EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

[Lisa_Bonami]

I'd like to know from one of the EA developers if we are close to having the technology in EA's future Madden (and/or other sports titles) to finally replace repetitive motion capturing technology that triggers scripted animation captured sequences from a library?

Are we close to seeing truly infinite levels of real motion physics animations driven by realtime collision detection algorythms and such. I think that for video sports games to be truly immersive and feel completely different each time you play a contest, this really needs to be incorporated.

Im just curious to know how far off we are from this technology?....ballpark? 20 years? 2 years?, etc?

Also your thoughts on this subject?.....for and against it?

It would be an interesting topic of discussion

[RGiles36]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

It would be an interesting topic of discussion had it not been discussed several times now.

Ever since people started hearing about Natural Motion Euphoria tech, mo-cap for some reason has become the devil . With enough variety in animations, I don't think that "Full Physics based Animations" are necessary.

[Exonerated]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

Imo. Probably next-gen. ie ps4.

It isn't that far away. Its just that the ps3 and xbox360 aren't that powerful atm . They lack alot of ram.they only have 512 i believe. Whilst real physics would still be work, it'll take a lot of testing and probably would be better for next -gen.

[Ian_Cummings]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

Both Simon and I have answered this in a number of different threads over the past few months, but in short, it will be a LONG time until you see a FULL replacement of mocap. Recreating ALL of the movements of the human body through some sort of algorithm just isn't feasible or worthwhile yet.

Strictly for collision though, procedural manipulation of pre-existing motion capture is definitely the way of the future (and present). That isn't necessarily all-inclusive to physics either...there is much that can be done with IK, animation layering, dynamic animation swapping, and other technologies outside of physics to get more variety. Meaning, the problems like "lacking animation variety" and "immersiveness" can totally be solved without going to as drastic a solution as "replace all collisions in the game with procedural ones, ensuring that your art team now has zero control over the look of the animations in your game". It's kind of a throw the baby out with the bath water approach...(which is the approach we took in 06 btw and everyone now probably knows it just isn't the best approach in yearly iterative sports titles).

In short, using a combination of motion capture and other run-time manipulation will continue to give you the look that you must have when creating a simulation sports title - which is the look of professional athletes (and not of flailing dead bodies). I'm very curious as to whether NaturalMotion can deliver Backbreaker on their technology without running into that problem. I kind of hope they can, as that will be a great step in the industry towards a problem that has yet to be solved.

Hope that gives you some good background on what we're thinking, at least within the Madden team.

[Cusefan]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

Thing about this generation of Physic based Animations is that there is no weight behind the Animation. Look at GTA IV, when Niko gets hit by a car going about .5 MPH he goes flying like he is dead.

Now when you add weight that could create a more lifelike animation Things get alot more Complicated. Assuming Madden did use Euphoria style Physics engine, they would have to Multiply what GTA IV did by 22. That is a pretty tall order considering all of the Moves players can do.

Seriously, I do not see a real Physics based Engine in any major Sports game for quite some time, Maybe next Gen. And as Far as Backbreaker goes, lets Play the game before anyone uses it as an example...

[kmart2180]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

I dont know if this has anything to do with the topic but I put it here cause I seen the word physics and when I think physics in Madden I always think about the 300lb vs the 150lb guy difference.

How close are we to the day when a 300 lb guy mattered against a 150lb guy? Example: If I create a 300lb O-Lineman and a 150lb D-linemen, the 300lb lineman should kill the 150lb D-Lineman. It shouldnt even be close. Now the same thing if I created a 300lb RB (unrealistic) but I dont want to see a little cornerback be able to hitstick this 300lb rb like it was easy.

What Im getting at is How close are we to Weight actually mattering?

[NuJoiseySoulja]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

What Im getting at is How close are we to Weight actually mattering?

They actually did that in NFL 2K5. I tested it, I made a running back 5' 8" 160 lbs with break tackle set to 99. He would break the tackles but he wouldn't run them over, it was a totally different animation.. the 5' 8" 160lb guy was slowed down when he trucked over the linebacker (Urlacher). Then I went back put his break tackle at 70 and it was a lot harder to break tackles, the guy (Brian Westbrook edited to be 160) kept getting rocked and once in awhile i would break a tackle but most of the time, nothing. This was on All-Pro difficulty in practice.

[REDRZA]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

Both Simon and I have answered this in a number of different threads over the past few months, but in short, it will be a LONG time until you see a FULL replacement of mocap. Recreating ALL of the movements of the human body through some sort of algorithm just isn't feasible or worthwhile yet.

Strictly for collision though, procedural manipulation of pre-existing motion capture is definitely the way of the future (and present). That isn't necessarily all-inclusive to physics either...there is much that can be done with IK, animation layering, dynamic animation swapping, and other technologies outside of physics to get more variety. Meaning, the problems like "lacking animation variety" and "immersiveness" can totally be solved without going to as drastic a solution as "replace all collisions in the game with procedural ones, ensuring that your art team now has zero control over the look of the animations in your game". It's kind of a throw the baby out with the bath water approach...(which is the approach we took in 06 btw and everyone now probably knows it just isn't the best approach in yearly iterative sports titles).

In short, using a combination of motion capture and other run-time manipulation will continue to give you the look that you must have when creating a simulation sports title - which is the look of professional athletes (and not of flailing dead bodies). I'm very curious as to whether NaturalMotion can deliver Backbreaker on their technology without running into that problem. I kind of hope they can, as that will be a great step in the industry towards a problem that has yet to be solved.

Hope that gives you some good background on what we're thinking, at least within the Madden team.

What i want to know is are we CLOSE to a point where the animations are so varied and more random so that the same over effective catch animations cant be easily triggered to the point of abuse?? I mean one of the reasons we all love Madden is because of its PLAYABILITY (for lack of a better term), meaning the game has always gave us great ability to click onto players and make plays.

Since Madden 2004 when the big jump catch animations were added (ie; the Jetpack and Rocket etc) it seems these animations are too easily triggered to the point of abuse. I think not only more animations are needed, but more factors should be coded in to determine which animations play out. This goes back to the QB play, and how accurate all throws seem to be. Everything from the QBs pass ratings to even weather and wind conditions should effect pass accuracy. From there, even more factors come into play such as better DB AI and coverage. Positioning of both players in relation to the ball etc.

Many more catch animations are needed, and many more factors which determine which animations play out need to be coded in. Not just for catching and DB/WR play..this goes for tackles, blocks, and much more. Point is the game becomes mundane when all you see are the same over effective animations over and over. The game needs more live action randomness, like every game is unique and has its own life. To me that is what "NEXT GEN" gameplay was supposed to be.

This has been an issue since LAST GEN. Just want to know if the game can move past this?

[NYGmen56]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

I dont know if this has anything to do with the topic but I put it here cause I seen the word physics and when I think physics in Madden I always think about the 300lb vs the 150lb guy difference.

How close are we to the day when a 300 lb guy mattered against a 150lb guy? Example: If I create a 300lb O-Lineman and a 150lb D-linemen, the 300lb lineman should kill the 150lb D-Lineman. It shouldnt even be close. Now the same thing if I created a 300lb RB (unrealistic) but I dont want to see a little cornerback be able to hitstick this 300lb rb like it was easy.

What Im getting at is How close are we to Weight actually mattering?

Lol so true..

Ian, I hope you guys fixed the physics & momentum in Madden 2010, because there is NONE in 2009 thats for sure.. It's almost comical seeing 150lbs rb truck 300+ lbs DT's! Or QB's trucking DT's & LB's lol thats a joke.. I think thats another part of 2009 & before that makes Madden play and look truely unrealistic. I say this, if 2K sports did it 4 years ago, you guys should be able to do it as well.. no excuses..

[Lisa_Bonami]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

I think to truly bring sports games to that unlimited replayability potential, we need to see rendered animations driven by true motion physics with full ranges of human motion.

With that said, until that time arrives, I think we need to increase the motion captured animations library for each activity/event/action by at least 4 to 5 times. It seems like year after year, you play thru the first few games of Madden and are impressed with the few new animations that are added and then once you've seen them all, the limited way in which each occurrence plays out makes the game feel stale and repetitious very quickly.

For example, I constantly get the deja vu feeling everytime I have a reciever cutting accross the middle and I dump him a pass where he will reach back and extend to make the catch behind him then the covering LB horse collars him from behind with that swing around tackle to bring him down. This is just one example of how every situation just feels so limited in what can transpire. Its just such a "OK here we go, been here, done that...this is what is going to happen here..." feeling

In addition, the game could also use a few of those triggered animations that will happen once in a while in only the rarest of moments to give that "anything can happen surprise" feel to the game from time to time.

[Skyboxer]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

My biggest issue with the animations today is a good amount look pretty stupid. Not to mention we still see the "shoulder tackle" and players bouncing off each other and falling in different directions. I could care less how we get the animations as long as they remotely look like that would happen in the real world. Right now we are far from that.

[countryboy]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

My biggest issue with the animations today is a good amount look pretty stupid. Not to mention we still see the "shoulder tackle" and players bouncing off each other and falling in different directions. I could care less how we get the animations as long as they remotely look like that would happen in the real world. Right now we are far from that.

AMEN.

[hedphurst]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

I went to school for 3D animation back in 2003, and the technology for realistics physics was already available back then, and it didn't require a whole lot of CPU power or RAM to do it. The problem isn't the physics, it's the work required to create the player models that have weights, joints, and density properly set. It's pretty easy to create a player's body that looks right. It's a whole different ballgame to create a player with a belly that gives, shoulder pads that don't, and a torso that can twist realistically to absorb a hit, pivot to stretch for a first down, a knee that can hyper-extend, etc.

EA has demostrated over the years that they're not really willing to start from scratch unless it's absolutely neccessary, and as long as people keep buying roster updates in pretty new packaging, they're not going to go back to the drawing board to create realistic physics-based player models when most gamers just want the pixels to look right instead of caring about the muscle and skeletal system underneath.

[youALREADYknow]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

In short, using a combination of motion capture and other run-time manipulation will continue to give you the look that you must have when creating a simulation sports title - which is the look of professional athletes (and not of flailing dead bodies). I'm very curious as to whether NaturalMotion can deliver Backbreaker on their technology without running into that problem.

Correct me if I'm wrong, but aren't we talking about multiple technologies when we typically reference NaturalMotion as a company? NaturalMotion has already proven that it is possible to blend motion capture technology with physics based transitional animations, but my understanding of what they are doing with Backbreaker is abandoning mocap in favor of computer generated keyframes with the transitional animations used to create the realistic look.

Regardless of the approach that either game uses, the larger underlying problem in my opinion is the background logic that determines which motions, animations, and sequences are available to a given player in a situation.

For example, it would not matter much that Madden only has 20 different tackling animations from behind if the logic of what triggered each animation was thoroughly defined and made sense in regards to real world physics. Not using a full-fledged physics engine isn't an excuse to not code logic to abide by those laws whenever possible. Weight and height are known variables within the Madden world and yet they are largely ignored when I play the game.

In this regard, I'm left with the feeling as a consumer that the developer has been extremely lazy and/or inept at developing an in-depth football game. The tools are readily available and the only thing that seems to be truly lacking is the code and logic.

Sorry if that is too harsh of a criticism and I'm sure that there are tons of people who dedicate long hours to making the game as good as it can be, but all I can judge as a consumer are the results. Looking forward to 2010 being a step in the right direction, but I hope that these kinds of game-wide issues are being looked at since they are likely to have more impact than any new game mode or enhanced feature.

[Lisa_Bonami]

Re: EA, Are we CLOSE to FULL Physics Based Animations to replace MOCAPs??

YouAlreadyKnow, Nicely said, to add in laymens terms to what you just said, in a nutshell, we have largely been playing the same game for the last 10 years or so with the addition of higher resolution textures and increased polygons. But the game is so underdeveloped in so many other areas and its about time to start giving attention and focus in these areas.

We need better ratings relationships to the results and interactions that result on the field. We need to see deeper relationships in terms of weight, height, player awareness, consistency, veteranship, leadership, Intelligent AI audible calling, late game situational awareness and tendencies.

We need more development in Team distinctions, philosophies, play styles, Player indivualities

We want real sound effects of shoulder pads and helmets colliding, not simulated car crashes and other over the top gumbo. We want dynamic crowds that react to realistically to home team actions and give us the sensation of being in those big play and game situations.

We need Playoff and Superbowl games to feel like playoff and Superbowl games...up the presentations, the commentaries, the head to head and player analysises during these games and make it feel special

We need more realistic and interactive, tackling, juking, running, Scrambling, defending. Tighter control and more immersive struggling for that extra yardage, fighting in the trenches, realistic blocking and pocket forming.....we are so far from the real things that need developing. This game hasnt advanced these areas over the last 10 years to where I feel and most of us feel they should be by now.

Im not intersted in getting every blade of grass accurately depicted in shade of green in 1080P resolution.

I want gameplay that grabs me by the ears and drowns me into the trenches of real gridiron video game football and leaves me gasping for air!

I want a football that will finally surpass what 2K has already delivered years ago