This is actually a bigger question than you might think. I have not seen any shots at 2160p, no. But the reported numbers from Digital Foundry won't be the cap either, because they're ragged and uncertain. So, when a game has dynamic resolution, how do we tell what the ceiling is if it never gets there? I chose to answer this based on two principles: the statistical nature of game analysis, and the precedent of framerate reporting.Are we sure Assassin's Creed Origins go all the way up to 2160p ? Correct me if I'm wrong.
It's partially due to constraints. Paradoxically, using pictures takes up less space, in terms of post limits. The number of characters needed to link to each pic is fewer than the number needed to present all the textual data within the picture. This becomes more true when you add all the BBCode tags that'd be needed to properly tabulate everything. ResetEra's per-post character limit is higher than the list's old home, but could still have been an issue. (Note that pictures have a different limit, the max number allowed per post, so at some point I'll have to overflow anyway. But it'll take months longer to reach that point.)One thing I always wondered though is why images when a text chart would do just fine (and require less work/space)? The amount of bandwidth this thread will take up by the end of the gen is, well the thought is quite scary.
This is actually a bigger question than you might think. I have not seen any shots at 2160p, no. But the reported numbers from Digital Foundry won't be the cap either, because they're ragged and uncertain. So, when a game has dynamic resolution, how do we tell what the ceiling is if it never gets there? I chose to answer this based on two principles: the statistical nature of game analysis, and the precedent of framerate reporting.
When Digital Foundry or anyone else examine a game, the material looked at is a small sample of the entire experience. DF don't give figures, but it's basically guaranteed that they pixel count less than 100 screenshots per game. (I'd wager the number is closer to 20.) Likely only a couple hours of video is run through their fps analysis tool. (Only a few minutes are presented publicly.) All this is supplemented by the technician closely observing longer runs of the game while playing. But as useful and probative as such examination is, it's necessarily less objective and metricized than the more rigorous methods.
None of this undermines the general results. It is a fact that even a small number of samples from a large population can let us draw strong conclusions if collection was unbiased. But when attempting to describe extremes--the lowest or highest endmembers--samples are a less sure guide. Chances that we've selected the very best (or worst) frame are very low. In this case, I know that Digital Foundry didn't, because I myself have counted sizes above their reported maximum, all the way up to 1660p. But in turn, chances I've seen the very highest are low, so the ceiling is probably higher still.
There's already an area where we encounter such fluid and uncertain data: framerate. A title that runs at 50-55fps all the time is not typically described as a "52fps game", and almost never as a "55fps game". It will be called a 60fps game, even if it never reaches that. We simply assume that the target is a standard milestone, despite not being hit at any time.
So that's how I reported the AC Origins target: a standard size above the highest mark actually seen. But as I've been answering your question, the point occurred to me that--unlike framerate--there are multiple common landing points in excess of the measurements. Given the big gap between observation and proposed target, it would be more conservative to posit an 1800p limit. Therefore I'll change the info when next I update.
It's partially due to constraints. Paradoxically, using pictures takes up less space, in terms of post limits. The number of characters needed to link to each pic is fewer than the number needed to present all the textual data within the picture. This becomes more true when you add all the BBCode tags that'd be needed to properly tabulate everything. ResetEra's per-post character limit is higher than the list's old home, but could still have been an issue. (Note that pictures have a different limit, the max number allowed per post, so at some point I'll have to overflow anyway. But it'll take months longer to reach that point.)
Second, and perhaps more importantly, pictures let me fully control the design and presentation. A wall of grey text columns can be harder to parse than when broken up by stronger graphical elements; compare the IGN list linked in the OP, for example. It also reduces the possibility of weird break points on mobile.
Of course, there are some infelicities. Search is harder, there's more setup time, and as you say bandwidth used is much higher. But the entirety of all pictures used is about 6MB. You'll burn up way more bandwidth looking at a single SunhiLegend GIF. In short, I think the clear and distinctive look is worth the peccadillos.
They doubled the resolution from 1080p to 1440p so it will be more clear on 4k screen especially background.
This is actually a bigger question than you might think. I have not seen any shots at 2160p, no. But the reported numbers from Digital Foundry won't be the cap either, because they're ragged and uncertain. So, when a game has dynamic resolution, how do we tell what the ceiling is if it never gets there? I chose to answer this based on two principles: the statistical nature of game analysis, and the precedent of framerate reporting.
When Digital Foundry or anyone else examine a game, the material looked at is a small sample of the entire experience. DF don't give figures, but it's basically guaranteed that they pixel count less than 100 screenshots per game. (I'd wager the number is closer to 20.) Likely only a couple hours of video is run through their fps analysis tool. (Only a few minutes are presented publicly.) All this is supplemented by the technician closely observing longer runs of the game while playing. But as useful and probative as such examination is, it's necessarily less objective and metricized than the more rigorous methods.
None of this undermines the general results. It is a fact that even a small number of samples from a large population can let us draw strong conclusions if collection was unbiased. But when attempting to describe extremes--the lowest or highest endmembers--samples are a less sure guide. Chances that we've selected the very best (or worst) frame are very low. In this case, I know that Digital Foundry didn't, because I myself have counted sizes above their reported maximum, all the way up to 1660p. But in turn, chances I've seen the very highest are low, so the ceiling is probably higher still.
There's already an area where we encounter such fluid and uncertain data: framerate. A title that runs at 50-55fps all the time is not typically described as a "52fps game", and almost never as a "55fps game". It will be called a 60fps game, even if it never reaches that. We simply assume that the target is a standard milestone, despite not being hit at any time.
So that's how I reported the AC Origins target: a standard size above the highest mark actually seen. But as I've been answering your question, the point occurred to me that--unlike framerate--there are multiple common landing points in excess of the measurements. Given the big gap between observation and proposed target, it would be more conservative to posit an 1800p limit. Therefore I'll change the info when next I update.
The mode is enabled by default. Note that, while Blue Reflection targets 60fps on Pro, it is double-buffer vsynced. This means that when it drops at all, it snaps all the way down to 30fps rather than smoothly varying up and down. As a result, it will not feel like consistent 60fps. Battle sequences especially, with alpha effects on attack, are liable to be 30fps more often than not. But parts of them, some of otherworld exploration, and almost all of school exploration and menus are at 60fps.Does anyone know how to enable Pro features in Blue Reflection? Is it just on by default when inserted into a Pro? I'm using a 1080p TV and would love to have 60 FPS. The game does not appear to run at 60 for me currently.
Increase from native 1080p to native 1440p. Some improvements like clearer filtering and less artifacting or shimmering on distant objects comes naturally with the resolution increase.
Besides resolution, it's a bare-bones patch. No increase in draw distance, no higher quality effects etc. The only other thing it improves is a few instances in cut-scenes where the OG PS4 hard locks to 30 FPS for a few seconds (like when you first rescue Miller), the Pro patch lets those moments stay to 60 FPS.
http://www.eurogamer.net/articles/digitalfoundry-2017-what-does-the-mgs5-ps4-pro-upgrade-actually-do
The return of the best thread ever!!
I would like to know if Wolfenstein 2 is downsampled on HDTVs.
I can't find an official confimation anywhere.
Liabe Brave
have you had a chance to sample Wolfenstein with or without the 1.02 patch which reportedly cuts back on lighting to improve other things ?
Could someone explain me what jittered means? Or at least explain me the con of YS VIII
Wait, the MGSV patch doesn't downsample on 1080p??
What's weird about this one is they pretty much said ahead of time it wouldn't downsample. Wording of the Tweet announcing it specified that owners of 4K displays would get a benefitWait, the MGSV patch doesn't downsample on 1080p??
FucKonami and fuck you Sony, fix your shit!
Thanks. You know your shit.The mode is enabled by default. Note that, while Blue Reflection targets 60fps on Pro, it is double-buffer vsynced. This means that when it drops at all, it snaps all the way down to 30fps rather than smoothly varying up and down. As a result, it will not feel like consistent 60fps. Battle sequences especially, with alpha effects on attack, are liable to be 30fps more often than not. But parts of them, some of otherworld exploration, and almost all of school exploration and menus are at 60fps.
Anyone know the answer to my question a couple of posts above?
Inside a game engine, scenes are described in great mathematical detail. Your screen that the game will be displayed on doesn't have resolution that high, so the game decides what color each pixel will be by taking a sample from the underlying scene description. These samples can be taken in all sorts of patterns. A regular grid is the simplest to imagine, but using simple patterns can exaggerate the sharpness of transitions between color zones, the results being perceived as aliasing or shimmer.Could someone explain me what jittered means? Or at least explain me the con of YS VIII
You're right, it's always been phrased weirdly. I was trying to compact too much info into too few words, and the image didn't help. I've made revisions to both; please let me know if it's clearer now.So the modes with the red X next to it cannot be used on a 1080p TV at all? That section about Downsample modes is really confusing.
Yes, it does down sample. This was mentioned in the DF video.
edit: I can't seem to find it, but I'm sure I saw this somewhere.
edit2: It's mentioned in the NXGamer video. Down sampling is supported:
https://www.youtube.com/watch?v=XXcejKC__V0
Inside a game engine, scenes are described in great mathematical detail. Your screen that the game will be displayed on doesn't have resolution that high, so the game decides what color each pixel will be by taking a sample from the underlying scene description. These samples can be taken in all sorts of patterns. A regular grid is the simplest to imagine, but using simple patterns can exaggerate the sharpness of transitions between color zones, the results being perceived as aliasing or shimmer.
Jittering is when the the sample position is moved around slightly each time, so that you get less predictable, less harsh transitions. But you probably don't want to just take samples at random, so often the "jitter" is repeated over and over in larger units. This is what I think we might be seeing on edges in Ys VIII. Here's a 400% zoom that shows what I mean.
Note how there's not a sharp edge, nor a gradually smooth edge, but noise broken up into chunks along the outline. It's not impossible that this is artifacts from checkerboard rendering (CBR). But it really doesn't look like it to me. As the name implies, CBR uses a regular grid. When it messes up, its artifacts are gridlike too. Here's a 400% zoom of a waving flag:
(Note that CBR is almost never this bad; Watch_Dogs 2 has by far the worst CBR artifacting I've seen. I picked it precisely because it's unusual, an extreme example that's very obvious.)
Keep in mind that I include these details because I observed them and I try to be open about my analysis. In practice, even the worst case may be barely visible, from viewing distances on a regular screen. The noise in Ys VIII is not necessarily bad; it does soften edges and reduce aliasing. And on Pro, the game has a much higher resolution than on standard PS4. My personal feeling is that the tradeoffs are worth it, but you should make your own decisions. Here's the full 4K shot my zoom comes from. Check it out without zooming and see if it bothers you.
You're right, it's always been phrased weirdly. I was trying to compact too much info into too few words, and the image didn't help. I've made revisions to both; please let me know if it's clearer now.
I apologize, but I won't have time to rectify the footnote problem today. It seems to control the alignment properly I'll have to make them part of the images again. This will increase the vertical extent of the lists, but they're too unruly otherwise. I should have this done tomorrow.
As always, any screenshots or video you can provide would be much appreciated. Any issues, comments, or suggestions are also welcome. Thanks!
Uncharted 4
The Lost Legacy
Ratchet and Clank
Horizon
Hitman
Resident Evil 7
FFXV
GT Sport
Infamous
What game not on this list has the best HDR implentation?