[obonicus]

Mod : Some posts merged here from Ratchet and Clank thread.

I'm not sure if this is the right place for this (it addresses future R&C games) but Mike Acton put up a very interesting blog post on framerate. It'll probably upset people here, but hopefully we're not going to freak out about it like I've seen elsewhere. It's nothing we haven't heard, though.

http://www.insomniacgames.com/blogca..._acton/1503082

Don't want to quote anything since the choicest bits are also really inflammatory and may cause people to not even read the post.

[Billy Idol]

interesting read.
I (maybe get banned for this ) fully agree with the blog post: framerate doesn't affect me, I am not sensitive to framerate and I don't mind 30 fps at all.
I really like the extra bling going the 30fps route, compared to a super smooth okaish looking ~60 fps game.
Up to now, only one game I have played worried me due to framerate (I am looking at you Mass_Effect.ppt) and I could nonetheless enjoy this game!
What affects me more is (framerate) stuttering due to in-between level loading (now, I am looking at you KZ2 and GeOW2), which was kind of annoying, but in the end did not hinder me to enjoy the games.

[obonicus]

I don't know. I think he's right. I've seen that sort of thought in tons of sandbox games. I've seen people do crazy stuff in Crackdown just to see how much they could make the otherwise solid engine chug.

[Dr. Nick]

Originally Posted by obonicus

I don't know. I think he's right. I've seen that sort of thought in tons of sandbox games. I've seen people do crazy stuff in Crackdown just to see how much they could make the otherwise solid engine chug.

Yeah and in some cases go online and complain about how the framerate isn't smooth or post videos that while they may seem impressive to some makes the game look bad to others.

I said this before in a previous thread that even next gen you guys should expect the majority of games to run at 30fps.
You guys might go wow I maybe I can play UC4,or whatever will be their main title next gen, at 60fps but once you're able to see what ND can do at 30fps you're going to spend all of your time talking about that game while most of you will complain that Tekken 7, DMC6, NG3, DOA6, and FORZA4 doesn't look all that hot or had to sacrifice resolution to get their framerate up that high while keeping their graphics up to par with other games of the generation.

[Cornsnake]

Originally Posted by obonicus

I don't know. I think he's right. I've seen that sort of thought in tons of sandbox games. I've seen people do crazy stuff in Crackdown just to see how much they could make the otherwise solid engine chug.

I suppose it's a lot to ask from a sandbox game to have to keep it's graphics simple enough to maintain a solid framerate even in the rare occurrence that players cause an extremely heavy load on the engine.

Insomniac's games so far have all been build around individual levels. So it should be that much easier for them to maintain a solid framerate.

[ultragpu]

30fps rock steady with better visual is the way to go for me, definitely a much better viewing impact than a 60fps with lower details. In Insomniac's case I just wish they really implement the much needed HDR and a better texture streaming engine for their next game. SSAO helps alot these days too. I'm just happy now that they have a clear direction in the priority and hopefully they will ascend to the top tier in graphics.

[Prophecy2k]

In terms of processing resources (assuming a single developement platform) how much processing resources would a drop from 60fps to 30fps free up?

Rachet & Clank games look amazing enough to me that I thought they were indeed 30fps games. If they can free up significant resource for adding lots more visual "pop" then why not?

[obonicus]

They may not have to sacrifice resolution anymore, for one. Joker has said here that going from 30fps to 60fps is more than twice as hard.

[obonicus]

Like homerdog said, I think people may notice which looks better. Enough to influence sales? AFAIK Acton isn't saying that 60fps isn't superior to 30fps, it's just that 60fps isn't a selling point. It's not going to push anyone to buy the game.

[Npl]

It depends highly on the gameplay if 60FPS is important or not. if you look at all games and "average" them, 60FPS doesnt matters. It doesnt matters for "The Sims", it doesnt matter for WOW, it doesnt matter for most FPS`s (since the emphasis is on hiding nowadays), it doesnt matter for RPGs, it doesnt matter for RTS, it doesnt matter for most driving games.

Most of the popular games dont require it, but its still a stretch that this is also valid for all genres/niches. Would Burnout/DMC/Ratchet be as well received if they werent 60FPS? I dont know but I know it would hurt them disproportionally more then the rest of the games.
I would be very skeptical and think twice if those where dropped down to 30FPS. So while having 60FPS wont get you new customers it might aswell lose you old ones.

[Arwin]

It's all about the gameplay, imho. And by that I don't mean that it doesn't matter how good your graphics or sound or story are etc, but what I mean is that it depends on the type of gameplay you offer what kind of impact various technical choices have.

When it comes to Insomniac's current position, I think that there is a fundamentally different problem. For Ratchet & Clank 4, the gameplay was already not enough of an advance over the previous genre, but because it was the first next-gen version, the game got away with it partly thanks to very nice graphics and presentation, which managed to stand out even at 60fps that many games running at 30fps weren't even able to match.

But R&C 5 is facing a different market situation in almost all of these ways. In this case I feel the Eurogamer review is mostly spot on in pointing out that it is above all the gameplay that is lacking. It's high time for completely new IP. If I were Insomniac, I would try out a bunch of DLC games with wildly experimental graphics and gameplay if possible, or really go to town with the new motion controller. Do something really new in a small format and test the waters. Maybe make a family game with lots of multiplayer options, similar to Nintendo's work on the Wii, as with the PS3 Slim and soon the motion controller the Playstation will move towards the more casual gamer of its PS2 era.

And to come full circle, with highly stylised family oriented graphics and fun multiplayer, 60fps will become important again.

[Joshua Luna]

Originally Posted by liolio

All this "30fps is not enough talk" needs some serious blind testing imho.

There are a lot of approaches to deal with your question, but probably most telling would be apply a scenario where "X is not enough" and challenge the basis--and see people's response. Or turn it into a challenge statement:

UC2 graphics are enough.
Crysis graphics are enough.
2 button controllers are enough.
One racing sim is enough.
5 guns is enough.
One FPS game is enough.

In the end market demand determines what is enough--even if market segments can see the difference.

Where the rubber meets the road is when the primary audiance (e.g. racing fans) demand a feature (e.g. 60Hz) because they do not and anything less is not enough.

Whether 60Hz is noticable isn't really debatable, the questions are is it distinct enough to most consumers? A selling point? Make the game better? Lack thereof make the game unplayable? A good tradeoff for other features?

Would Gears 2 be a better game at 60Hz with cut down graphics, AI, etc? Probably not. On the other hand Forza 3 would be less compelling to its core demographic at 30Hz.

Originally Posted by Shifty Geezer

Billy Idol - Banned
Obonicus - Banned
Mike Acton - He's banned. Twice!
...I imagine the findings are most people can't perceive the difference.

About time! No mercy for those heathens! Now if you can have the same approach to the baiters, trollers, and obstinant you will get Mod of the Year As for 60Hz, I agree, I bet most cannot notice.

Of course a lot of people don't notice the finer points of gameplay, either. I think a very strong core really set the tone for a lot of games and that there is a huge disconnect between the casual "that was so cool" market and the "I want every inch of gameplay world to be perfectly mapped for optimal competitive carnage!"

Btw, this is why the B3D Live group rocks--we actually discuss game design, technology, and all the bad posts here during gameplay :P

Originally Posted by Prophecy2k

In terms of processing resources (assuming a single developement platform) how much processing resources would a drop from 60fps to 30fps free up?

Rachet & Clank games look amazing enough to me that I thought they were indeed 30fps games. If they can free up significant resource for adding lots more visual "pop" then why not?

Carmack noted Doom 4 would look ~3 times better than Rage based on the general larger margine 30Hz gives you. Lets say you need to target 75fps +/- 15fps to have a stable 60Hz game. At 30Hz since you are rendering half as often you can lower some of your tolerance (what was once, e.g., a 2ms window for a task now becomes 4ms, and you also can lower some of your tolerances).

Originally Posted by obonicus

Like homerdog said, I think people may notice which looks better. Enough to influence sales? AFAIK Acton isn't saying that 60fps isn't superior to 30fps, it's just that 60fps isn't a selling point. It's not going to push anyone to buy the game.

With rare exception (racing sims--although NF Shift vs. FM3 sales will be interesting!!) I don't think sales are hugely impacted and reviewers and their scores (gameplay, graphics) are impacted much by framerate and it isn't something that can be spotted in a) magazine pics b) box art c) online pics d) most gameplay video feeds so on and so forth.

In most cases better still graphics has more impact.

[Billy Idol]

I really wonder what the purpose of this Mike Acton blog post was/is?

A) give an excuse and explanation for the fans (i.e. fan service), why they go 30 fps?

B) denounce the fact that > 30 fps is not taken fairly into account when judging the graphics (see for instance the B3D forum R&C: ACiT thread), i.e. we do 60 fps games if you start to acknowledge properly?

C) a gamer civil war (and in consequence world domination)

D) ...?

[obonicus]

Open discourse with fans? He's being very (and, judging from GAF's reaction, perhaps overly) candid about his motivations.

He's not trying to sell us 30fps as a way to a better game; he's explaining that at 30fps IS' games won't sell any worse and, if they can improve the graphics, might even sell better.

Mike Acton's a member here, too (and even has his own forum, the hotshot), he seems pretty forthcoming about information in general.

[Billy Idol]

Originally Posted by obonicus

Open discourse with fans? He's being very (and, judging from GAF's reaction, perhaps overly) candid about his motivations.

yeah, this would rather fit to IS' games mindset

[Richard]

While I know this was a blog post let me make a couple of observations. On the topic of scores:

Originally Posted by Mike

[...] they found that there was a clear correlation between graphics scores in reviews (where they are provided) and the final scores.

It's good to have independent research on this, despite the fact I, and many others, have been preaching this for years to the "gameplay > graphics" crowd.

And they found no such correlation between framerate and the graphics scores nor the final scores.

I believe Mike but his post doesn't have this data.

As an interesting side-note, our team also found no direct correlation between gameplay scores and final scores, however it does appear that gameplay scores are also influenced by graphics scores. i.e. Better looking games appear to be more “fun” to reviewers, in general.

This is what I have the most problems with. Firstly the post also doesn't show any data to support this lack of correlation and secondly, how did they infer that the correlation was evidence of causation?

About the poll: are 741 non-validated reponses a statistically relevant sample of the target demographics? The poll options also seem unbalanced. What does it mean not interfering with the game? Simply absolute ("don't drop below 30fps") or relative ("if the game runs at 60fps, never drop below 50fps")?

Personally, framerate matters but the actual lower bound is influenced by each game genre and engine (for instance, I can withstand lower absolute fps in DOOM 3 and Crysis than I can with any Source engine games where anything below 40fps seems like a slide-show). Also, a slow FPS (DOOM3, Bioshock, etc.) for me doesn't need the same fps as a BF2, ETQW, etc. Most FPS don't need the same fps as a racing game, or an RPG, etc.

That's why I, and a lot of other people, will hold off playing games until we upgrade our PCs: because we try the game, the fps is not at what we consider playable for that particular game, and we wait until it does. That seems to me framerate DOES have an inpact on our enjoyment, in spite of what this blog post pushes forward.

To reiterate:
I have a problem with the lack of data in the post, that is only hinted at.
I have a problem with the inferrence followed.
I have a problem with generalising fps/graphics to any and all games, engine technology and gametypes.

It's a worthwhile discussion to have and I suggest Mike publish more information about their research.

[Shifty Geezer]

Originally Posted by Richard

It's a worthwhile discussion to have and I suggest Mike publish more information about their research.

As I said earlier, the industry as a whole should commission some proper research. Maybe Gamasutra could do an investigation if the IGDA won't?

[Joshua Luna]

While I am not convinced it will help sales being this open/public, I really like the approach. Thanks Mike.

[patsu]

He's just saying 60fps is no longer a sacred cow in Insomniac. Sounds fair to me.

I agree with Arwin that Insomniac should always focus on gameplay/fun first and forth most.

And did they find reviewers gave overall scores with little co-relation to the high gameplay scores ? That's the reviewers' problem. People may come for nice graphics, but they stay usually because of the gameplay. This allows the developers to capture repeated business easier. A lot of companies die because they can't retain repeated customers. So gameplay is important.

Good graphics can be achieved with a combination of art direction and technology. I feel there are more leeway here. Insomniac is strong in technology, so they may want to experiment and play more with the softer issues -- without losing that technical sharpness.

EDIT: e.g., The ability to capture people's imagination via concepts, story, character, gameplay mechanics, art, etc. See The Last Guardian, Uncharted, Demon's Souls, Flower, LittleBigPlanet, BioShock, Batman:AA, Pan's Labyrinth (the movie).

I think it's also how everything is put together instead of just one or two independent traits.

[dobwal]

Of course great graphics tend to correlate with the overall score and thats because pubs that put alot of work in visuals tend to put alot of work in their overall product. There doesn't tend to be alot of "WOW" games in term of visual, yet "Man this is a POS!!!" in terms of gameplay.

There no correlation between overall score and 60 FPS because high FPS isn't, in and of itself, a metric of quality development. 60 FPS is an option even for a lazy or untalented dev, because 60 FPS isn't the hard part, its trying to balance maximum visuals and 60 FPS thats difficult for just about everybody.

Devs tend to sacrifice 60 FPS because only fraction of the market can tell the difference. Furthermore, that phenomenom is helped because only a fraction of games are release at 60 FPS so the majority of market is left without alot 60 FPS experiences .

Just about any gamer can discern the difference between framerates and resolution given the right circumstances. This is why its important to create studies that mirror reality. And the reality is that people's ability to see or feel a difference in framerate and resolution is determine how fresh the experiences are in their memory. Most mainstream consumers can easily see the difference between 1080p and 720p in a setting like BestBuy because that setting allows you to easily compare two different images. But let those same consumers go around random to their neighbors' homes with HDTVs, they would be practically unable to tell which one was a 720p set or a 1080p.

Same goes for framerates and overall game IQ. If you really wanted to know if people can tell the difference between AA settings. Space the exposure of the two images to your subjects by two days or more. The point of spacing out the exposures is to disallow the creating of a strong baseline for the subjects and basically judge the each image more independently. A person really sensitive to framerates, AA settings and resolution will tend to be able judge an image without the need to compare against fresh memories.

[fearsomepirate]

Originally Posted by dobwal

Most mainstream consumers can easily see the difference between 1080p and 720p in a setting like BestBuy because that setting allows you to easily compare two different images. But let those same consumers go around random to their neighbors' homes with HDTVs, they would be practically unable to tell which one was a 720p set or a 1080p.

Anecdote: I got a PS3 recently and was playing Bioshock. The frame rate was a bitt jittery, so I dropped the machine down from 1080p to 720p. On the PS3, this requires stopping and restarting the game. Even with just the one or two minutes needed to switch, I didn't even notice the lower resolution (certainly not the way I noticed switching from 800x600 to 640x480). I did notice the game running a little more smoothly, though.

[obonicus]

Mike sorta discredits the sample by pointing out that it's probably self-selecting, which is why the 'do not cares' are so low.

[Dr. Nick]

Meta Critic does seem to be important. The scores seem to be a marketing tool and from what I've heard some devs get paid more depending on the score. If a game scores a 10 in the graphics department it might push that 8.5 up to a 9 or a 9.5 which in turn gets a great deal of attention. Based on what we have seen so far this generation it seems like getting a 10 in graphics is more in reach at 30fps than at 60fps.

[patsu]

We need to be careful here. A high metacritic score may not be translatable into a 30fps vs 60fps argument. Games look good and play well for a good many reasons.

Don't want to put the cart in front of the horse. I agree that 60fps is not a must. That is all. v_v

There are still a lot of work and variables after this point.

EDIT: Bah... I should just give Mike more room to express his views without overreacting and overanalyzing. In isolation, I see nothing wrong in 60fps or no 60fps.

[Joshua Luna]

Originally Posted by patsu

EDIT: Bah... I should just give Mike more room to express his views without overreacting and overanalyzing. In isolation, I see nothing wrong in 60fps or no 60fps.

But it isn't isolated. Better graphics trend toward better reviews. Better reviews trend toward better sales (all things being equal). 60fps doesn't necessarily mean better graphics; indeed in many cases I would believe if a developer had a 60Hz version and 30Hz version w/ extra eye candy the 30Hz would be considered superior.

There are exceptions (e.g. racing sims, fighting games), but you won't see a huge docking of points (lets say 2-3 points) for not hitting 60fps in those games. How many reviews gave Shift a 5/10 for graphics?