*spin-off* Next Gen Gameplay and/or Graphics Differentiators

[Silent_Buddha]

Silent Buddha I have to say that reading this post has been a revelation for me and a bloody awesome breath of fresh air, that at least one person on the boards i post on finally "gets it" with regards to traditional RPG systems and twitch/skill-based "dudebro" mechanics.

I'm with you all the way brother on every point, and I thank you sincerely for helping me to realise that i'm no longer alone in the gaming universe with my views.

I've been banging on that stuff for quite a while now, and I'm not the only one either. Although granted it's probably banged on about more in the PC forum than the console forum.

They don't hav to be RPGs though. Take a game like Warhawk. You could have a character who's an ordinary soldier with fairly wild aim. You could invest points in aiming to improve the quality of his autoaim. Or you could invest points in something else maybe . Or even have a difficulty slider perhaps so those with awesome skills can do their thing and those without can enjoy the game without needing to learn new skills. Yes, there'll be issues in balancing, and maybe you'd separate those with real skill from the amateurs or something. The underlying point though is that I feel games should be smarter and read that many gamers haven't got perfect control and when the avatar doesn't act as they want it to, it annoys them. Platformers are another good one. I've seen people play LBP and get frustrated with the mechanics. I think it'd be possible to add 'driver aids' to make the control easier.

Nod. I just used RPGs as an example, as that's a personal pet peeve of mine that most so called RPGs aren't really RPGs anymore. But that what most people consider to be an RPG now days is just a FPS/TPS/other action game type with a light veneer of RPG like stuff thrown on.

But you're correct there are many other ways a developer could remove the twitch/reaction/physical skill component from the equation to allow for more casual players to enjoy the games. Difficulty scaling is obviously one method, but IMO a flawed method. As even casual players will feel a bit let down if they feel their enemies are just purposely suiciding on them or that the developers are humoring them. Things that can allow casual players to defeat difficult content (stat based/turn based RPGs for instance) or give them the illusion of greater skill (your example) are a couple ways to tackle this without making the player feel like they are purposely being pandered to.

The problem with that is you then run into lots of resistance from the "traditional" "core" console player who is looking for the next action fix that demands levels of skill beyond many of your typical consumers.

I posted that in this thread. That's the context that you missed.

It would have been nice to see more games use Euphoria this gen other than RDR, GTA4, LA Noir, SWTFU, SWTFU2, but with the increased power of nextgen and a collaboration by Sony and/or MS to handle the licensing & business issues, this could be a reality for more than just a handful of games nextgen.

Sure it's always nice to see good tech implemented. RDR, GTA4, LA Noir, SWTFU, SWTFU2 all had serious drawbacks in other areas that it would have been great to see addressed. But time, hardware, and budget constraints means that every single one of those had significant and glaring compromises and limitations.

That isn't going to change significantly in the next generation. Developers are still going to have to decide between feature X and feature X+1...feature X+1000 whenever they make a game. Whether due to technical limitations of the hardware (which most certainly aren't going away), budgetary limitations (also not going away), or time limitations (again not going away).

The X360/PS3 were a significant bump over the Xbox/PS2, and yet there is no universal set of "features" that are implemented in all or even most of the games. The same will hold true for Xbox/PS next.

Just like I'm hoping for games to adopt tesselation of geometry, real time lighting, better physics, better texture resolution and techniques (POM for example), destructable environments (when it makes sense), natural human motion (when it makes sense), more realistic shadowing, significantly better AA or rendering techniques removing MANY of the rendering artifacts that exist in current gen games, and the list goes on and on and on. I could sit here for the next 2-3 hours typing out the technology and features I'd love to see implemented as graphics differentiators.

But one things remains true through all of that. It will be impossible to implement all of them. So you wanting universal Natural Motion is no more "correct" than me wanting universal terrain and scene tesselation. But the fact is that not all games will want to use or focus on either of those at the expense of technologies they deem crucial or key to the game they are developing.

In other words, just as everyone has been trying to tell you. Natural Motion is something good to strive for, but it is not, and never will be universally better than any other technology that can possibly be implemented because all developement projects will focus on different key technologies that they deem applicable or necessary for their vision of their game. Because as mentioned by me and many others; Technical, monetary, and time limitions are not going away just because the hardware is more capable, or budgets might go up, and there's certainly no way to get more time.

Regards,
SB

 

[TheChefO]

In other words, just as everyone has been trying to tell you. Natural Motion is something good to strive for, but it is not, and never will be universally better than any other technology that can possibly be implemented because all developement projects will focus on different key technologies that they deem applicable or necessary for their vision of their game.

Oblivion
Skyrim
Gears series
Halo series
Resistance
Killzone series
Fallout series
Bioshock series
Batman AA/ACCOD
BF3
GTA4
Red Dead
LA Noir
All sports games
Uncharted series
Assassins Creed series

And any other games which feature human-like characters would benefit from a Euphoria like approach.

It's hardly a niche corner case feature. And with the intended consumer of games being humans, replicating natural motion, behavior, and interaction for humans (and human-like characters) in games should be a higher priority in reproducing a believable world than most any other periphery IMO.

For games that don't feature human-like characters or that are more cartoonish in nature, sure, ignore the feature. But there are fewer and fewer cases which that will be the case moving forward.

For the ones that do, the sacrifice in processing resources is worth it. And should be standard.

As for the time it takes to integrate, I'm sure the process isn't optimal at this point, but that's why MS/Sony have deep pockets... to help promote their platforms. However long it is though obviously isn't so prohibitive that it isn't worth pursuing given the fact that it is shipping in a half dozen games on current gen hardware.

Agree to disagree if you wish, but IMO, it's like contemplating if self shadowing is worth it, or if texturing is worth it.

 

[Shifty Geezer]

Euphoria is just one piece of the physics engine in RDR. I don't believe they use it on the horses at all, except for maybe when they ragdoll in death. Try to find some slow-mow footage of people getting shot off of their horses etc. They do use it in the player movement, for controlling the placement of his feet, and in the inertia as you turn while running. That's just a guess.

I'd have thought that, but the first YT vid I watched when looking for horses had the player moving like any other game. When turning the player's character rotates without adjusting feet position. So Euphoria is used in part, but doesn't break the gameplay. I guess it's more called upon for spot events in the same way we switch to ragdolls.

It would have been nice to see more games use Euphoria this gen other than RDR, GTA4, LA Noir, SWTFU, SWTFU2, but with the increased power of nextgen and a collaboration by Sony and/or MS to handle the licensing & business issues, this could be a reality for more than just a handful of games nextgen.

It would have been nice to see better animation, whether Euphoria or not, this gen, where it wouldn't interrupt gameplay. Would also have been nice to see better AI, lighting, shadowing, IQ, etc. There are many reasons that prevent devs for incorporating the features we'd like as they have to compromise on finite hardware. Something has to go. It'll be nice if a collaboration by Sony and/or MS with a whole load of middleware vendors would create fabulously easy and versatile tools for devs to use, not just for animation, but in business terms it's hard to justify, and in development terms it's extremely questionable if such a product could exist and be workable across a lot of different engine. If there's only UE4 and CryEngine next-gen then that simplifies things, but then we'll lose a lot of system power to engine overhead. The more flexible the engine, the less efficient it is.

List of games

You've already posted that list of games. How does posting it a second time change the argument or present a different case so those who didn't agree with you last time will see things differently this time?

And any other games which feature human-like characters would benefit from a Euphoria like approach.

Why have you not replied to any of the comments about gameplay that doesn't work with realistic animation? Just because a game uses a human character, doesn't mean the gameplay can cope with the slow-down issues of realistic animation. And if you are going to compromise like RDR seems to with its rotating-on-the-spot in the vid I saw, than how much benefit is there really to physics based animation versus the costs? Could the resources yield better returns elsewhere? That'd be answered case-by-case. For open, realistic worlds, having a populace that moves realistically would be valuable, as would having them behave realistically and having them not all look like clones (I'd argue the latter point is more important to preserving the illusion, but I dare say that's subjective). For many action games, the lack of realistic animation doesn't seem a problem to the experience, whereas slow, clunky controls would, so it makes sense in shooters and fighters to steer clear of behavioural physics. At which point it becomes a niche feature that serves only some genres, so is it really worth building that into the whole system SDK? Or mandating it for games? Even some games - devs would really not appreciate if in the prototyping stages they have to work on the siple animation gameplay they want and a slower behavioural animation in case when they submit the console company they decide this one should be one of those with behavioural animation. Console companies telling devs and publishers what to make and how to make it doesn't go down well. One of the reasons Sony started so well in this biz was Nintendo was heavy handed whereas Sony was much more open.

 

[Acert93]

Significant Next Gen Difference Maker Graphically: High Detail Facial Scans and body scans (3D and Texture via Kinect 2) that is used in games: You become the star/hero of many games. You could even have variable personas for different games where your hair style and wardrobe can be altered as well as some voice modulation (one game you could use a clean cut version of yourself; another game a version with piercings, died hair, tats, etc; one version a freakish body builder and another a slothly fat person and so forth).

 

[TheChefO]

It would have been nice to see better animation, whether Euphoria or not, this gen, where it wouldn't interrupt gameplay. Would also have been nice to see better AI, lighting, shadowing, IQ, etc. There are many reasons that prevent devs for incorporating the features we'd like as they have to compromise on finite hardware.

It's your right to express which of the above is a priority to you and make the case to others. I'm making the case that a Euphoria-type system should be a standard which is invested in by the platform holder to be present in the SDK available for all devs to use because in the long run (nextgen) we will see more games replicating a believable world than we did this gen and the gen before. It helps MS/Sony to showcase future games using the technology as "this is our standard on our platform, the games you play will have believable worlds with believable characters and this is the standard bar". Devs could take it from there, build on it, bypass it using their own custom tools, or ignore the believable world game-type.

I think everyone on this board understands finite hardware. The point of this thread IMO is: "with the increased hardware spec coming nextgen, how would you allocate the processing budget?"

With that, I say take a chunk of the new CPU, and leave a space for Euphoria-type animation/AI/physics system. The reason for bringing up existing games this gen which use the system is to show that it isn't an impossible or unreasonable case to make.

You've already posted that list of games. How does posting it a second time change the argument or present a different case so those who didn't agree with you last time will see things differently this time?

The list is of popular games which are aiming for more believable worlds which Euphoria helps. The reason it's listed multiple times is nobody has agreed/disagreed with the list's intention of representing the most popular games this gen which are aiming to represent believable worlds which is key to the point of investing time/money to having Euphoria developed into the sdk.

Take look at the list from last gen (mostly ps2 games given the dominant nature of the platform) and you'll see a lot of the big franchises were not aiming for realism.

Jack and Daxter series
SlyCooper series
Crash bandicoot
Kingdom Hearts series
Ratchet & Clank series

And compare again to popular ps1 games and you'll see the trajectory is toward more realism/believable worlds.

Using Euphoria for the game doesn't mean it would have to be used for the main character. In my original list, there are cases where the main character control would be compromised if driven with Euphoria, but the rest of the human/human-like inhabitants of the game world would benefit from such a system as Euphoria, even if the main character would not.

Why have you not replied to any of the comments about gameplay that doesn't work with realistic animation? Just because a game uses a human character, doesn't mean the gameplay can cope with the slow-down issues of realistic animation.

And if you are going to compromise like RDR seems to with its rotating-on-the-spot in the vid I saw, than how much benefit is there really to physics based animation versus the costs? Could the resources yield better returns elsewhere? That'd be answered case-by-case. For open, realistic worlds, having a populace that moves realistically would be valuable, as would having them behave realistically and having them not all look like clones (I'd argue the latter point is more important to preserving the illusion, but I dare say that's subjective). For many action games, the lack of realistic animation doesn't seem a problem to the experience, whereas slow, clunky controls would, so it makes sense in shooters and fighters to steer clear of behavioural physics.

see above

At which point it becomes a niche feature that serves only some genres, so is it really worth building that into the whole system SDK? Or mandating it for games?

As I said, believable worlds is a growing game concept which is finding its way into more popular games as time goes on and hardware power increases. The niche case is for games that aren't aiming for a believable world. That doesn't mean it replicates reality 100%. It means the things one sees and interacts with in the game world are believable. Surreal perhaps, but convincing to the gamer. And in interactive environments, a big part of that is interaction, not just pretty, non-interactive graphics. Interaction with the gamer, and with the game world.

Even some games - devs would really not appreciate if in the prototyping stages they have to work on the siple animation gameplay they want and a slower behavioural animation in case when they submit the console company they decide this one should be one of those with behavioural animation. Console companies telling devs and publishers what to make and how to make it doesn't go down well. One of the reasons Sony started so well in this biz was Nintendo was heavy handed whereas Sony was much more open.

Devs didn't appreciate multithreading either. But time moves on and as progress takes place, there are bound to be bumps in the road and bitching along the way. But it's for the betterment of the industry. Having more devs on board and familiar with a Euphoria-type system will likely help to push the tech along even further as devs get their hands on it and come up with new ideas, workarounds, optimizations, etc.

For some, they will want a custom animation and control scheme for the hero character (GOW, Ninja Gaiden, maybe batman, etc). For others, they'll want a regular "joe" as the main character and Euphoria-type makes sense. But in either case, the rest of the characters in the game world would benefit from a Euphoria-type scheme.

 

[Laa-Yosh]

High Detail Facial Scans and body scans (3D and Texture via Kinect 2) that is used in games: You become the star/hero of many games.

That doesn't really work like that...

Either it's a high quality face scan (already unlikely with the current Kinect - try to get a 2K resolution face texture with a VGA camera) that can't really move or do anything.

Or it's a low quality generic face model and a blurred texture that kinda looks like the player, from a certain angle, from a certain distance, but at least it can use the canned animations that someone did on the generic head. Think something worse than the custom Shepard heads in ME1.

If it was just so easy to automate, why would Naughty Dog hire artists to spend months on their ingame characters?

 

[TheChefO]

That doesn't really work like that...

Either it's a high quality face scan (already unlikely with the current Kinect - try to get a 2K resolution face texture with a VGA camera) that can't really move or do anything.

Or it's a low quality generic face model and a blurred texture that kinda looks like the player, from a certain angle, from a certain distance, but at least it can use the canned animations that someone did on the generic head. Think something worse than the custom Shepard heads in ME1.

If it was just so easy to automate, why would Naughty Dog hire artists to spend months on their ingame characters?

Right, but with a higher res kinect2, it becomes more possible.

AAs for ease of use in automatic the texture mapping to the character, that's where MS and devs come in! I don't think anyone would expect the end result to be picture perfect, but I think the idea of using real world objects in the virtual world games is a great concept and worthy of the investment. Especially given that the initial Kinect promo video featured as much!

 

[Shifty Geezer]

I'm making the case that a Euphoria-type system should be a standard which is invested in by the platform holder to be present in the SDK available for all devs to use because in the long run (nextgen) we will see more games replicating a believable world than we did this gen and the gen before.

Why Euphoria as standard an not anything else? there are lots of techs that could do with being supported, but they come at cost. Hence the reason the console companies mostly provide an SDK and let middleware serve the tools and engines market.

With that, I say take a chunk of the new CPU, and leave a space for Euphoria-type animation/AI/physics system. The reason for bringing up existing games this gen which use the system is to show that it isn't an impossible or unreasonable case to make.

As long as we appreciate that there is a cost involved, and likewise there will be a cost involved. That's something we cannot quantitfy. I can agree it'd be nice to see improved animation (and we have had animation improvements this gen) but I can't agree it should be mandated. Some devs may want to do their own thing, or go with something else entirely.

The list is of popular games which are aiming for more believable worlds which Euphoria helps. The reason it's listed multiple times is nobody has agreed/disagreed with the list's intention of representing the most popular games this gen which are aiming to represent believable worlds which is key to the point of investing time/money to having Euphoria developed into the sdk.

Okay, I never understood what you were getting at. You're saying that the common target of these consoles as denoted by most popular titles is believability, hence supporting that common standard out of the box gives devs working on such titles a leg up. Similar in principal to providing a stearing wheel controller out of the box to support racing games. I'm not sure it's wise to go anticipating where games will go or try to support one style more than another though.

Take look at the list from last gen (mostly ps2 games given the dominant nature of the platform) and you'll see a lot of the big franchises were not aiming for realism.

Hmmm. I don't think realism is the key. You also use the word believability, and these worlds were going for believable within their own ruleset. Bioshock wasn't really realistic, but it was believable. I agree that more fluid motion would lend credibility to the characters in these games, but I also point out that clones is a similarly serious problem. Are we to want to console companies to make engines that solve content duplication too? Where are the lines drawn as to what the console companies should provide for devs and what should be left for the industry to innovate? One big issue with a console-standard is a lack of opportunities for new ideas and thinking from other middleware providers.

Using Euphoria for the game doesn't mean it would have to be used for the main character.

Fair and valid point I concede.

And in interactive environments, a big part of that is interaction, not just pretty, non-interactive graphics. Interaction with the gamer, and with the game world.

Interaction on that level is way more than just behavioural physics. You need data formats for dealing with objects and materials and people. That's a hell of a lot of overhead and complexity that most games can do without. The difference between Gears where characters can crouch behind preset rectangular walls, and Gears where characters can adapt their positoin behind objects of different shapes including destructibel objects, is huge. So you'll be left with characters still confined by very obvious limits only slightly better animated in some cases (mostly less repetitive animation. Good animation alread exists but it repeats as 'captured death animation 3' etc.).

Devs didn't appreciate multithreading either.

Whole different thing. Multithreading is essential for all games as that's the way tech is moving. A feature like physics or AI being forced on devs is just a damned nuisance.

I don't disagree with looking forwards to advances in animation tech. It's this notion of mandating it, and having the software that powers it run by the console company or even a joint venture that doesn't fit with me. Behavioural animation is no more important in that respect than many other features. The only rationale would be to work on an XBox Engine or PS Engine as a common game framework offering all these pieces, which doesn't work for reasons mentioned before. There is no one-size-fits-all solution and cannot be unless we want to seriously gimp the results of our hardware. We're not even using the convenience of something like single code-base JIT compiled languages, by compiling natively to each platform. It should remain the developer choice what techniques they use and what they don't, while the industry will evolve techniques and tools to help enable that.

 

[Acert93]

That doesn't really work like that...

Either it's a high quality face scan (already unlikely with the current Kinect - try to get a 2K resolution face texture with a VGA camera) that can't really move or do anything.

Or it's a low quality generic face model and a blurred texture that kinda looks like the player, from a certain angle, from a certain distance, but at least it can use the canned animations that someone did on the generic head. Think something worse than the custom Shepard heads in ME1.

If it was just so easy to automate, why would Naughty Dog hire artists to spend months on their ingame characters?

I get what you are saying, and I was just tossing it out, but putting on a "consumer hat" for a moment "connecting" with the protagonist or antagonist is an important aspect of a game. There is the traditional route--awesome artists like yourself creating compelling, high quality, and surreal characters that effectively draw people's imagination.

But, if the point is to effectively draw people's imaginations then the art assets are just a means. Why could not a lower quality yet personable avatar not create a similar hook for some games?

Put concisely: For consumers there may be enough incentive to have lower quality models (in some games) if the inclusion of a digital representation of the player themselves in the game offers a more compelling experience.

I know, not what an artist wants to hear

I have already seen, first hand, in games that support avatars how my kids LOVE seeing their avatar in-game. e.g. Powerup Heros but they get really excited in Your Shape 2012 because they see themselves and think it is exceptionally cool.

Ps- I agree the current Kinect is light years away from this, this is why I mentioned Kinect 2. And if Kinect 2 is not light years ahead of Kinect 1 I see no compelling reason to get a Xbox 3 unless the internals are pretty advanced.

 

[AlNom]

Well, you already mention games that do use the avatar function (there's Guitar Hero 5 too ), so clearly a lower quality avatar can be acceptable provided the rest of the game art flows with that.
Using a higher quality avatar than what already exists would be similarly acceptable but only for the same types of games, no?

I mean we shouldn't really expect to see it in big budget "hardcore" titles such as CoD or other shooters etc unless the art is a throwback to Team Fortress 2 or Battlefield Heroes or even the new Epic game with shopping carts and zombies, where in these particular titles, it doesn't matter how goofy you look.

 

[Silent_Buddha]

It's your right to express which of the above is a priority to you and make the case to others. I'm making the case that a Euphoria-type system should be a standard which is invested in by the platform holder to be present in the SDK available for all devs to use because in the long run (nextgen) we will see more games replicating a believable world than we did this gen and the gen before. It helps MS/Sony to showcase future games using the technology as "this is our standard on our platform, the games you play will have believable worlds with believable characters and this is the standard bar". Devs could take it from there, build on it, bypass it using their own custom tools, or ignore the believable world game-type.

I think everyone on this board understands finite hardware. The point of this thread IMO is: "with the increased hardware spec coming nextgen, how would you allocate the processing budget?"

Just as long as it is never mandated, enforced, or pushed on devs. that's not a problem. We all have wishlists for things we'd like to see prioritized (my list is obviously different than yours). So too devs have a list of things they'd like to be able to do from which they pick the things they most want to use.

I think just about everyone can agree on that, but everyone isn't going to agree on what things should be a priority.

It's similar in some ways to the how spoken dialog in games evolved. It didn't happen overnight. It didn't even happen with 1 or 2 generations. And it still isn't universal in all games especially for all NPCs. Although for voice, it's more a data storage + how much money do you want to budget at recording people speaking (and in how many languages).

But that's pretty similar to what we're talking about with anything. It is always a tradeoff. And as long as developers are the ones making decisions with regards to that then no problem.

But if console makers start dictating game impacting features then it becomes unwieldy. Just like no console manufacturer mandates Spoken Voice for every game (much less every character or NPC in the game), they shouldn't mandate anything resembling natural motion.

Regards,
SB

 

[TheChefO]

Why Euphoria as standard an not anything else? there are lots of techs that could do with being supported, but they come at cost. Hence the reason the console companies mostly provide an SDK and let middleware serve the tools and engines market.

IMO, A Euphoria-type solution would provide a substantial bang for the buck due to the increasing number of game worlds aiming for believability/realism. Whether it is actually Euphoria or another solution isn't important, but I think the amount of research that has gone into the Euphoria solution is worthwhile to just buy them out and take it from there instead of trying to start from scratch for the sake of time.

For the sake of simplicity, when I say Euphoria, I mean a system which enacts a similar result in-game.

As long as we appreciate that there is a cost involved, and likewise there will be a cost involved. That's something we cannot quantitfy. I can agree it'd be nice to see improved animation (and we have had animation improvements this gen) but I can't agree it should be mandated. Some devs may want to do their own thing, or go with something else entirely.

The mandate is for a standard level of quality for character behavior (AI, physics, animation) in games which aim for a believable game world. This would be provided to the developers in the SDK. If devs choose to go their own route or modify the existing Euphoria provided in the SDK, that's fine as long as the game exhibits realistic behaviors for their world. So when a character bumps into another character, there is a response. Not the typical response which is roughly the same as running into a concrete wall. When a force is exerted on a character, the character responds in kind.

That's the mandate (for games with an intended believable world). And the SDK would be provided with the tools necessary to efficiently implement the solution. Obviously this would likely need a good deal of investment by the platform holders to get to this point of being efficient and easy to integrate, but worth it IMO.

Okay, I never understood what you were getting at. You're saying that the common target of these consoles as denoted by most popular titles is believability, hence supporting that common standard out of the box gives devs working on such titles a leg up. Similar in principal to providing a stearing wheel controller out of the box to support racing games. I'm not sure it's wise to go anticipating where games will go or try to support one style more than another though.

If driving games were constituting the vast majority of games sold and that trend were increasing, I'd say an integrated wheel would be a great idea!

I haven't broken down the actual percentages, but I think it's safe to say the vast majority of popular games fall into the category of intending to represent a believable world which have human or human-like characters.

This is likely to continue to trend in this direction as hardware gets more capable of representing a believable world. What game devs decide to do with that world is entirely open to them, but again, the one aspect which is almost a certain commonality is human or human-like characters.

Hmmm. I don't think realism is the key. You also use the word believability, and these worlds were going for believable within their own ruleset. Bioshock wasn't really realistic, but it was believable. I agree that more fluid motion would lend credibility to the characters in these games, but I also point out that clones is a similarly serious problem. Are we to want to console companies to make engines that solve content duplication too? Where are the lines drawn as to what the console companies should provide for devs and what should be left for the industry to innovate? One big issue with a console-standard is a lack of opportunities for new ideas and thinking from other middleware providers.

Bioshock was indeed a believable world. Would have been even better with more realistic action/reaction from characters.

Content duplication for multiple characters is another area of concern. While I agree that it would be worth looking into, the impact of believable character motion and interaction trumps character visuals IMO. Having said that, I don't see a Euphoria-type solution prohibiting a solution to generate procedurally created characters or character variables. Fallout did a good job with this.

As for this standard library limiting creativity for Euphoria-type solutions, I disagree. I think widespread exposure of developesr to the tech will encourage innovative solutions and or optimizations.

As I said, the intent isn't to lock in developers exclusively to Euphoria as it exists. Modifying it to expand beyond the current capabilities will take place, and for those that have better alternative solutions, they can continue to innovate and stand on their innovations which go above and beyond Euphoria.

Fair and valid point I concede.

:smile:

Interaction on that level is way more than just behavioural physics. You need data formats for dealing with objects and materials and people. That's a hell of a lot of overhead and complexity that most games can do without. The difference between Gears where characters can crouch behind preset rectangular walls, and Gears where characters can adapt their positoin behind objects of different shapes including destructibel objects, is huge. So you'll be left with characters still confined by very obvious limits only slightly better animated in some cases (mostly less repetitive animation. Good animation alread exists but it repeats as 'captured death animation 3' etc.).

Agreed.

But a good place to start is with human and human-like characters and setting a standard to build from.

Whole different thing. Multithreading is essential for all games as that's the way tech is moving. A feature like physics or AI being forced on devs is just a damned nuisance.

Agreed.

It will be a nuisance. But it will also be necessary for most games moving forward if my theory of greater hardware = more realism/realistic game-worlds is correct.

The idea of the standard is to make it as painless as possible to implement.

It should remain the developer choice what techniques they use and what they don't, while the industry will evolve techniques and tools to help enable that.

Agreed.

Again, the concept isn't a take it or leave it proposition. It's more along the lines of: "get to this end result however you wish, and here is a reference design on how to get there included with the sdk."

 

[TheChefO]

But if console makers start dictating game impacting features then it becomes unwieldy. Just like no console manufacturer mandates Spoken Voice for every game (much less every character or NPC in the game), they shouldn't mandate anything resembling natural motion.

The difference would be if platform holders provided a spoken voice library in the SDK.

__________________

The concept of not forcing things on devs and devs having all the freedom in the world to do whatever they want is a bit misleading.

Devs can't target KB/M on xb360 ... they have to target a gamepad and/or kinect.

Devs can't use openGL on xb360 either.

Devs cant target 640x480 or lower for retail games.

In fact, at the start of this gen, MS wanted to institute a 720p standard and AA standard. But the hardware wasn't quite up to snuff for such a standard with only 10MB EDRAM.

I agree that introducing gameplay standards is a slippery slope, but for the cases I've outlined, it's a worthy standard IMO.

 

[function]

There was a video recently where two of the Virtua Fighter design team talked about (among other things) how they had to move away from realism to make things look right and play right.

Enforcing physically correct motion will break things that rely on none physically correct motion to look and play correctly. Even in Halo, even with the human characters, they move in ways that are unnatural in order to allow them to operate correctly within the game.

Allowing things to operate outside of a physically correct or consistent set of rules is one of the great freedoms that games allow. It's, like, one of the things that makes games great.

 

[Laa-Yosh]

I didn't say that 'scanned' avatars are a bad idea or that low quality results could not work in a game.

All I said is that tech for creating high quality avatars isn't really available right now. The higher you raise the quality bar, the more complex the technical implementation needs to be, and there's no way to automate it, not even in the near future.
In fact, even professional tools for high quality scanning are very, very rare and expensive - the 360 degree laser stuff used for most movie VFX costs a fortune, there's only a few of them in the world. And even those systems provide a dense point cloud that can't be used on its own, a CG modeler needs to go in and rebuild a more efficient and ordered version, then the two can be compared to generate basic displacement and normal maps. But it's usually cleaned up and processed before that, using digital sculpting tools like zbrush or mudbox.

So don't expect something like a new Mass Effect series that scans the player to become the new main character for Kinect 2, maybe even for Kinect 3...

 

[Shifty Geezer]

How's about a system of modification like EA's sports customisation, that uses a default mesh and allows parts to be shifted? I can imagine a system where a head and face can be scanned and metrics derived and applied to a standard mesh. Getting a texture could be quite hard as there'll be ambient lighting, but maybe it could select from a number of preset pieces for a photofit type composition? You'd never get high assets, but I feel like a decent construct could be managed with a Kinect 2. Kinect has shown some amazing 3D world construction. The trick would be generating the mesh IMO, which, as I say, could be handled like a character customisation process on an existing head mesh.

 

[TheChefO]

Enforcing physically correct motion will break things that rely on none physically correct motion to look and play correctly. Even in Halo, even with the human characters, they move in ways that are unnatural in order to allow them to operate correctly within the game.

Allowing things to operate outside of a physically correct or consistent set of rules is one of the great freedoms that games allow. It's, like, one of the things that makes games great.

It's funny because at the same time I agree, and I see what your saying and can apply it to many instances, I disagree as I think the thing that makes games great is the interactivity.

Bottom line I know that a straight enforcement of Euphoria on games where a group/team of humans are jammed together in a tight area or need to quickly follow a intricate route would end in disaster as they fumble on themselves entering stagger mode over and over again, but that's where further R&D pays off.

New Euphoria rules which determine force strength to impose stagger or to impose stubbed toe, finger, elbow instead without breaking stride. Or about keeping a certain distance between characters to avoid collision.

The last thing you want is a permanently broken game, but at the same time, interactivity should trump static scripted events.

 

[Silent_Buddha]

The difference would be if platform holders provided a spoken voice library in the SDK.

__________________

The concept of not forcing things on devs and devs having all the freedom in the world to do whatever they want is a bit misleading.

Devs can't target KB/M on xb360 ... they have to target a gamepad and/or kinect.

Devs can't use openGL on xb360 either.

Devs cant target 640x480 or lower for retail games.

In fact, at the start of this gen, MS wanted to institute a 720p standard and AA standard. But the hardware wasn't quite up to snuff for such a standard with only 10MB EDRAM.

I agree that introducing gameplay standards is a slippery slope, but for the cases I've outlined, it's a worthy standard IMO.

Seriously? A spoken voice SDK? The simplest thing in the world to include and you believe an SDK would boost availability in games?

Here's a hint, an SDK would do absolutely Zero to boost the prevalence of spoken voice in a game. Spoken voice is limited mostly by available disk space and far more importantly developement costs. Voice actors aren't necessarily cheap if you want decent ones. Lets say a console manufacturer decided to mandate it as a requirement. All NPCs had to have spoken dialog no matter how much dialog there was. You'll see many smaller developement project cancelled or their scope reduced. Titles with lots of dialog will be canned or dialog simplified. Developement costs of AAA titles will skyrocket. Etc.

And then somehow comparing implementing Natural Motion as similar to what controller is used? Again, seriously? How does choice of controller impact rendering quality or features? Zero. Gameplay? Only makes RTS games difficult. Computation costs? Zero. Developement costs? Zero. Added time to implement? Zero.

Natural Motion on the other hand? Rendering quality will be impacted as you'll have to drop other features in order to implement even in next gen consoles and even if they were an order of magnitude more powerful than the most powerful consumer PC a person can get today.

Gameplay? Enemy AI will have to be adjusted as them instantly turning to respond to a players presence will become impossible. Enemy response slowed down to be significantly slower than a player as they will have to turn their bodies, stay balanced, bring their weapons up to bear, and then take a shot. In essence making ALL shooters significantly easier.

Computation costs? Significant and thus requiring compromises in other areas of the game whether it be graphics, physics, AI, game logic, or whatever.

Developement costs? Significant. Physics SDKs still require significant money and time investment to get working correctly according to the needs of the project. And this despite the fact that we have had physics SDKs available for over a decade now. And those are still greatly limited in what physics they can emulate on consoles and computers (which have far more computational ability than consoles) even when employing hardare GPU acceleration (now going into multiple orders of magnitude more computational ability). Natural Motion is like a whole order of magnitude more complex. Especially if you want to use it for more than just the occasional animation which canned animation doesn't fit. Even your vaunted GTA IV example only uses it sparingly because to do otherwise would have resulted in far more feature compromises and most likely less realistic total animations. Euphoria is signficantly more expensive to license and implement than something like Havok not to mention Bullit or PhysX.

And a large part of that is that Euphoria really isn't an SDK. You are actually hiring the company to create natural motions for you. We're still a LONG way away in not only time and money from any company even MS + Sony combined being able to create some sort of SDK that would allow more than just specific use cases.

And trying to compare it to rendering resolution? Seriously? Besides the fact that MS is more than willing to allow for lower resolution if a game proves to be impossible without lowering the resolution, going down to 640x480 will likely mean greatly reduced sales anyway. Then again if a game really needed to, I'm sure it might be allowed. Some games already go as low as 540p which is a far cry from 720p.

Microsoft wanting to enforce MSAA was due entirely to the fact that they had assumed that with the hardware it would be FREE. In other words, no rendering compromises, no increased developement budgets, no costs that would require changing gameplay or rendering quality in any way.

What was the reality? The prevalence of deferred renders making MSAA not as useful or noticeable when the later lighting stages completely obscure the fact that polygon edges had been AA'd in a previous pass. As well, there ended up actually being a cost to rendering speed if a scene needed to be tiled. So what happened? It ended up not being a requirement.

So lets see. All these things you try to compare them to have no impact on rendering quality, rendering techniques, developement costs, computation costs, time to implement costs, etc... And the only area they impact gameplay is with RTS titles. Hmmm...

That's a far cry from natural motion which will require compromises in all of those areas. Every single one of them.

Hence, I'm fine with a developer "choosing" to implement it at the expense of something else. I'm also fine with a developer "choosing" not to implement it in order to implement something else. I'd LOVE to see natural motion implemented in next gen games a differentiator (the post topic) as long as it fits the game and the developer chooses it over other things they could implement to differentiate themselves from other titles. Doesn't change that I feel other things would be more noticeable and do more for potential immersion in a game at a far FAR lower cost in both computation and moneytary budget.

I'm certainly not fine with a console manufacturer forcing/mandating it's implemention which ultimately means potentially (actually guaranteed) reduction in graphics quality and increased developement costs. For developers who would have already chosen to use it, it'd have little to no impact obviously. For developers who would not have chosen to do it? Yup, higher dev costs and lowered graphic fidelity.

Regards,
SB

 

[TheChefO]

Seriously? A spoken voice SDK? The simplest thing in the world to include and you believe an SDK would boost availability in games?

Here's a hint...

Here's a hint, the word: "library" from my reply. Now assuming you can follow along, this implies that a large sampling of common mutterings could be plugged into an otherwise voiceless game world as the imaginary voice library was compiled provided to the developers. It was an example based on your suggested comparison.

In the case of Euphoria, the motion sampling has already taken place and the library base exists. Now that library just needs to be added to, and tweaked.

And then somehow comparing implementing Natural Motion as similar to what controller is used? Again, seriously? How does choice of controller impact rendering quality or features? Zero. Gameplay? Only makes RTS games difficult. Computation costs? Zero. Developement costs? Zero. Added time to implement? Zero.

It's a comparison of a restriction imposed on developers. I'm pretty sure you can follow the logic there.

Natural Motion on the other hand? Rendering quality will be impacted as you'll have to drop other features in order to implement even in next gen consoles and even if they were an order of magnitude more powerful than the most powerful consumer PC a person can get today.

Shockingly, games are more than just graphics. Interaction is a key element. Having more realistic graphics just helps to point out other elements which stand out as being unrealistic in the game world.

Gameplay? Enemy AI will have to be adjusted as them instantly turning to respond to a players presence will become impossible. Enemy response slowed down to be significantly slower than a player as they will have to turn their bodies, stay balanced, bring their weapons up to bear, and then take a shot. In essence making ALL shooters significantly easier.

You mean enemies couldn't spin on a dime anymore once "alert_mode" was activated? Forgive me as I wipe away my tears to mourn the loss of unrealistic actions in modern games.

Computation costs? Significant and thus requiring compromises in other areas of the game whether it be graphics, physics, AI, game logic, or whatever.

Significant indeed. In fact, I'm not even sure they be able to render enough characters on screen to create a believable open world, much less throw cars in the mix while still having enough CPU power to have AI for them all and then render on top of it? No way...

...Euphoria is signficantly more expensive to license and implement than something like Havok not to mention Bullit or PhysX.

I'm not sure how many more times I need to say it, but I'll try again, my concept is for Euphoria to be bought out and improved by the platform holders to be provided in the SDK to developers use.

And a large part of that is that Euphoria really isn't an SDK. You are actually hiring the company to create natural motions for you. We're still a LONG way away in not only time and money from any company even MS + Sony combined being able to create some sort of SDK that would allow more than just specific use cases.

The specific use case which is available has already been researched and implemented in multiple games. Do you honestly think every game Rockstar makes has to go back to square one with Natural Motion for them to try and "figure it out" again.

And trying to compare it to rendering resolution? Seriously?

Again, a standard which restricts developers choice.

Microsoft wanting to enforce MSAA was due entirely to the fact that they had assumed that with the hardware it would be FREE. In other words, no rendering compromises, no increased developement budgets, no costs that would require changing gameplay or rendering quality in any way.

As I'm sure you're aware, there is always a cost. 5% was their figure. The intent with the original mandate was for MS to uphold a visual standard on their platform.

What I'm suggesting is an interactive standard.

What was the reality? The prevalence of deferred renders making MSAA not as useful or noticeable when the later lighting stages completely obscure the fact that polygon edges had been AA'd in a previous pass. As well, there ended up actually being a cost to rendering speed if a scene needed to be tiled. So what happened? It ended up not being a requirement.

Indeed. But that doesn't change the intent.

As for the standard I'm proposing, it wouldn't be for all games, which I've made clear multiple times now.

So lets see. All these things you try to compare them to have no impact on rendering quality, rendering techniques, developement costs, computation costs, time to implement costs, etc... And the only area they impact gameplay is with RTS titles. Hmmm...

Real Time Strategy titles?

That's a far cry from natural motion which will require compromises in all of those areas. Every single one of them.

And yet, multiple developers and titles did not have an issue with them and in fact, were fairly successful. Whoda thunk it with such a huge disadvantage on their hands?

Doesn't change that I feel other things would be more noticeable and do more for potential immersion in a game at a far FAR lower cost in both computation and moneytary budget.

And you have every right to suggest that belief. I believe that investing broadly into Euphoria for the entire platform by the platform holders and mandating some form of it for a certain class of games will yield more innovation in that field and create more believable worlds and characters with greater interactivity.

Instead of depending on a small company such as a developer or publisher to invest on a per game basis, the platform holder can absorb costs incurred to develop the tech into the broad range of the platform's AAA games which are for the most part featuring human or human-like characters.

Graphics are great, but interactivity helps to sell the believability (and fun) of the world just as much if not more so.

I'm certainly not fine with a console manufacturer forcing/mandating it's implemention which ultimately means potentially (actually guaranteed) reduction in graphics quality and increased developement costs. For developers who would have already chosen to use it, it'd have little to no impact obviously. For developers who would not have chosen to do it? Yup, higher dev costs and lowered graphic fidelity.

Investment in the tech by platform holders would ideally result in optimized code and a developer friendly interface.

Having said that, games aren't all about graphics. If you want ultra high definition and realistic graphics, you can pop in Avatar. For nextgen games though, I'm saying a portion of the processing budget should be allocated to having those beautiful graphics not only look great, but act and interact great too.

Games that have used Euphoria this gen weren't exactly Dogs in the visuals either. So while I'm sure there is an impact for running the system, it obviously isn't a crippling impact ... and this is on current gen systems. That impact on nextgen systems will be even smaller.

 

[dobwal]

But you're correct there are many other ways a developer could remove the twitch/reaction/physical skill component from the equation to allow for more casual players to enjoy the games.

I am curious to know why you assume that "twitch" based game mechanics are somehow less inviting to casual players especially when there is plenty of data to suggest otherwise. I am also curious of your encouragement of a return to a more stat based rpg mechanic to attract casuals when stat based rpgs (outside of wow and a few other titles) arent especially known as casual friendly.

Ipad and Android the two biggest platforms for casual gamers arent really crowded with rpgs and have a healthy representation of twitch based games as popular downloads.

Maybe the reason rpgs arent as stat based and are moving to a more twitched based mechanic is because developers are trying to widen the demographics of the genre.