, PhysX supporting titles, etc, folks will have a body of evidence to reference as far as which features truly enhance gameplay, which are pleasant but ultimately irrelevant, and how much those features cost in terms of performance overhead.

But call me biased, and of course I'm going to take issue. I'm the guy whop turns off PhysX for my 3dMark runs because it creates an unrealistic handicap. I'm doing my best to keep things on an even keel, here.

Instead, I'm getting slammed with charges of bias from some folks, regardless. And I have to wonder, what would have made them happy?
After testing the game and seeing the results, was I supposed to bury them and find another title that showed equal performance so that I could be considered "unbiased"?

That doesn't make sense to me at all. In some cases (and I'm not saying yours), it seems more like certain people weren't happy seeing their favorite solution perform less-than-ideally, and immediately called foul.

 

[curnel_D]

Ehh, I may have been out of line with the biased comment, but I still think overall that this review was slightly subjective towards the Nvidia only AO. But I suppose this is a gaming industry problem, not a reviewer problem. Heh. As much as I like the performance Nvidia cards can pump out, AO is a stupid technology, and literally a patch over a missing DX10.1 section in their hardware to make people think that Nvidia has something new, fresh, and cool. (None of which is true).

Comparative before and after screenshots and an in-depth explanation of both AO and DX10.1's Global Illumination would be one way to Compare, but you're right in saying it will be tough to find a game that does both. Despite that, subjective or not, a majority will declare one or the other better, which is exactly what people seem to be looking for.

 

[maximus559]

Wow - this looks like a really good game. Unfortunately, I'll probably never play it. I remember getting all excited when I first heard about this title, then having my hopes dashed when I found it has SecuRom. So sad... and to everyone here who has played it on PC, you do realize you payed to have malicious software installed on your machine, right? Not trying to stir up an argument - just sayin'.

 

[cleeve]

[citation][nom]Curnel_D[/nom]As much as I like the performance Nvidia cards can pump out, AO is a stupid technology, and literally a patch over a missing DX10.1 section in their hardware to make people think that Nvidia has something new, fresh, and cool. (None of which is true).[/citation]

I don't think that's necessarily true; a good, UNIVERSAL, AO option would be a very cool feature; something that could let you to play an older game and really enhance the lighting model would be really cool. Unfortunately, it doesn't have universal game support. As well - from what I've seen in Dead Space so far, anyway - this doesn't appear to make a hell of a lot of difference, it doesn't provide a lot of benefit. Frankly, I don't think I kept that a secret in the review, I pointed all of this stuff out.

As for Dx10.1, I've seen it do some really amazing things in HAWX, but that's the only game I've seen it enhance appreciably so far, and unfortunately it's not much of a game. What ATI needs to do is get more agressive on the developer side and get Dx10.1 support in some triple-A titles.

But I'll be keeping abreast of the new stuff and I'll keep my eye out for new Dx10.1 titles that aren't just visual treats but also offer solid gameplay. If you guys catch wind of something coming, PM me and I'll take a look at it.

 

[San Pedro]

Well I didn't think the review was biased. Like Cleeve said sometimes ATI wins and sometimes Nvidia wins, but in the end most cards were playing this game at ultra high resolution.

Also someone complained about the controls. I've never had a problem with them, but I use a wireless XBOX 360 Controller, and it works great (very well spent $60), I can change weapons no problem. However, I think reviews for the game did state that keyboard mouse controls were pretty bad.

 

[Lans]

I don't think the review is biased but this site as a whole seems favor Nvidia. Admittedly Nvidia got lots more features ATI doesn't (PhysX, AO, and better support with CUDA). I think is the right thing to review each feature in its own right but that is just it. Where are the DX10.1 reviews? Sure there are limited titles but likewise how many titles currently support AO and PhysX?


Apparently I missed the news on Friday that Nvidia will have DX10.1 parts soon so DX10.1 review will be coming (when and not if).

And I agree it might be nice to replay out titles with AO.

 

[cleeve]

I don't think the review is biased but this site as a whole seems favor Nvidia.

Seriously, have you seen the last few monthly Best Cards for the Money articles? Because I write those as well, and ATI has been absolutely dominating lately. And that's not because I have a boner for them, it's because they're offering the best hardware for the cash. That's commendable.

And to be fair, Nvidia has dominated it in the past.

I don't think we favor Nvidia or AMD. I don't think we favor anybody. We just try to call it like we see it.

 

[curnel_D]

[citation][nom]Cleeve[/nom]I don't think that's necessarily true; a good, UNIVERSAL, AO option would be a very cool feature; something that could let you to play an older game and really enhance the lighting model would be really cool. Unfortunately, it doesn't have universal game support. As well - from what I've seen in Dead Space so far, anyway - this doesn't appear to make a hell of a lot of difference, it doesn't provide a lot of benefit. Frankly, I don't think I kept that a secret in the review, I pointed all of this stuff out. As for Dx10.1, I've seen it do some really amazing things in HAWX, but that's the only game I've seen it enhance appreciably so far, and unfortunately it's not much of a game. What ATI needs to do is get more agressive on the developer side and get Dx10.1 support in some triple-A titles.But I'll be keeping abreast of the new stuff and I'll keep my eye out for new Dx10.1 titles that aren't just visual treats but also offer solid gameplay. If you guys catch wind of something coming, PM me and I'll take a look at it.[/citation]
Lol, and we all know how that will turn out. AMD/ATI has some of the worst marketing skills in recent memory. Funny thing, the 3x series made the green team take notice, and now the 4 series has made the green team actually pick up the pace and really put some prices lower to the ground. Any company who can do that has a little bit of cash to throw into advertising. *grumbles about AMD's awful mismanagement.*

 

[Guest]

@JWL3: You dont know the game shortcuts....???

 

[shagrathdex666]

to jwl3

obviuolsy when u download it u'd never get the tutorial on how to play it ,awesome game i'd plat it again you bet,the controls are so perfect guess you forgott to read the instruction manual i guess...

answers...
question 1.. wasd great direction buttons for isaac.

part 2..just scroll with the numbers like in bioshock to change your weapons(you can even customize the order of them)

3 what better way to feel impotent when you get to watch how they dispose of you...tsshhh...

5:use q to utilize a medpack,r to reload,and other letters to quick acces your medkit of choice...

overall a kick ass game..

 

[wavebossa]

I'm reading these comments, and this is just absolutely ridiculous...

All of you guys saying that this is biased somehow need to learn how to read. I mean, this is a review to test out a new NVIDIA feature, NOT AN NVIDIA VS ATI review.

How is he supposed to review a new technology without you fanboys getting so offended all the time? Seriously...

Great article, and very informative. It seems AO will go the way of PhysX for now, but hopefully, it will mature quickly.

 

[Lans]

I still maintain there is a Nvidia slight tilt and I am not saying you guys try to spin everything in Nvidia's favor and hopefully never suggested that. However, choosing to feature some technologies but not other(s) seem biased to me and in this case, just so happens to favor Nvidia.

I still find the reason given for lack of in-depth DX10.1 reviews to date seems superficial... From wikipedia, there is: Assassin's Creed (removed with 1.02 patch), BattleForge, Far Cry 2, STALKER Clear Sky (1.5.0.6 or later), Stormrise, and HAWX (as mentioned).

Pretty small list of titles but as far as I know PhysX and AO is very limited at the moment also. Far Cry 2 is a pretty big title? PhysX and AO are nothing but visual eye candies at the moment so I am still not satisfied with reason given for in-depth reviews of those but not DX10.1 and until then my opinion does not change.

I would like to think of myself first and foremost as a techie and don't care who implemented which technology.

 

[bikeracer4487]

Hey, any chance you could post savegames of where you benchmarked from? I think it'd be really handy if you guys posted the save games and/or timedemos that you use to benchmark so we could run them on our own systems to see how we stack up.

 

[cleeve]

[citation][nom]Lans[/nom]I still maintain there is a Nvidia slight tilt and I am not saying you guys try to spin everything in Nvidia's favor and hopefully never suggested that. However, choosing to feature some technologies but not other(s) seem biased to me and in this case, just so happens to favor Nvidia.[/citation]

Well, for me, this is the second article in the game performance series. In the first, ATI showed a massive performance advantage over Nvidia in Burnout Paradise when AO was enabled, and in the second, Nvidia showed a performance advantage and an exclusive AO feature.

It looks like 1-1 at this point, hardly an Nvidia bias IMHO.

I tell you what though, if the majority of the game performance analysis reviews I do in the future turn out to primarily feature games with Nvidia-only features, and none with Dx10.1 or ATI features, I might put more weight to your 'slight tilt' theory. But for now, ATI is dominating our recommended cards list, and the game titles we've run since this young series began show a 1 win-1 loss record for Radeons. I'm not seeing the tilt, bro.

But if it happens, I encourage you to keep me honest.

 

[g-thor]

The reviewer says, on the Medium Detail page, "the Radeon HD 4650, which can only muster an average 39 FPS and a minimum 22 FPS. It might be playable at this setting, but it's far from smooth." Then, on the High Detail page says, "the GeForce 9600 GT is pulling a minimum frame rate of 25 FPS, which is quite playable but a little below the ideal 30 FPS."

This seems to indicate a personal bias - one is far from smooth at an average 30 FPS, the other is quite playable at 25 FPS. Fortunately he does state that the game does favor nVidia over Radeon, but that doesn't excuse a blatant statement that is so self-contradictory.

Please, try to go for more even-handed journalism.

 

[bikeracer4487]

[citation][nom]g-thor[/nom]The reviewer says, on the Medium Detail page, "the Radeon HD 4650, which can only muster an average 39 FPS and a minimum 22 FPS. It might be playable at this setting, but it's far from smooth." Then, on the High Detail page says, "the GeForce 9600 GT is pulling a minimum frame rate of 25 FPS, which is quite playable but a little below the ideal 30 FPS."This seems to indicate a personal bias - one is far from smooth at an average 30 FPS, the other is quite playable at 25 FPS. Fortunately he does state that the game does favor nVidia over Radeon, but that doesn't excuse a blatant statement that is so self-contradictory.Please, try to go for more even-handed journalism.[/citation]


You're overreacting. First off, the 9600 GT was 4 FPS faster at a much higher resolution and on higher detail, so he's not contradicting himself since one card was, in fact, faster than the other. Secondly, when commenting on the Radeon he said it was far from smooth, not that it wasn't playable. And when he commented on the 9600 GT he said it was "quite PLAYABLE" not that it was "quite SMOOTH".

Please, try not to make unfounded accusations.

 

[cleeve]

[citation][nom]g-thor[/nom]Fortunately he does state that the game does favor nVidia over Radeon, but that doesn't excuse a blatant statement that is so self-contradictory.Please, try to go for more even-handed journalism.[/citation]

Wow. Are you serious?

I said they were *BOTH* playable. I said the 4650 was 'far from smooth' and the 9600 GT was 'below the ideal'.

I'm not sure how they interpret english where you hail from, but 'far from smooth' *IS* 'below the ideal'.

Calling that 'blatant' and 'self-contradictory' is an excercise in hypersensationalism.

Good for a chuckle, but if you're not joking... wow. Dude. You need to have a gbeer and get off the internet before you give yourself a coronary.