Euphoria
acidicX
Join Date: 2004-07-08 Member: 29795Members, Constellation
Join Date: 2004-07-08 Member: 29795Members, Constellation
-> <a href="http://www.naturalmotion.com/euphoria.htm" target="_blank">NeuralMotion Euphoria</a>
Is it possible to include it into the source engine for NS2 ? GTA IV uses it, the most realistic model animations i've ever seen. They've got an office in SanFran, you guys could check it out?!
It would be that great.. and it even saves you time animating the player models! Although I think it will just work for the marines.. they specialized on human reaction ;-)
acidicX
Is it possible to include it into the source engine for NS2 ? GTA IV uses it, the most realistic model animations i've ever seen. They've got an office in SanFran, you guys could check it out?!
It would be that great.. and it even saves you time animating the player models! Although I think it will just work for the marines.. they specialized on human reaction ;-)
acidicX
Comments
i take the vote : <b>NO WAY</b>
It was no bad idea to "mix" it into the RAGE engine either..
i'm just not confident it'll happen <img src="style_emoticons/<#EMO_DIR#>/sad-fix.gif" style="vertical-align:middle" emoid=":(" border="0" alt="sad-fix.gif" /> time, costs, etc.
I do like cool and good looking animations but they are eye candy in my opinion rather than something thats important to the game. I think its a shame that lots of games these days get built and sold on their eye candy value rather than their gameplay or narrative. If ns2s game play is all sussed and done to perfection then may be this should be added in but i wouldnt expect this to revolutionise ns 2.
I do like cool and good looking animations but they are eye candy in my opinion rather than something thats important to the game. I think its a shame that lots of games these days get built and sold on their eye candy value rather than their gameplay or narrative. If ns2s game play is all sussed and done to perfection then may be this should be added in but i wouldnt expect this to revolutionise ns 2.<!--QuoteEnd--></div><!--QuoteEEnd-->
Animations are important to gameplay too though. Seeing where a marine is in his/her reload animation, for example, is important to a player. Also, as far as eye candy goes, animations really make or break immersion. I'd rather have great animations with poor graphics, than poor animations with great graphics.
Though taking Zim's lead, Im also concerned about "fps issues"... If I understood correctly Euphoria is aimed at CONSOLE games where you have constant computation-power as opposed to PCs.
<!--quoteo(post=1677943:date=May 7 2008, 09:27 PM:name=locallyunscene)--><div class='quotetop'>QUOTE(locallyunscene @ May 7 2008, 09:27 PM) <a href="index.php?act=findpost&pid=1677943"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Also, as far as eye candy goes, animations really make or break immersion. I'd rather have great animations with poor graphics, than poor animations with great graphics.<!--QuoteEnd--></div><!--QuoteEEnd-->
Very true, I'd have to agree. But I have one thing to add: something with poor animations and poor graphics is (probably) better than something with poor animations and great graphics. I think it has to do with an incongruency.
the heirarchy then, would be:
great animations + great graphics
great animations + poor graphics
poor animations + poor graphics
poor animations + great graphics
as for the fps issues, can't they base it around the minimum specifications?
Unfortunately I don't see it making business sense to replace realtime physics sequences with realtime physics sequences that look a little nicer. On top of that, this algorithm looks like it would create more server load in a game that's already calculating DI and structures (far more effort than most games put on servers).
If bots were implemented, this could be used, but it seems like total overkill as its not in the NS spirit to play with bots.
<u>If bots were implemented, this could be used, but it seems like total overkill as its not in the NS spirit to play with bots.</u><!--QuoteEnd--></div><!--QuoteEEnd-->
As seen here on the NS2 about page (http://www.unknownworlds.com/ns2/about/):
<!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->Instant Gratification
Rated maps and powerful matchmaking allow you to play with people of any skill level. Play on servers from Easy to Expert or compete in tournaments. Play on the Internet, on a LAN or offline with any number of players. <b>Clever bots will do the rest.</b><!--QuoteEnd--></div><!--QuoteEEnd-->
I'm pretty sure the developers did say they would try to implement bots for NS2, so it just may be in the NS spirit.
But that is besides the point, if I understand Euphoria right, its about replacing rag doll and other labor intensive processes of animation, with a procedurally generated (AI) animation, sorta like in <i>Spore</i>. It wouldn't be about taking control away from the player, it would be about what the player sees happen to others, mostly when they die it seems, the rest of the time it seems on hit would be something like the model still stands its ground but reacts to a hit or on knock back hits, takes a step back and reacts to the hit.
I wonder if Euphoria was ever meant to work for a model other than human, that would definitely factor in if it can work just as well for a kharaa as it does a human.
Of course this a moot point if they already have animations done, but if not, they really should see if its technically and financially possible, it really would add a ton to the atmosphere to have these quality procedural animations.
<img src="style_emoticons/<#EMO_DIR#>/biggrin-fix.gif" style="vertical-align:middle" emoid=":D" border="0" alt="biggrin-fix.gif" />
in source games you have great ragdolls. they whould couse pretty much the same.
i had think euphoria is a AI system to create more RL AI and muscles, so that the AI soldier hold another AI soldier when he fall down a cliff etc. instinct like movements and such.
but for source? no way, i dont know, we need to keep in mind that there are people who havnt a high end PC.
and Euphoria is disignt for AI behavior, and the only use in NS2 could be death animations.
It was no bad idea to "mix" it into the RAGE engine either..<!--QuoteEnd--></div><!--QuoteEEnd-->
1. Way to much work for a 4-man team.
2. It is a bad idea to attempt to "merge" engines.
If you want NS2 to be fully of sloppy code, patches, and redundancy glitches then by all means support having a ####ed up piece of #### engine that will crash more than crack addict in rehab.
3. There are legal rights to source code and other intellectual property. Often even "open source" is both limited in scope and limited in rights (meaning you can't go commercial with it)
This whole idea is a resounding can-of-worms, unrealistic for time & money constraints, etc.
In fact, I can't think of single damn positive about this idea other than the general notion of wanting NS2 to have great model animations (which is full possible in source and is by and large almost entirely up to the model animators, not the programmers), which of course we all would realistically want to have anyways.
<strike><!--sizeo:3--><span style="font-size:12pt;line-height:100%"><!--/sizeo--><b>vote <!--coloro:#FF0000--><span style="color:#FF0000"><!--/coloro-->no<!--colorc--></span><!--/colorc--></b><!--sizec--></span><!--/sizec--></strike>
~edit~
vote temporarily nullified (see page 2)
edit:
from the website: euphoria:core consists of a small kernel, AI controller modules (for adaptive behaviours) and a rigid dynamics module (for physics). euphoria is physics-engine independent and works with <i>all commercially available engines </i>(as well as proprietary ones).
/edit
if it turns out to be too difficult to implement or too costly to consider the the dev team will obviously not go down that path.
personally, i think it would be awesome to see this in game. picture a skulk lunging at a marine and instead of the skulk just stopping when it came in contact, the skulk would actually knock the marine back and you would see him react realistically. this would also take into account what would happen when a large alien lifeform moves through a crowd of marines. if its just one then the marine would obviously get pushed down, but if its a couple the the lifeform would be slowed, or even stopped, depending on its size. it would make going onos much less annoying to play because you cant get completely stopped by a lone marine.
a fade swipe would actually cause a marine to react as if it had just gotten hit by a big friggin claw.
on the other end of the spectrum, a jetpacking marine would be able to fly into a fade trying to block him, and instead of just getting stopped dead in his tracks he would knock the fade back. while he would be slowed down some and even jarred, perhaps thrown into a wall, he wouldnt defy the laws of physics.
keep in mind its a multiplayer game.
A, the server whould lag or crash
B, no player in any game will lose the controll over his body
C, its a 4 man team, not valve or ubisoft. it whould push the release even further as it now, and then we can play NS2 in 2020.
D. if that happen, pushing around. how going it playable? when something pushing you always around and you cant do anything?
I vote: NO
euphoria is disignt for singleplayer games, or coop play, not for multiplayer,
also not everyone have a super-high-end PC
like said before: the only reason whould be the animations, and these going to the player animator, not to the coder.
<b><u>stop saying no because of the technical aspects!</u></b>
no one except the developer team can say that it would be too difficult to implement/too laggy to implement/too expensive to implement/too time consuming to implement/too high end to implement
and if any of those are true then they wont implement it.
<b><u>stop saying no because of the technical aspects!</u></b>
no one except the developer team can say that it would be too difficult to implement/too laggy to implement/too expensive to implement/too time consuming to implement/too high end to implement
and if any of those are true then they wont implement it.<!--QuoteEnd--></div><!--QuoteEEnd-->
Would you read my post above again please?
I'm not saying it isn't something UWE should check out? Who knows there might be something about it that inspires a new idea?
I'm not saying that better model animations are bad. In fact I wholeheartedly agree with locallyunscene's comment, "Also, as far as eye candy goes, animations really make or break immersion. I'd rather have great animations with poor graphics, than poor animations with great graphics"
Where I'm saying a resounding NO to on is the technical aspect of "engine feature merging" as per acidicX's comment, "It was no bad idea to "mix" it into the RAGE engine either.."
Yes, 6john, there <i>are</i> technical problems with that. Look at any software that just tries to quick-fix with a bit of cut & paste when the structure of the source code is fundamentally different. You either get code that kind of buggy, or you have to spend many-fold of untold hours trying to fix the glitches and get it streamlined and efficient. And even then you can still miss some memory leaks or have inefficient overlapping code. Can it be done? Of course it could, but unless that feature is a neat-standardized OOP design that you can just "pop-in", then it's almost one of those "looks simple but isn't cases." Look at it from a legal-technical side. That's intellectual property you just can't "include" without paying somebody, UWE doesn't need more financial debt than what they have to.
Considering all that, you tell me which is more important:
1. Great animations > Great Graphics
2. Stable source code > Great animations
Of course #1 is good (nobody here has said it isn't), but isn't #2 <i>more</i> important?
And on top of all of that... 6john, I'm tired and annoyed of professional software developers trying to develop at hyperspeed and having to use constant patchwork fixes (not just updates, talking core instabilities) in their initial public release. This was the case with Grin's GRAW for PC when they ported it over from 360 design, this was the case with NWN2 when it was first release when they rushed transition programming due to timetable constraints, and heck it's even the case of Microsoft with Vista initial release. Be creative, be inspired, be efficient, but don't release buggy software. Please.
<!--quoteo(post=1678235:date=May 10 2008, 12:44 PM:name=6john)--><div class='quotetop'>QUOTE(6john @ May 10 2008, 12:44 PM) <a href="index.php?act=findpost&pid=1678235"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->edit:
from the website: euphoria:core consists of a small kernel, AI controller modules (for adaptive behaviours) and a rigid dynamics module (for physics). euphoria is physics-engine independent and works with <i>all commercially available engines </i>(as well as proprietary ones).
/edit
if it turns out to be too difficult to implement or too costly to consider the the dev team will obviously not go down that path.<!--QuoteEnd--></div><!--QuoteEEnd-->
(you may or may not have seen this quote already since it was edited in, and there's more info on the website - have you looked at the <a href="http://www.naturalmotion.com/euphoria.htm" target="_blank">website</a>?)
If you really want me to, I can quote the relevant parts; but it's really very straight-forward so I don't think it'll be necessary.
NS2 will already have a licensed copy of the source engine, and to add anything else will just up the cost. I'm not sure about Unknownworld's budget, but I imagine they will not dabble in anything until the have they financial backing and security to do so, especially since adding the euphoria engine could put the game back at least another 6 months.
Secondly, in a world where everything is user controlled (multiplayer) the applications for euphoria would only include death scenes (which wouldn't be that significant because euphoria is designed to emulate conscious movement), or new actions that would requires huge changes to gameplay dynamics (like stumbling over a block in game, or a reaction to being hit in the leg), which would annoy players who are used to having a relatively constant first person view unhampered by shifitng or tilting.
how do you justify this? euphoria is meant to cut down on production time because there are far fewer animations to be done through conventional animation techniques. as for death animations, that's what ragdoll physics are for.
<!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->which would annoy players who are used to having a relatively constant first person view unhampered<!--QuoteEnd--></div><!--QuoteEEnd-->
CoD4, anyone?
Euphoria is not a complete new game engine, it's even wrong to say the engines will be "mixed" (I thought I made that clear in my 2nd posting). You can't just say no because you fear Euphoria will be implemented quickly and bad by UWE. You just don't know it - you just fear it. And, as far as I understandt it, it is only a physics engine, so it doesn't even have to run on the servers. It could easily be implemented client-based only. So please just see it as an ADDITION to the Source engine.
I also never stated this is a must-have, but it would be a great feature AND it can cut-down on on production costs also - I know UWE has a limited budget and they already licensed the Source engine. We also don't know any of the licensing costs of Euphoria. My intention was to show it to the devs, so that they can have a closer look at it.. maybe they can give us some answers in the next podcast or so ;-)
its a multiplayer game, that means there is no place for anhanced physics or great AI.
euphoria: if you could bring it in, whitout much time loss and work, and it have acctually A USE instead of droping FPS, then ok.
but it isnt so easy: euphoria is disignt FOR singleplayer games, or simulators, or next next next generation games, but not for a 100% player controlled, physic less, source engine FPS
if it were to be in ns, it would have to be toned down a lot from how it is in gta, so we dont have marines bumping into each other and knocking each other down, but im sure thats possible. i really just like the idea of having leap at a marine, and when it makes contact if actually feels like contact was made and the other players can see it.
what if the marines lose too much controll and games end always whit a "aliens win" message?
or the marine have too much controll and can easy kill the skulk, who have a "trauma" from the hit.?
the true is: we dont know ANYTHING about the game
Secondly, as stated many times before in other discussions, realistic physics based gameplay is still way too detailed to implement in a game, even for the ephoria engine, and especially in a game centered around controlled balance. You can implement quisi-physics which several engines like source and euphoria do, but even in these games, the more that's added, the buggier, more exploitable games become.
I wouldn't mind euphoria strictly for animations, but when a get knocked into a rail, I don't want my visuals flopping everywhere. I wouldn't mind seeing a marine brace himself, but from the marines perspective nothing should change much. Hell I'd even vote for marines getting knocked over, but view point of marine should never be as disorientating as it would in real life, because in real life we can shift eyes, turn heads, and shoot behind our heads if we really wanted too. Euphoria would work great in a thrid person shooter like max payne but not any first person shooter. Animations mixed with standard view and gameplay would be ideal. The most extensive use I'd make out of it though would knocking players back visually realistically, but once again not disorientating their view much. In other words this may not be super complex to implement, but it would require modifying the engine most likly.
One thing I would like to see would be more complex animations and movment in the game with creatures such as onos who currently hop over small object like a bunny in stead of stepping over them or pushing movable objects out of the way. I sure would think it would be awesome to see and onos with leg on box and the rest on the ground and not seeing two legs floatingin midair while on a box.
And like another user said, it's quite possilbe the engine could cut down production time. The real question then being, is cutting down production worth the cost of the engine liscense and if its worth the cost to only use some of the engine functionality, particularly animations, rather than just implementing more advanced animations designed specifically around the game. For example, would it be better and more cost efficiant to learn how to use euphoria to apply to the structure an onos, lerk and skulk, or would it easier just to make new more advanced animations from scrap.
Also remember this is an rts. The more freedom a player has from physics based engine, the more exploitable and harder to balance a game it is.
And lets admit it physics based animations also have downside unrealistic affects, the most common example is explostion effects. I love blowing up bots out of view in Battlefield and blowing people sky high in golden eye in slowmotion. This is unrealistic looking, albeit fun, I never understood why they just don't add a force cap on objects to prevent these issues.