The Framerate Analysis Thread part 2

That'll learn me for skim-reading. Of course it's possible, but the video graphs are far more illuminating and there'll be one of those for InFamous some time this week.
 
That'll learn me for skim-reading. Of course it's possible, but the video graphs are far more illuminating and there'll be one of those for InFamous some time this week.

Demo or full game? The press demo is from an April build. The PSN demo is retail as far as I know.
 
Last edited by a moderator:
Got a bunch of screens here showing the lighting is entirely dynamic - showing the same image with different times of day and how it effects both Cole, all characters and all the scenery. Pretty cool.
 
Demo or full game? The press demo is from an April build. The PSN demo is retail as far as I know.

It's from the PSN demo.

Got a bunch of screens here showing the lighting is entirely dynamic - showing the same image with different times of day and how it effects both Cole, all characters and all the scenery. Pretty cool.

Isn't that pretty much standard fare for this style of game? What is noticeable is that the lightning effects from Cole illuminate characters and scenery but cast no shadows.
 
Isn't that pretty much standard fare for this style of game? What is noticeable is that the lightning effects from Cole illuminate characters and scenery but cast no shadows.

Mind you, that could be potentially bad-looking and distracting (low res shadow-maps flickering all the time). :p The performance penalty would be fairly high, and the framerate certainly doesn't need that. :)
 
It's from the PSN demo.

Isn't that pretty much standard fare for this style of game? What is noticeable is that the lightning effects from Cole illuminate characters and scenery but cast no shadows.

Yeah OK...I just thought it was cool and I wasn't sure if character self-shadowing was affected in other open world games. :/
 
Hey grandmaster, seems like you don't have to do that Uncharted "no-tearing" video anymore. Judging from the small Uncharted 2 teaser on gametrailers right now (teasing the full unveiling in a few hours on gttv) Naughty Dog has done it for you. :LOL:
 
It's more a demonstration of the 'clean-up' tool than just showing off Uncharted 1 without tear.
 
This assumes of course that Uncharted 2 is v-synced and the footage hasn't been processed (you'll note that tearing is absent in Uncharted's cut-scenes so ND clearly know how to do it). But knowing ND I'm sure the sequel is v-synced, which makes me even happier bearing in mind how cool the game looks overall.
 
This assumes of course that Uncharted 2 is v-synced and the footage hasn't been processed (you'll note that tearing is absent in Uncharted's cut-scenes so ND clearly know how to do it). But knowing ND I'm sure the sequel is v-synced, which makes me even happier bearing in mind how cool the game looks overall.

Just in case you didn't know, ALL of the cut-scenes in UC1 were pre-rendered.. with the same in game assets of course ;) it's a common trick used in many Sony First Party games such as Resistance, Ratchet, Heavenly Sword etc. KZ2 is pretty much the only game doing everything in real time :cool:
 
It's more a demonstration of the 'clean-up' tool than just showing off Uncharted 1 without tear.

Oh I know, and I'm still waiting for it. :D My post was more of a compliment to the fantastic and smooth visuals of Uncharted 2. :)

This assumes of course that Uncharted 2 is v-synced and the footage hasn't been processed (you'll note that tearing is absent in Uncharted's cut-scenes so ND clearly know how to do it). But knowing ND I'm sure the sequel is v-synced, which makes me even happier bearing in mind how cool the game looks overall.

As MazingerDUDE said, they just pre-rendered the cutscenes, frame by frame no matter how long each frame took, so there's no need to touch it up. Of course, they could have used your technique in this particular gameplay video. Anyway, hell you must be so busy with the blog moving and all, but hey, whenever you do have time, it'd be great to see an analysis of this small segment gametrailers showed. Perhaps we'll have more footage by then, though. :)
 
Even though there's only the 30fps vid to look at, I'm pretty much convinced it is v-synced based on the way it drops frames.

Here's an interesting question. Should frame rate analysis (with graphs etc) only be carried out on final games and playable demos? Is it 'unfair' to do it on things like the CryEngine trailer and the Uncharted 2 vid when the game is not complete and - as of now - there is no gameplay experience or buying decision at stake?
 
Even though there's only the 30fps vid to look at, I'm pretty much convinced it is v-synced based on the way it drops frames.

Here's an interesting question. Should frame rate analysis (with graphs etc) only be carried out on final games and playable demos? Is it 'unfair' to do it on things like the CryEngine trailer and the Uncharted 2 vid when the game is not complete and - as of now - there is no gameplay experience or buying decision at stake?

Well the new 5 min. of gameplay shows clearly but I don't know if it's a real vsync, but something of alternative. It seems that for a fraction of second isn't totally vsynced I don't know how to explain , is it possible?
 
Even though there's only the 30fps vid to look at, I'm pretty much convinced it is v-synced based on the way it drops frames.

Here's an interesting question. Should frame rate analysis (with graphs etc) only be carried out on final games and playable demos? Is it 'unfair' to do it on things like the CryEngine trailer and the Uncharted 2 vid when the game is not complete and - as of now - there is no gameplay experience or buying decision at stake?

Grandmaster, get Gamersyde's wmv encode of the 5 minute video (500MB+)and take a look at it if you don't mind. I went through some of the most intense scenes (the first with the truck chase in the alley being one of them), frame by frame (tapping the right arrow in media player classic, old-school style :LOL:), and each and every frame was a new screen. Either my "technique" is flawed or they're doing post-processing like you can do, or else they're pretty much always above 30. :)
 
Even though there's only the 30fps vid to look at, I'm pretty much convinced it is v-synced based on the way it drops frames.

Here's an interesting question. Should frame rate analysis (with graphs etc) only be carried out on final games and playable demos? Is it 'unfair' to do it on things like the CryEngine trailer and the Uncharted 2 vid when the game is not complete and - as of now - there is no gameplay experience or buying decision at stake?
The question is how often does a unstable framerate get stabilized in retail code.
Devs that are going to prioritize framerate are usually going to do so early on in development.
 
Grandmaster, get Gamersyde's wmv encode of the 5 minute video (500MB+)and take a look at it if you don't mind. I went through some of the most intense scenes (the first with the truck chase in the alley being one of them), frame by frame (tapping the right arrow in media player classic, old-school style :LOL:), and each and every frame was a new screen. Either my "technique" is flawed or they're doing post-processing like you can do, or else they're pretty much always above 30. :)

You tapped the right arrow over 10,000 times?! Try doing it at this point: http://img36.imageshack.us/img36/8179/drakecopy.jpg - just one example though there are a fair few. You might argue that the static scene is generating false positives on duped frames (Uncharted 1 does this which in part explains why we're late with the tech analysis) but that wouldn't explain why the fire on the right isn't updating every frame ;)

I'm not going to take issue with the frame rate because a) where it does drop it is not especially apparent and b) the game isn't finished any way and c) it is, quite frankly, the best thing I've seen on current gen console. I am almost literally aching for this game.

Curiously, the encode I got was a lot smaller than Gamersyde's effort, but of a higher quality. First generation copy versus second, I suspect.
 
The question is how often does a unstable framerate get stabilized in retail code.
Devs that are going to prioritize framerate are usually going to do so early on in development.

You know I used to think the same thing, but then I saw Burnout Revenge 360 optimisations transform the game in terms of refresh rate within a week of development. Playable demos are put out as a sampler of a game experience and so are fair game, but I'm not sure that beta code video is. Although in this case I reckon it probably will stay much the same.
 
You tapped the right arrow over 10,000 times?! Try doing it at this point: http://img36.imageshack.us/img36/8179/drakecopy.jpg - just one example though there are a fair few. You might argue that the static scene is generating false positives on duped frames (Uncharted 1 does this which in part explains why we're late with the tech analysis) but that wouldn't explain why the fire on the right isn't updating every frame ;)

I'm not going to take issue with the frame rate because a) where it does drop it is not especially apparent and b) the game isn't finished any way and c) it is, quite frankly, the best thing I've seen on current gen console. I am almost literally aching for this game.

Curiously, the encode I got was a lot smaller than Gamersyde's effort, but of a higher quality. First generation copy versus second, I suspect.

Haha no, I didn't do that for the whole vid, just some scenes, and yes I have found a few drops during the bus crash and immediately after, at least. :) They're pretty isolated and it's extremely smooth overall though. I 'd say it doesn't even come down to 25fps (in that scene).

Oh btw feel free to upload your encode somewhere. :D
 
Back
Top