Multiball Ghost-Ball Collisions

[Crush3d_Turtle]

I have seen several posts around here where people talk about ball collisions during multiball where the balls will collide without actually hitting each other. Amazingly, I noticed that we do in fact have video footage of this happening in the latest DLC video posted by FS. Here is a link to the video a few seconds before the "ghost collision" happens. It happens right after the game says "Taxi, taxi, taxi." Using the arrow keys to slow the video down, you can clearly see that the left ball has never even left the flipper when the collision happens about half way up the table. I have seen this happen many times while playing the game and was quite shocked to see it during one of their demo videos. At least we now how conclusive evidence that it does happen and hopefully this can get addressed.

 

[laughing_lunatic]

If you go back, there's another one at the 2:11 mark, Also, slowing the video right down, you can clearly see that the ball often bounces off the slingshots without actually touching them. Hard to see full speed, but it's at the 2:53 mark. At 2:26 a drop target drops by having a ball go over the top of it on the rail. What is going on with the physics engine ??

 

[Fungi]

Are we so sure it's TPA? YouTube ain't exactly 60 frames per sec.

 

[laughing_lunatic]

true...true... GSpot says it's only 30

 

[Crush3d_Turtle]

Are we so sure it's TPA? YouTube ain't exactly 60 frames per sec.

From what I see in the video, I believe that there is no way that this can be a product of either the video recording or Youtube playback. At the time I mentioned in the first post, the ball is colliding with another one half way across the table. I sometimes see odd stuff like this when playing TPA so I am thinking it is something with the physics engine.

 

[BonzoGonzo]

yes, finally a video of it happening!!! this is actually the first bug i noticed when i began playing tpa... balls colliding without really touching

 

[Jim O'Brien]

Hate that sound it makes when they do or don't collide it's annoying.

 

[countcb]

My guess is, that the physics engine is running with more fps than the video output of the game.

physic calulates position and the current position gets drawn on screen
physic calculates the next position which is not beeing drawn
physic calculates again. checks that there is a colision, handles the collision, changes speed and direction, calculates position which is again not being drawn
physic calulcates new position and current position gets drawn on screen.

so there have been 2 calculated positions and one collision that was not shwon on screen.
I think this is what causing these "ghost" collisions. That means the collsion engine works as it should. it is just not showing every calculated frame on the screen.

And that the physic calculation is running more often then the screen is drawn is pretty normal and the way it should be.

 

[BonzoGonzo]

when the ball goes through solid objects (flippers, rail, ramp, scoop) and the balls collide with eachother without touching i say no, thats not the way it should be

 

[countcb]

the ball going trough objects is when the physic is not running fast enough to register a collision. this is a bug and should not happen.

The balls colliding without touching isn't actually happening. It just LOOKS like that because the physics run faster than the screen updates.
Thanks for taking the time to read my whole post and not only the last sentence.

 

[laughing_lunatic]

well as pinball is a game that relies on seeing where the ball is, and where it's going, and to an extent predicting where they'll go, having balls colliding without looking like they do, is a MAJOR problem that has to be fixed.

 

[countcb]

if you predict where the ball is going you should come to the same conclusion as the physics and know they will collide

serious answer:
The game is drawing with 60fps... the physics engine is running faster then that. I don't see any way how the problem that the physics calculates something that isn't drawn can be fixed.
This will always be a problem as long as the framerate of the game and the rate of the physics are different.

So as long as your console can't draw as fast as the calculations are done this will remain a problem. Slowing the physics down to 60fps is not an option, because that would result in crappier physics.

 

[Mike Reitmeyer]

I think what you are seeing is the fact that youtube videos are 30fps, when the game runs at 60fps, so you are missing in between frames where the collisions actually happen. I've never seen this while playing.

 

[countcb]

I see this all the time when playing multiball.
Would you be allowd to say how often the physics are updated? Because if it is faster then 60 times per seconds there is actually no way that this could NOT be happening.

And if there actually is (because I'm overlooking something) I would really like to know it
But as a game developer myself I rellay can't see a way to avoid this if the physics are updated with a faster interval then the screen is updated.

 

[Fungi]

Funny, I've never noticed this phenomena due to the fact that I can't look away from the flippers when in multiball.

 

[BonzoGonzo]

I think what you are seeing is the fact that youtube videos are 30fps, when the game runs at 60fps, so you are missing in between frames where the collisions actually happen. I've never seen this while playing.

sorry mike, but i've seen this happening plenty of times in multiball, and as i said earlier it was the first thing i noticed when playing tpa and have been mentioning it on this forum quite a lot. the balls sometimes collide when they are not in fact touching.

it is not just youtubes 30fps limitation sadly

 

[BonzoGonzo]

the ball going trough objects is when the physic is not running fast enough to register a collision. this is a bug and should not happen.

The balls colliding without touching isn't actually happening. It just LOOKS like that because the physics run faster than the screen updates.
Thanks for taking the time to read my whole post and not only the last sentence.

please dont patronize me, kthx

 

[countcb]

Didn't want to patronize you. (Actually had to look in a dictionary what "patronize" meant)

English is not my native language so maybe sometimes things sound different then meant. Sorry if I offended you, but your first reply to my first posting implied to me that you didn't even read what I wrote. Maybe I misunderstood you, so sorry again if my reply wasn't appropriate...

 

[BonzoGonzo]

if you predict where the ball is going you should come to the same conclusion as the physics and know they will collide

nobody is disputing the fact they will collide. the thing is, they get to collide in wrong positions on the playfield.

example (data is made up, i'm just making a point here): 2 balls hitting eachother 2cm above the tip of the left flipper yields in the lower ball richocheting further to the left 2 cm above the left flipper and the second ball further up/left 4 cm above the left flipper. when the ghost collision happens, in the same example the lower ball gets hit to the left 2cm above the left flipper as before, but the second one gets 'hit' before the actuall contact so it doesnt get to just 4 cm, but 6.5cm up/left above the left flipper. becouse of the ghost collision the balls move differently/have different positions on the table even if the angle of the ricochet stays the same/proper

 

[BonzoGonzo]

Didn't want to patronize you. (Actually had to look in a dictionary what "patronize" meant)

English is not my native language so maybe sometimes things sound different then meant. Sorry if I offended you, but your first reply to my first posting implied to me that you didn't even read what I wrote. Maybe I misunderstood you, so sorry again if my reply wasn't appropriate...

lol, dont sweat it we are all human (well i'm actually only half alien, so there)

btw, off course i read all of your post, just didn't agree with the proposed reasoning for the problems cause