I don't trust anything from Digital Foundry, especially considering their obvious bias toward MS and the 360 in nearly every one of their head-to-head comparison. They tend to smear the PS3 version of multi-plat games and do everything in their power to make people think that the 360 version is the "best" one
Digital Foundry Article Technical Discussion Archive [2012]
- Thread starter AlNom
- Start date
-
- Tags
- digital foundry
- Status
Says 'PSman'. They don't smear PS3, but point out shortcomings that do exist. A few of the more recent comparisons I've read have even heavily favoured the PC. They're also written by a range of contributors so will have different flavours.
Gears 3 uses 100% baked environment shadows.
And Killzone games use a mix of both. Games with cascade shadow maps are exceptions rather than the norm.
Digital Foundry has a heavy bias to graphics. And when it comes to graphics, multi-platform games, for a tonne of reasons discussed here almost endlessly, multi-platform games tend to almost always perform a little better graphically on the 360. That is a fact, and Digital Foundry uses methods such as pixel counting and frame-rate measurement to show the differences as objectively as possible.
It is hard to find fault with Digital Foundry in that respect. And I say this as a big Playstation 3 fan, who has both a 360 and Playstation 3, and almost exclusively play on Playstation 3.
Not for foliage.
It uses baked ones for trees, and for foliage like shrubs and bushes it just uses contact shadows.
Some moving objects, like the swaying trees in the last few levels have real time shadows that fade out at a fairly close distance.
https://www.youtube.com/watch?v=8IGRfSfob5U&hd=1
What you say is true but mainly for multiplayer and indoor environments in campaign.
My point is that not 100% of Gears 3 shadows are prebaked.
Is the thoroughness of the analysis related to how much technical info the developer has divulged? How much technical info has been in the vidocs for Halo 4? I haven't watched. From what I've seen, there really haven't been any presentations or whitepapers, and it doesn't look like DF was able to get an interview with 343. What are they supposed to do? The early tech heavy analysis were largely based on devs being forthcoming about their technology, not screenshot analysis.
not really, previous one were just based on whoever wrote the article while he was playing through the games. I remember ND tweeted to check out the U2 tech analysis because DF was very close to what they have described. A lot of tech interviews comes after their tech analysis articles. This halo 4 one just feel very sloppy, if this is truly the first ground breaking technical marvel like they said, I think it deserve a lot more afford and tear down from chapter to chapter. I finished the game and personally I find it rather disappointing after being hyped by DF.
Billy Idol
Legend
That might be true Scott, that they did not get any help from devs this time. But it is one of the biggest console releases, and they could have used a bit more time...imo. See the Killzone 2 special, where they went through the game lvl by lvl and analized everything (comments by Alex Goo or something like this iirc) - this was a masterpiece of tec analysis.
This tec analysis includes some bla bla about new dev and franchise and than some few tec facts...why not take the time and go through it lvl by lvl?
I admit, that we don't know the circumstances, time schedule and all...I just hoped for more.
You're not getting it. You shouldn't be comparing Halo to UC, Kz, Far Cry, or Gears. You should be comparing it to Halo.
Your constant mentioning of the shadows is silly when they work similarly to shadows in previous Halo games (outside of some improvements shown in forge). Learn about these engines you criticize, maybe then some things will become clear for you.
Ground breaking technical marvel? Where do they say that? I got the impression that it can hang with the best of them, not that it's the next breakthrough like we saw with UC2.
Also I don't understand why you (or Billy) expect a level by level break down when they've only done that for one game. Looking back, the last two analysis articles covering LoU and Forza Horizon were also one page. Seeing how they hired additional writers to help Rich out, I doubt he has time to repeat what he's done for KZ2. I too would have liked more of a break down, but I'm not sure what more they could have covered without any involvement from the devs.
Look at the ODST analysis, that's way more thorough. Where's the latency and co op tests here?
Latency tests should be a standard feature in addition to resolution and frame rate checks. Disappointing that the NFS MW and Forza Horizon didn't have them either considering what was achieved in the devs previous games.
SedentaryJourney
Regular
The last three comparisons either favoured the PS3 or were dead even. I do think the reporting on certain issues is inconsistent, for example the sparks issue on the PS3 version of NFS:MW was mentioned earlier in the article as a bug but was summarized as a pared back feature. There's also claims in the comment section of motion blur being applied on the PS3 in instances where neither the 360 or PC version feature it, but this isn't mentioned in the face-off.
I've also found that the texture and LOD disparity between the PS3 and 360 versions of MOH:Warfighter is neither obvious nor clearly pointed out in DF's comparison shots, though it's one of the main points in their summary.
Overall though there aren't extreme cases of bias equivalent to Lens of Truth comparing Warfighter on PS3 to the Xbox 360 version minus the HD texture pack.
My view is that it's much easier to write a long analysis of a game where the developers publicly promote their technology as part of their sales pitch for the game at E3 and other events, as well as in interview and their own promotional videos. Naughty Dog, Epic and others are particularly interested in doing this. I haven't seen very much of anything for Halo 4 by comparison. I'd rather see them write a short analysis based on what they know to be true, or very close to the truth, rather than write a long analysis full of speculation based on visual impressions and be almost entirely wrong. To this point, is the analysis short because they didn't spend much time on it, or because they were only comfortable with the details they provided and nothing more?
The comments that say they should have done tests on split-screen play and controller lag are good points, but I think split-screen is probably more helpful for people purchasing the game for that feature, where controller lag is largely irrelevant if there is not another console release to compare to.
So I'm being silly when I pointed out the obvious? No one on the planet knows 100% inside out about this engine except for their lead engine programmer for obvious reason, but some stuffs don't take an expert to notice. Maybe you're the one who should simply admit the fact that there are no cascaded shadow maps for the environment in Halo 4.
Dominik D
Regular
There's this thing called reality. It's not always even-steven. What you call bias is a difference in quality of multiplatform games. Get a job at Sony, improve their toolchain and help developers tune their games towards PS platforms. It's that simple.
As for the quasi-technical observations (in case it matters: I played the game so it's not based on YT videos)
- shadows fading in the distance are much better than shadows popping-up but the distance at which it happens is disappointing; I was also wondering if it would be feasible to fade from correct shadows to black circular blobs as shadows in the distance - would it look good enough if fade happened at the reasonable distance from the player?
- particles in general and explosions specifically are super underwhelming; didn't really pay attention if they were bouncing on the z-buffer like it used to work in Reach but there were fewer particles and explosions just didn't have the OOMPH! i was expecting (this may be art issue but probably it was dictated by the GPU budget)
- enemy disintegration looks awesome
- character details and animations have improved significantly: faces finally look acceptable and skin SSS in the cutscenes interacts wonderfully with lighting in the environment
- I like the lighting in general: it's less flashy than it used to be which paired with the art style may finally (but slowly) put down the "canned dude is shooting pink lasers to yellow fairies" BS
- light shafts are overused but I can see why they did that: they're faked but they do give the sense of volume to the light and environment which is not necessarily the case with other implementations (e.g. Bulletstorm had two suns for double the god rays awesomeness but the way they looked was IMO shit)
- I really like the voxelized loading screens
- menus and navigation are confusing as usual
Also: kudos to Laa-Yosh!
Art style-wise (yeah, non-technical) I don't like some of the enemy designs 343i created. But that's completely subjective.
BTW: Corrinne Yu got patent filed on Single Pass Radiosity From... and I can't decipher the remaining two words. Depth Buffer?
- Status
Similar threads
