The Framerate Analysis Thread part 2

I've already rendered the 45% graphs with a small vertical bar to indicate a torn frame. What I think I will do is overlay a new version with a larger vertical bar - this will be based on analysis of the 712 lines of resolution. So a non-perceptable tear will have a short indicator, a noticeable one will be larger. In theory then everyone will be happy :)
 
I really think it should, many people make an effort to make sure they have no overscan and those are likely the exact same people that would be bothered the most by tearing. Include both figures if you want, but to my eyes any tearing at all looks hideous and any games that features heavy tearing should be called out for it. Its a pet peeve of mine, I just see no instance where tearing is an acceptable compromise, and I force triple buffer vsync in every PC game I play precisely because of that.

It's interesting how tastes vary from individual to individual.

For me v-sync + triple buffering is the antichrist when it comes to action games, especially if they require a swift response as there's always that little bit of input lag. In such cases, while not optimal I'd prefer tearing over the input lag.

In other cases, such as point and click adventure games, then it doesn't matter as much and I don't mine a little mushy feel in exchange for no tearing.

But I'd agree, if a cropped analyses is made, an uncropped one should also be included.

Regards,
SB
 
It's interesting how tastes vary from individual to individual.

For me v-sync + triple buffering is the antichrist when it comes to action games, especially if they require a swift response as there's always that little bit of input lag. In such cases, while not optimal I'd prefer tearing over the input lag.

In other cases, such as point and click adventure games, then it doesn't matter as much and I don't mine a little mushy feel in exchange for no tearing.

But I'd agree, if a cropped analyses is made, an uncropped one should also be included.

Regards,
SB

Even with triple buffering there's still a lot less input lag playing a PC game at 60fps with a USB mouse to your average unstable 30fps console game with a wireless controller. I sure as hell don't ever get between 200ms and 300ms of input lag as has been recorded in Killzone 2 for instance.

The huge increase in IQ without scarificing any performance (well save for some video RAM but my card has more than enough of that anyway) is well worth the small increase in input lag which I don't tend to notice at all.

I agree its all down to individual preferences, which is honestly why I love the PC platform so much as I get to decide where the performance of my GPU is put. Something like the reent Red Faction videos actually made me feel sick, the tearing means that the console version would be totally unplayable for me. If the PC port is somewhat decent (or at least not a total mess like Saints Row 2 was) then I could quite happily buy the game as getting rid of the tearing is just a click away.

Oh and I'm sure I speak for most here when I say a big thanks to going to all this effort, grandmaster, as stated in the article there's probably not a load of us that are interested in this kind of thing, but for the few of us that are, its genuinely fascinating.
 
Last edited by a moderator:
The huge increase in IQ without scarificing any performance (well save for some video RAM but my card has more than enough of that anyway) is well worth the small increase in input lag which I don't tend to notice at all.

I agree its all down to individual preferences, which is honestly why I love the PC platform so much as I get to decide where the performance of my GPU is put. Something like the reent Red Faction videos actually made me feel sick, the tearing means that the console version would be totally unplayable for me. If the PC port is somewhat decent (or at least not a total mess like Saints Row 2 was) then I could quite happily buy the game as getting rid of the tearing is just a click away.

Oh and I'm sure I speak for most here when I say a big thanks to going to all this effort, grandmaster, as stated in the article there's probably not a load of us that are interested in this kind of thing, but for the few of us that are, its genuinely fascinating.

While I disagree that tearing (or in this case removing tearing) is a huge IQ increase, I'm sure others would make the same cases of whether AA is a huge increase in IQ or not.

But that's where we come to whole hearted agreement. In general PC games at least allow each individual to tailor what aspects of IQ they wish to have and which ones they disable if it doesn't affect them as much as others.

And yeah, thanks for the effort grandmaster, interesting stuff to look at.

Regards,
SB
 
Thanks. Looking forward to the reveal of all the 'surprises' (whatever those will be).

:D

PS: Wondering why nobody is replying to the 'First fruits of the v-sync clean-up tool' :D, because almost everybody wanted to see 'Uncharted: Drake's Fortune' (yes, i know, it's not the full analyses, yet).

:eek:
 
Thanks. Looking forward to the reveal of all the 'surprises' (whatever those will be).

:D

PS: Wondering why nobody is replying to the 'First fruits of the v-sync clean-up tool' :D, because almost everybody wanted to see 'Uncharted: Drake's Fortune' (yes, i know, it's not the full analyses, yet).

:eek:

I dont like it. I understand DF is a business, but from our point of view isnt it just going to allow publishers to create "bullvideos" by stripping the tearing out of their preview game footage?
 
I dont like it. I understand DF is a business, but from our point of view isnt it just going to allow publishers to create "bullvideos" by stripping the tearing out of their preview game footage?


As noted on the website, Microsoft already mandates that the videos be free of tearing for XBLM... And besides, it's all for marketing purposes. If gametrailers or gamersyde want to put up direct captures as is, that's totally within their power...
 
I dont like it. I understand DF is a business, but from our point of view isnt it just going to allow publishers to create "bullvideos" by stripping the tearing out of their preview game footage?

Understanding what you mean, but, as AlStrong already said, maybe you should have a look at this:

http://www.digitalfoundry.org/blog/?p=694

Well, Microsoft’s conditions for Xbox Live Marketplace video include a stringent guideline that all footage must be v-locked.

:eek:
 
I dont like it. I understand DF is a business, but from our point of view isnt it just going to allow publishers to create "bullvideos" by stripping the tearing out of their preview game footage?

Well that is of course true, but MS's guidelines have been in place since the launch of Live and nobody has complained yet. It means that some poor fool has been going through their trailers frame by frame and copying and pasting frames. In extreme cases, they've been taking frames into Photoshop and reconstituting them.

Also consider the basic internet video point: the vids are 30fps, the raw output of the console is 60fps. Every other frame is being lost. Now, either you're going to lose a torn frame or you're going to lose a clean one. If you lose the clean frame, the torn one will be displayed for twice as long as it will be when you play the game. So on balance, I know which I would rather have.
 
Infamous is v-synced but frame rate varies between 20fps and (very, very rarely when there's a lot of sky) 40-50fps. Across an eight minute selection of clips, the average is 29.9fps.

Comparing it to its nearest relation on 360, Crackdown is generally the much smoother game, but that loses v-sync whenever the engine is stressed.
 
I've only observed it go above 30fps in two of the cutscenes (fleetingly) and when you're on the quayside by the water. Most of the time it's at 30, or below if there's any action. Frame rate isn't one of this game's strengths as the forthcoming analysis will show.
 
Infamous is v-synced but frame rate varies between 20fps and (very, very rarely when there's a lot of sky) 40-50fps. Across an eight minute selection of clips, the average is 29.9fps.

Comparing it to its nearest relation on 360, Crackdown is generally the much smoother game, but that loses v-sync whenever the engine is stressed.

Is their anyway you can list standard deviation stats?
 
Oddly enough this was suggested to me some time back. There is no reason why this cannot be done for 480p on Xbox 360. PS3 is a bit more complex as the games are signed to run at either 480p or 576p (depending on the region) and rarely both.
 
Oddly enough this was suggested to me some time back. There is no reason why this cannot be done for 480p on Xbox 360. PS3 is a bit more complex as the games are signed to run at either 480p or 576p (depending on the region) and rarely both.

I think you misunderstood me not standard definition but the standard deviation of how much the fps are deviating from the median. I just interested because the average sometimes does not tell the whole story especially when the framerate is varying by so much from the 20's to the 50's.
 
Back
Top