Realizing Bullet Time in Realizing Bullet Time in movies: visual - PDF document

Bullet time Bullet time Realizing Bullet Time in Realizing Bullet Time in � movies: visual effect combining slow motion movies: visual effect combining slow motion � with dynamic camera movement with dynamic camera movement Multiplayer Games with Multiplayer Games with � computer games: player can slow down the computer games: player can slow down the � Local Perception Filters Local Perception Filters surroudings to have more time more time to make decisions to make decisions surroudings to have � easy in single player games: slow down the game! easy in single player games: slow down the game! � Jouni Smed, Henrik Niinisalo, Harri Hakonen Jouni Smed, Henrik Niinisalo, Harri Hakonen � how about multiplayer games? how about multiplayer games? � Turku Centre for Computer Science (TUCS) Turku Centre for Computer Science (TUCS) Department of Information Technology, Department of Information Technology, University of Turku, Finland University of Turku, Finland Bullet time in multiplayer games Local perception filters (LPFs) Bullet time in multiplayer games Local perception filters (LPFs) � introduced by Sharkey, Ryan & Roberts (1998) introduced by Sharkey, Ryan & Roberts (1998) � two approaches: two approaches: � � � a method for hiding communication delays in a method for hiding communication delays in � speed up the player speed up the player � � networked virtual environments networked virtual environments � slow down the other players slow down the other players � � exploits the human perceptual limitations by exploits the human perceptual limitations by � � if a player can slow down/speed up the time, if a player can slow down/speed up the time, � rendering entities slightly out- rendering entities slightly out -of of- -date locations date locations how it will affect the other players? how it will affect the other players? based on the underlying network delays based on the underlying network delays � localize the temporal distortion to the immediate localize the temporal distortion to the immediate � � causality of events is preserved causality of events is preserved � surroundings of the player surroundings of the player � rendered view may have temporal distortions rendered view may have temporal distortions � � but how to do that? but how to do that? � � rendered view ≠ real view rendered view ≠ real view � Rules of LPFs Rules of LPFs Entity types Entity types Player should be able to interact in real Player should be able to interact in real- -time time � active: indeterministic, unpredictable (humans) active: indeterministic, unpredictable (humans) 1. 1. � ⇒ players ⇒ with the nearby entities. with the nearby entities. players Player should be able to view remote Player should be able to view remote � local: residing in the same computer local: residing in the same computer � 2. 2. interactions in real interactions in real- -time, although they can be time, although they can be � remote: connected over a network remote: connected over a network � out out- -of of- -date. date. � passive: deterministic, predictable (projectiles, passive: deterministic, predictable (projectiles, � buildings etc.) buildings etc.) Temporal distortions in the player’s perception Temporal distortions in the player’s perception 3. 3. ⇒ ⇒ entities entities should be as unnoticeable as possible. should be as unnoticeable as possible. p r p r q q n n 1

Interaction between players Example: Temporal distortion Interaction between players Example: Temporal distortion � interaction = communication between the players interaction = communication between the players � � local players: immediate local players: immediate � Blue view Blue view � remote players: subject to the network latency remote players: subject to the network latency � � time frame = current time time frame = current time – – communication delay communication delay � � interaction = players exchanging passive entities interaction = players exchanging passive entities � passive entities are predictable ⇒ ⇒ they can be rendered in the � passive entities are predictable they can be rendered in the � past (or in the future) past (or in the future) � a passive entity can change its time frame dynamically a passive entity can change its time frame dynamically � � the nearer to a local player, the closer it is rendered to the the nearer to a local player, the closer it is rendered to the � Orange view Orange view current time current time � the nearer to a remote player, the closer it is rendered to its the nearer to a remote player, the closer it is rendered to its � time frame time frame Temporal contour (from the blue Temporal contour (from the blue Example: Pong Example: Pong player’s perspective) player’s perspective) � two active entities: two active entities: � t t paddles paddles � movements movements � unpredictable unpredictable d d � one passive entity: ball one passive entity: ball � y y � movements movements � predictable predictable � latency of latency of d d seconds seconds � x x The view of The view of the blue the blue player player The view The view of the of the red red player player t t t t 2

Temporal contours in Pong Worth noting Temporal contours in Pong Worth noting Blue player Blue player Red player Red player � original paper: continuous temporal contours; original paper: continuous temporal contours; � here: simple linear functions here: simple linear functions � LPFs are the ‘opposite’ of dead reckoning LPFs are the ‘opposite’ of dead reckoning � � no prediction for remote players no prediction for remote players � � the closer the players get, the more noticeable the closer the players get, the more noticeable � the temporal distortion becomes the temporal distortion becomes � in critical proximity interaction becomes impossible in critical proximity interaction becomes impossible � � no mêlée no mêlée � Example: Temporal contours 2½- -dimensional temporal contour dimensional temporal contour Example: Temporal contours 2½ d ( d ( p p , , r r ) ) t t p p r r x x y y d ( d ( r r , , p p ) ) x x x p p r r x Many players Many players Problems Problems d ( ( p p , , s s ) ) d � original approach: visual disruptions on impact original approach: visual disruptions on impact � d ( d ( p p , , r r ) ) ⇒ ⇒ shadows (see the paper for details) shadows (see the paper for details) d ( d ( p p , , q q ) ) � sudden changes in the player’s position or delay sudden changes in the player’s position or delay � can cause unwanted effects can cause unwanted effects x x p p r r q q s s � if a player leaves the game, what happens to the if a player leaves the game, what happens to the � temporal contour? temporal contour? d ( d ( p p , , s s ) ) � third party instrusion: someone with a high delay third party instrusion: someone with a high delay � d ( d ( p p , , r r ) ) ‘blocks’ the incoming entities ‘blocks’ the incoming entities � jitter: entities start to bounce back and forth in time jitter: entities start to bounce back and forth in time d ( d ( p p , , q q ) ) � x x p p r r q q s s 3

p shoots shoots r r while while p p is using bullet is using bullet p Adding bullet time Adding bullet time time time � player using the bullet time has more time to player using the bullet time has more time to � d d ( ( p p , , r r ) ) react react b ( b ( p p ) ) p p r r ⇒ the delay between bullet ⇒ the delay between bullet- -timed player and the timed player and the other players increases other players increases x x � add artificial delay to the temporal contour add artificial delay to the temporal contour � d ( d ( r r , , p p ) ) r r p p x x b ( ( p p ) ) b 2½- 2½ -dimensional temporal contour dimensional temporal contour p shoots shoots r r while while r r is using bullet time is using bullet time p and bullet time and bullet time b b ( ( p p ) ) t t d d ( ( p p , , r r ) ) p r p r x x y y b ( ( p p ) ) b d d ( ( r r , , p p ) ) p r r p x x x x MaxMazeDemonstrator MaxMazeDemonstrator Future work Future work http://staff.cs.utu.fi/staff/jouni.smed/mmd/ � non non- -linear temporal contours linear temporal contours http://staff.cs.utu.fi/staff/jouni.smed/mmd/ � � how to compute quickly? how to compute quickly? � � noticeable benefits (if any)? noticeable benefits (if any)? � � numerical evaluation numerical evaluation � � measuring the distortion and its effects measuring the distortion and its effects � � practical evaluation practical evaluation � � how well does it work? how well does it work? � � does it allow new kinds of games? does it allow new kinds of games? � 4