CosmoKramer
Newcomer
I know very little about "Next-gen graphical pet peeves", given that there are no next-gen games (or consoles) yet. Current gen has some things that could be improved though.
Next-gen graphical pet peeves
- Thread starter Bohdy
- Start date
I know very little about "Next-gen graphical pet peeves", given that there are no next-gen games (or consoles) yet. Current gen has some things that could be improved though.
<Snicker!>
He is obviously referring to the Xbox 360 and PS3. When PS2s stop selling like hot cakes at IHOP's Sunday brunch we can reclaim the "next gen" moto for the 4th 3D generation
HDR On the average, I think it hurts more than it helps. There are some cases where it's done with good taste but more of those where HDR is used just because "you are supposed to". I get that it is the latest gimmick, and everyone and their uncle are doing it, but does every game have to look like it is a soft porn film shot through a vaseline-smudged lens? Yes, the bright spots are glowing, I can see that, enough already, seriously!
Hmm, excuse me if you weren't there at the time, but you guys were big bloom-offenders with Fable 1. Glowing chickens??
Nice to read that some devs are also concerned about the flickery shadow problem, though.
TurnDragoZeroV2G
Regular
Hmm, excuse me if you weren't there at the time, but you guys were big bloom-offenders with Fable 1. Glowing chickens??
Nice to read that some devs are also concerned about the flickery shadow problem, though.
That doesn't mean he had any say in the decision if he was there!
My personal peeves have been repeated alot by now. Shadowing is a big one, both lack of shadowing and quality. I also dislike having such mixed bags with lighting. Very nicely lit worlds, for the most part, but a bunch of objects that are lit in such a way that it makes them appear as if they aren't even part of that world. So, yeah, Lighting, shadowing, animation, and especially the plastic appearance to materials that clearly aren't plastic.
cloudscapes
Regular
Crappy death animation.
The later-gen precanned animations were often very good, even though they didn't fall onto furnature properly. It's still much more pleasing to see than the gelatine-filled dolls that either plop or float down when killed with ragdoll. Endorphin FTW!!
Abusive HDR and bloom (I know theyre two different things).
Yes darker darks and brighter brights is "realism", but at the moment monitors don,t have the dynamic range to make it work properly, unless if it's used in a subtle way. Unfortunatelly not all devs want to be subtle, and we end up moving through a blurry ultra-bright and ultra-dark mess.
http://www.vgcats.com/comics/?strip_id=224
EDIT:
VOICE ACTING!!!
PLEASE!! More voice acting like HL2 and the guy in Heavenly Sword, it's so worth it. Lump quality dialogque in here too. I'm tired of even high-profile games that feature bland voices that do little more than say what's on the lined paper in a perfect rehearsed fashion!
In real life, people pause, "hmmm..", grunt, and breathe when they talk. They don't just emit a string of words and then put a period at the end! All it takes is a better director and a bit of patience during recording (and arguably slightly better actors).
The later-gen precanned animations were often very good, even though they didn't fall onto furnature properly. It's still much more pleasing to see than the gelatine-filled dolls that either plop or float down when killed with ragdoll. Endorphin FTW!!
Abusive HDR and bloom (I know theyre two different things).
Yes darker darks and brighter brights is "realism", but at the moment monitors don,t have the dynamic range to make it work properly, unless if it's used in a subtle way. Unfortunatelly not all devs want to be subtle, and we end up moving through a blurry ultra-bright and ultra-dark mess.
http://www.vgcats.com/comics/?strip_id=224
EDIT:
VOICE ACTING!!!
PLEASE!! More voice acting like HL2 and the guy in Heavenly Sword, it's so worth it. Lump quality dialogque in here too. I'm tired of even high-profile games that feature bland voices that do little more than say what's on the lined paper in a perfect rehearsed fashion!
In real life, people pause, "hmmm..", grunt, and breathe when they talk. They don't just emit a string of words and then put a period at the end! All it takes is a better director and a bit of patience during recording (and arguably slightly better actors).
Last edited by a moderator:
That's why we have tone mapping. Keep in mind that also are not devs (as programmers) deciding these things..but artists.
1. Shadows. Covered ground.
2. Imperfect interaction/collision/clipping between complex, skinned models. One of the big things I thought would change with next-gen.
3. Visible LOD transitions and differences. Be it hugely obvious mipmaps or level geometry morphing into the proper shape as I get closer, if it's not transparent it's a problem.
2. Imperfect interaction/collision/clipping between complex, skinned models. One of the big things I thought would change with next-gen.
3. Visible LOD transitions and differences. Be it hugely obvious mipmaps or level geometry morphing into the proper shape as I get closer, if it's not transparent it's a problem.
scooby_dooby
Legend
And Cloudscapes, our resident Ubisoft dev
.On the shadowing, the shadows in MAss Effect used to pretty terrible, but in the new screens look extremely soft and smooth. We haven't seen them in motion, but any theories on how they are accomplishing this??
For example:
http://xbox360media.gamespy.com/xbox360/image/article/768/768658/mass-effect-20070227050237296.jpg
What you're talking about is not HDR.
HDR makes much more realistic lighting, reflection mapping and a few other things possible. Including the bloom effect that you're actually talking about.
non unified lighting would be the biggest thing.
also like most ppl have mentioned, overused bloom, (at least half the games suffer from it, this is easily solved by using a separate bloommap + materials ).
lack of solidity in the scene, 2d plants eg speedtree is a nono
repetitive textures
restricting where the player can go, eg instead of placing some convenient fallen pillar or log in front of an area, let the player enter it ( even if the area is just a infinite flat plane )
also like most ppl have mentioned, overused bloom, (at least half the games suffer from it, this is easily solved by using a separate bloommap + materials ).
lack of solidity in the scene, 2d plants eg speedtree is a nono
repetitive textures
restricting where the player can go, eg instead of placing some convenient fallen pillar or log in front of an area, let the player enter it ( even if the area is just a infinite flat plane )
I have the impression that it's still not well understood in the community..but I might be wrong.
30 fps + motion blur in a fast moving camera games are simulation/motion sickness inducing formula to me. That combination doesn’t look smooth for me, more like blurry jerky pictures. Like in a movie where the camera moved too fast for its own good. From the looks of things there are going to be less 60 fps games compare to last generation. And I thought 60 fps is barely adequate. I'll look forward to a generation where a stable 120 fps is a mandatory checklist for release.
ShootMyMonkey
Veteran
Or even all that well-known. A number of rendering guys I know at various studios have never even heard of it, let alone understand the concept.
Otherwise, among those who are aware of it, they consider bugs like the light bleeding to be unacceptable (which is fair), so I don't think the idea often gets much further than being written on a whiteboard.
Apart from the light bleeding the only other 2 issues I have with it are:
1) prefiltering has a linear costs compared to PCF quadratic costs, but PCF runs on pixels you need
to shade, while prefiltering runs all over your shadow maps and often shadow maps pixels count is quite higher than frame buffer pixels coiunt
2) it consumes more memory (expecially at 32 bit!) compared to common shadow maps and memory is a precious resource, especially on consoles.
1) prefiltering has a linear costs compared to PCF quadratic costs, but PCF runs on pixels you need
to shade, while prefiltering runs all over your shadow maps and often shadow maps pixels count is quite higher than frame buffer pixels coiunt
2) it consumes more memory (expecially at 32 bit!) compared to common shadow maps and memory is a precious resource, especially on consoles.
I agree with this one, but I think we're all still figuring out optimal tone mapping. The problem is many fold though:
- We can have artists tune it for our tv's in the office, but everyones tv at home will look a little different so our tone mapping choices will likely never quite be optimal.
- Worse yet, some tv technologies don't handle low brightness details very well, so even though our intension is for example to darken the main scene when there is predominantly bright sky visible, the end result on some tv's is that the darkened areas appear as almost pitch black on some tv's.
- Compounding this problem is that most tv's are set to be uber bright from the factory, some people will adjust the settings, some won't. Which do you cater to?
-Typical tone mapping algorithms will use the entire scene to determine a new tone, which isn't really correct. What you really need to be sampling is the area the users eye is looking at, and the tone mapping should be weighted more in favor of that region of the screen. On tv broadcasts the camera man will usually do this manually, but what can you do on a console game? There is no easy way to know where the users eye is looking. In an ideal world a console would have a built in "eye toy" type device to know where the users eye is looking, then we could adjust tone based on exactly the region they are looking at.