I have to admit that they were impressive titles in their own way. But I didnt get the same wow factor I got from some PS3 exclusives. Alan Wake was the closest thing to Uncharted 2 but watching the two together you see lots of compromises done on Alan Wake like in character modeling, resolution, texture detail etc. If it wasnt for the scale and enemy numbers I wouldnt have considered Reach an impressive looking title. But that could be just me
Digital Foundry Article Technical Discussion Archive [2011]
- Thread starter user542745831
- Start date
-
- Tags
- digital foundry
- Status
As far as Uncharted goes, I felt 1 in 2007 was an impressive showcase for quite awhile on PS3, but by the time 2 arrived I felt there were more advanced games. At the moment, Guerrilla Games and Santa Monica have amazed me with what they've been able to accomplish. In any case, I think RDR and GTA IV are the most impressive console games overall, so I've never had the impression the PS3 was the stronger system based off linear games like Uncharted or Killzone because the 360 won those sandbox match ups w/o much trouble.
(((interference)))
Veteran
I agree, but Remedy is a far smaller team (~30 people) and their tech dates back to like 2005 so stuff like the animation is sub par, as for texture detail the whole game was supposed to be open world (which is evident by the far off vistas in the missions which you actually travel to in the game and by the driving sections) which means they can't go to town on the textures.
With Uncharted, I haven't played the 2nd one besides the demo (waiting on a friend to lend me his copy) but from what I've seen/played I don't find it better looking than say KZ2, which is almost like playing a CG movie - Uncharted 2 still looks like a game in that sense. GoW3 is also stunning, but its a bit difficult to directly compare since its 3rd person with cinematic camera.
Reach is impressive due to the scale of the environments, detailed weapon and character models, fantastic effects and the great AI (at least for enemies) which the game throws at you by the dozen (I rember seeing a Firefight game with 50 elites on screen at once). Also it has saved films (though I don't know what kind of overhead that costs).
It also does 4 player splitscreen (and you can go online) which is quite the achievement.
That being said their tech for Halo 3 was pretty weak so they had to effectively start from scratch with Reach rather than it being the second iteration of an already capable engine (like with Uncharted).
Perhaps we should wait to see how the tech heavyweight MP releases like RAGE or Crysis 2 perform on each console to see if they buck the trend of the 360 performing better. From what little i've seen Skyrim also looks promising.
Last edited by a moderator:
It would seem that Sony is banking on it's exclusives to push it machines.
I think the reason polyhony digital was given so much time on the car models is so that when PS4 comes along, the current models can be used but with higher LOD in gameplay. Imagine playing the photo travel models in-game? And playing killzone 4 at true 1080p at 60fps?
Also they might be using their engines on games for PSP2.
I really hope for a PS4 with upgraded cell and something similar to a nvidea GTX560. We would also get BC.
I think the reason polyhony digital was given so much time on the car models is so that when PS4 comes along, the current models can be used but with higher LOD in gameplay. Imagine playing the photo travel models in-game? And playing killzone 4 at true 1080p at 60fps?
Also they might be using their engines on games for PSP2.
I really hope for a PS4 with upgraded cell and something similar to a nvidea GTX560. We would also get BC.
It's a risky strategy, but it could be the right one. I'm just not sure - I predicted (as one of many probably) that when GT4 had Photomode and featured somewhat better models in Photomode that were pared back in gameplay, that this would be a big help for when GT5 was produced. I particularly also predicted that they could probably reuse some of the rendering tech (motion blur and such) directly, and that this could result in GT5 being released pretty shortly after the PS3 was launched. That has turned out a bit of a double-edged sword though, hasn't it?
Right now my worry is that a) producing a model on that level could be done much easier and faster in the future, and b) you may get different requirements by then for instance for tesselation and what not, and other breakthroughs (3D textures?).
But I guess that with the level of fidelity that the models have now received, yes, they probably won't stand out negatively on a next-gen system - the quality of the photomode car models is really outstanding.
Right now my worry is that a) producing a model on that level could be done much easier and faster in the future, and b) you may get different requirements by then for instance for tesselation and what not, and other breakthroughs (3D textures?).
But I guess that with the level of fidelity that the models have now received, yes, they probably won't stand out negatively on a next-gen system - the quality of the photomode car models is really outstanding.
Good point on a). I dare say 3D photography will provide a means to automate high-fidelty models in future. It may not produce efficient models, either needing tidying up or just use up more resources than necessary, but then future hardware should be able to cope with a little excess. Regards b), it should be possible to transform and and export from any 3D mesh into newfangled formats, just as normal maps can be derived from models.
Don't have too high expectations for automated modeling approaches, the geometry they produce is rubbish. Shading artifacts on polished reflective surfaces like car bodies are unacceptable, UV mapping is problematic, polygon efficiency is troublesome...
Most tools and workflows in the CG industry today are very well polished and studios need some radical changes elsewhere to abandon their methods. Like, move to fully voxel based rendering, replace texture mapping with something, and so on...
Most tools and workflows in the CG industry today are very well polished and studios need some radical changes elsewhere to abandon their methods. Like, move to fully voxel based rendering, replace texture mapping with something, and so on...
Lucid_Dreamer
Veteran
It will be interesting to see the results of Infamous 2. If they achieve 720p at near 60fps (sandbox game), I wonder how well that point will hold up. The first Infamous was no slouch at 720p and around 45fps. The textures were still equal or better than the sandbox/open world games out now (and that was with a multiplat engine).
I never understood why game like RDR or GTA IV are the most impressive console games for someone
I mean, are excellent but close to GOW 3 or the others major ps3 esclusive not seems so impressive... and only because aren't 'linear' game? There are even Two Worlds, almost ignored, more impressive to me and on the ps3 it's fantastic (not compromise how Rockstar game), by the way I continue to found more impressive the sony exclusive. However will be curious how The getaway 3 like on the ps3 (if coming out) and if this theory of the supremacy of the sandbox game vs linear will continue to exits imho.
Last edited by a moderator:
I never understood why game like RDR or GTA IV are the most impressive console games for someone
I would imagine that some are more impressed by sandbox games over the typical linear design due to the random nature of a sandbox game. You have greater control over what's being rendered and pushed on the screen in a linear game where in a sanbox game the designer has less control over what direction the player will be looking and what will be rendered on the screen at that time.
I can easily see how some would consider GTA4 or RDR to be the most technically impressive games this gen.
Regarding Sony's exclusives, I really think their studios do a great job of maximizing the PS3's strengths while designing around it's weaknesses. This is what makes them look so great IMO and really that's exactly what an exclusive game should be trying to do.
Still?! What's the hold-up? 3D scanning should be as easy as 2D scanning these days!
3D scanning generates a multi-million point cloud. You can mesh it but it's still a multi-million poly mess, mostly unusable for anything other than extracting normal maps from, but you still need an organized mesh for UV mapping, animation and such tasks.
There are many well developed methods to build good models on top of scanned data but automatic approaches are still rubbish and the manual work requires skilled personnel and time. Just because scanning gets cheaper and faster, it's still not going to replace them.
There are many well developed methods to build good models on top of scanned data but automatic approaches are still rubbish and the manual work requires skilled personnel and time. Just because scanning gets cheaper and faster, it's still not going to replace them.
By the way I did a lot of research about 3D scanning, mostly the portable laser or digital camera based ones and they're all quite mediocre.
I'm not sure what freeware apps can open OBJs but the trial versions of both Maya and 3ds max should be able to import them. Here are some example files:
http://www.artec3d.com/gallery/3d-models/
They're good as a reference, but you still need a skilled modeler to create anything usable. Some of the laser scanned stuff is also far too rough even for normal maps, look at this:
http://www.xyzrgb.com/lightbox_images/helmut_facescan.jpg
Then there's the issue of object scale: a complete human body requires very, very high end equipment and even there it takes several seconds, while most people just can't stand still even for that short amount of time...
I'll perhaps get back to this later if there's enough interest.
I'm not sure what freeware apps can open OBJs but the trial versions of both Maya and 3ds max should be able to import them. Here are some example files:
http://www.artec3d.com/gallery/3d-models/
They're good as a reference, but you still need a skilled modeler to create anything usable. Some of the laser scanned stuff is also far too rough even for normal maps, look at this:
http://www.xyzrgb.com/lightbox_images/helmut_facescan.jpg
Then there's the issue of object scale: a complete human body requires very, very high end equipment and even there it takes several seconds, while most people just can't stand still even for that short amount of time...
I'll perhaps get back to this later if there's enough interest.
Yeah, I thik that RDR, GTA4, JC2 and AC:B are the most impressive technically titles this gen. But linear games shouldn't be compared to open world ones.
For example: GOW3 is the graphics champion of slasher: Dante's Inferno has better framerate (60fps compared to ~45) but graphically it's way worse, Castlevania is close graphically but it runs in almost half the framerate.
For TPP shooters: both Uncharted games are graphically and technically better than Gears of War 1/2, Alan Wake has very nice lighting but res is low and game tears pretty bad in some places. I'm not even going to mention something like Army of Two etc.
GeOW 3 is bringing very nice changes (new lighting, cascade shadows, foliage etc.) so we'll see how it compares with Uncharted 3For FPP shooters: Killzone 2 still looks better than every other shooter on consoles, and Killzone 3 with graphical improvments is coming. Halo Reach is probably closer to linear game than open world one, it may be doing some nice thing technically (lots of enemies, good Ai, full res alpha buffers etc.) but KZ2 still looks better (but it's of course much more linear in most levels, some are comparable). Currently there are no other competitors, but games like C2 and Rage are coming.
All of this is just my opinion of course
I never understood why game like RDR or GTA IV are the most impressive console games for someone
They provide long distance visual scope combined with 24 hour lighting and smooth framerate, all on old console hardware. GTA additionally adds large amounts of mayhem to the mix and is able to somehow manage it all relatively smoothly. I'd personally add Assassins Creed 2 to the list as well. You can't bake your way out of visual issues if you have full 24 hour lighting, you have to make it all work. Getting stuff like shadows working at all times of day (both visually and computationally), getting textures to look nice at lunch time and under moonlight, making sure tone mapping gets the right look under bring sun and limited sun, etc, I mean it's a whole different ballpark of complexity. While the visual scope of a game like Halo Reach is real nice, it's fixed time of day makes it much less impressive, at least to me anyways.
I figure you don't actually spend any time tweaking that million point mesh, you just use it as a reference to get a killer normal map and to get correct scale on body parts for the mesh the artist is actually working on. So like they put the uber mesh next to the one they are actually working on maya/max so that they can get size and scale correct on arms, fingers, head, etc, for games that it matters on like sports games, or if you are modeling a known actor, etc...
Joker, I'm a lead modeler at a CG studio, we do this every day although most of our high res stuff is sculpted from scratch in zbrush, but we've dealt with scanned data too. We actually model right on top of the high res geometry and make the new model transparent with visible edges. Resurfacing, retopo, there are many names for this. My point is that it still takes significant work to get results, even for 10-20k game characters and such. And there are no proper automatic tools to replace human intelligence.
The main advantage for getting a real person's scan is that life provides more interesting varieties of shapes and forms and proportions, many of which an average artist wouldn't ever think about, and yet the results are - obviously - realistic. It's also the most reliable way for celebrity likeness, although even there it may bring surprises - a good director of photography and some clever lighting can create a completely false image of an actor sometimes, covering up unpleasant facial features or asymmetries or such
Also, movie directors used to prefer maquettes because they were easier to look at in physical form; but nowadays we're seeing a reverse workflow. For example the costumes in Tron, they were built in the computer on top of each actor's full body scan, then printed in 3D using CNC machinery to create molds for the final pieces. It was faster and cheaper to design and iterate this way
But most scan data, with the exception of super expensive stationary devices, is very noisy and full of holes, so usually you can't even use it for normal maps either. Then there's the issue of resolution, even 1mm is usually not good enough for fine little details and precise forms (for hard surface stuff). Think about it like voxels, anything smaller than the voxel size gets lost or isn't preserved properly.
although most of our high res stuff is sculpted from scratch in zbrush, but we've dealt with scanned data too. We actually model right on top of the high res geometry and make the new model transparent with visible edges. Resurfacing, retopo, there are many names for this. My point is that it still takes significant work to get results, even for 10-20k game characters and such. And there are no proper automatic tools to replace human intelligence.The main advantage for getting a real person's scan is that life provides more interesting varieties of shapes and forms and proportions, many of which an average artist wouldn't ever think about, and yet the results are - obviously - realistic. It's also the most reliable way for celebrity likeness, although even there it may bring surprises - a good director of photography and some clever lighting can create a completely false image of an actor sometimes, covering up unpleasant facial features or asymmetries or such
Also, movie directors used to prefer maquettes because they were easier to look at in physical form; but nowadays we're seeing a reverse workflow. For example the costumes in Tron, they were built in the computer on top of each actor's full body scan, then printed in 3D using CNC machinery to create molds for the final pieces. It was faster and cheaper to design and iterate this way
But most scan data, with the exception of super expensive stationary devices, is very noisy and full of holes, so usually you can't even use it for normal maps either. Then there's the issue of resolution, even 1mm is usually not good enough for fine little details and precise forms (for hard surface stuff). Think about it like voxels, anything smaller than the voxel size gets lost or isn't preserved properly.
Yeah I know, figured I'd pick your brain a bit
Wasn't sure if anyone was making more detailed use of these high res scanners for games as I'm out of the loop now.
I was imagining a system that'd take a high density 3D model and 'vacuum form' a mesh around it. AI techniques would identify essential creases and map key points. Similar to scanned-art to vector-art automated tools. Of course referencing those tools, they don't tend to create clean solutions. But surely there'd be a way to have a human brain select key data on the mesh and the computer to connect the dots and fill in the gaps. In the case of scanning cars for example, it's the same few basic forms in essence with variations on scales and postions. A human user should be able to train an AI system to produce a uniform interpretation from any scanned car. Although if the noise is bad as you say, then the issues with clean up or noise filtering will probably destroy the fidelity, and I can see something as engineered as a car benefiting from ground-up construction in that case.
It's disappointing that 3D data capture and manipulation isn't progressing anything like 2D data has. I can use a digital photo or scan to get artwork into any document or application, but getting a 3D mesh needs someone to build it.
It's disappointing that 3D data capture and manipulation isn't progressing anything like 2D data has. I can use a digital photo or scan to get artwork into any document or application, but getting a 3D mesh needs someone to build it.
- Status
Similar threads
- Locked
- Deleted member 11852
- Gaming Technology
- Locked