37 replies to this topic

[Thalamus]

Those round prims of yours is imo essential in getting the "round" feeling. it is very very hard to create a rubber in vpx that is actually round. And it shows for a lot of games. Especially top of the slings. I guess the problem with "not round" rubbers exists everywhere. But, you tend to notice more when it hits the top of the sling because of the possible drain to the outlane.

 

When talking about rubbers - yeah, don't be corny now. The pinball rubbers

 

BorgDog mentioned it years ago. The height of the rubber has quite a lot to do with the physics. I guess It kind of makes sense though. If a ball is standard 50 units. I do understand why we see odd things happening when the ball hits lets say a rubber band that is higher than the mid point of the ball at 25. So many tables have 30 or there about and they tug the ball like a girl you just don't want to let go off.

 

It is not to say that rubbers, as is can't be improved, you guys have imo absolutely valid points. But, if you are comparing physics, be very sure that you also are using the same height in the test cases. It is essential for how the rubber responds.

[batch]

The problem is that the rubber has a different elasticty depending where the ball hits it

 

Perhaps we can have several walls (or prims) with different settings

 

[Schlabber34]

I do get the round rubber prim thing and yes, it seems to be better!

Ben was right. The fall off is a approximation of the material compresion and the change of physics that comes with it! When you compress a rubber it is very soft and bouncy at low preassure but if you raise the preassure the rubber gets harder. A harder, compressed rubber material can't store energy in the same way as a soft, bouncy one! A ball that hits the rubber at low speed will get let's say 90% of the energie back from the rubber, at mid speeds maybe 70% and at high speed only about 60%. The fall off simulates that behaviour. The higher the value (0 to 1) the more the rebound energy will be reduced at higher speeds!

I mostly use a elasticity value of 0.8 - 0.84 for rubbers and a fall off of 0.3 - 0.4 depending on the table.

If you don't use fall off the table might play like an oldschool videogame pinball. The ball will sometimes reach speeds that are totally out of control and the ball will alot of times feel way too light. More like a golf ball.

Gesendet von meinem SM-G955F mit Tapatalk

[rothbauerw]

@batch You're right, but that's too much work   Besides, with the other issues discussed here, that difference in behavior would not be noticed by most, if any, players.  Worth considering though if we ever dive into the the physics handling of rubbers.

 

@Schreibi34 The Elasticity curve was something that nFozzy was never happy with, specifically for posts, sleeves, and pegs.  That's part of the nFozzy physics I apply to my tables.  It has more realistic, although I'm sure not perfect, elasticity fall-off behavior.  It looks something like this for posts and pegs:

 

Velocity 0 Elasticity 0.935

Velocity 3.77 Elasticity 0.935

Velocity 5.76 Elasticity 0.942

Velocity 15.84 Elasticity 0.874

Velocity 56 Elasticity 0.64

 

Multiply the elasticity by 0.85 for sleeves.

 

What it's still missing is the compression (acceleration/deceleration) Ben is referring to.  I don't see any way to address that outside of the physics engine.  I get comments from many people that there is an acceleration/deceleration aspect missing from VP tables, so Ben is not alone.  I always had a hard time putting my finger on what they were actually referring to, but I think this whole rubber discussion is probably the crux of the issue!

[batch]

Why are Slingshots allways Walls, couldn't they be Primitives like Rubbers ?

 

Don't you think they are the more important Rubbers of a table ?

[LynnInDenver]

Why are Slingshots allways Walls, couldn't they be Primitives like Rubbers ?

 

Don't you think they are the more important Rubbers of a table ?

I do find it an interesting choice, but I can see how it works from an engine standpoint - technically a slingshot is a device *behind* a rubber ring on a real pinball machine, so it makes sense that it's a separate device in the engine from the rubbers. I do agree that it's odd that it's part of the Wall element specification instead of being a separately implemented element, but that's probably a decision that was made back before VP9.

[rothbauerw]

I think it would be really cool to architect VP to have physics plugins.  So over time we could add new physics elements handled by the physics engine without a new release of VP.  Want to change how rubbers are handled.  Build a plugin.  Want separate slingshot element.  Build a plugin.  Want to change flipper physics.  Build a new flipper plugin.  Need a pop-up trap door or lifting ramp.  Build a plugin.

[Schlabber34]

I've finished the VPX high speed video captures! First, for those who missed it, the high speed video of a real table from randr: 

 

https://www.youtube....eature=youtu.be

 

First i wanted to say that i'm NOT creating new physics for VP!! All i do is share my experience and my knowledge about physics IRL and in VPX!

 

So here are the first videos from my Hot Tip table with a simple bounce off the top of the sling rubber:

 

 

First shot PF friction 0,045:

 

 

https://www.youtube....h?v=Q3uW6UDNVvE

 

 

Second shot PF friction 0,045

 

 

https://www.youtube....h?v=MepHAl799Pc

 

 

Third shot PF friction 0,15

 

 

https://www.youtube....h?v=5JI5lgu2pII

 

 

We can clearly see that the ball rotation is stopped on impact! That is with a low rubber friction of 0,6! We can also see that it takes a while before the ball starts rotating after the rebound even with the PF friction at 0,15!

 

Now soft plunger push with PF friction 0,045:

 

 

https://www.youtube....h?v=saMUXzanPFU

 

 

The ball seems to skid before he starts to rotate!

 

 

A simple post pass:

 

 

https://www.youtube....h?v=D6D27koz21U

 

 

The last one is really interresting! I was not able to pull this one off with the standard flipper settings i use! It's just a little wiggle of the ball when cradled. With the settings i used every tap endet in a violent post pass. So i fooled around with the coil ramp up EOS torque and return strength and was finally able to pull it off. These are the settings i used for Hot Tip:

 

Mass                      1,1

Strength              1700

Elasticity                0,6

Elasticity FO        0,35

Friction                 0,25

Return Strength   0,09

Coil Ramp up       0,75

SA                             0

EOS Torque         0,18

EOS TO Angle          6

 

https://www.youtube....h?v=JEtEvKmrRhg

 

 

The good news is, to me they play better and even more responsive then the old settings. Drop catches are easy to pull off even for me as a lousy pinball player! I haven't tried live catches yet! Nothing here is fine tuned! It's just what i discovered this afternoon. Those numbers might be old news for some of you, but i wanted to make sure that some people try these on the HT table and maybe tune them to perfection!

 

The only downside is, that the flipper moves down when hit by a ball (in up position, flipper button pressed).

 

 

I hope this helped a bit and like i said, this is not a new physics religion, this is just stuff to discuss!

[Ben Logan]

Here’s one more hack idea to possibly simulate acceleration / deceleration dynamics related to high energy ball impacts with rubbers...
 

Someone mentioned that high playfield friction settings can yield this effect to some degree, but also that high playfield friction can unfortunately have some other undesired effects.

 

This idea may be bonkers, but to mitigate the undesired effects of high friction playfield and keep the desired effect on acceleration / deceleration dynamics... What about the possibility of a “Dynamic Variable Playfield Friction Curve?”

 

Playfield friction could temporarily scale up after high speed impact with rubber (degree of scale up could be governed by strength of impact), and scale back down to normal setting as function of distance of ball travel. This might be one way to get the desired ball slowdown that happens as a result of rubber compression on a real table, rather here as a result of temporarily scaled up playfield friction curve. 
 

Ball leaves rubber —> playfield friction temporarily scales up. Ball travels along playfield—> playfield friction value dynamically scales back down toward baseline as a function of distance traveled 

 

Any thoughts? Maybe the amount of math required for calculation would melt our CPUs?! 

Edited by Ben Logan, 26 April 2020 - 01:02 AM.

[rothbauerw]

Now you're starting to highlight another missing piece to VP physics.  We only use one friction value.  In reality, there are two, static and dynamic.  I don't think messing with friction values will give us the more realistic rubber behavior we'd want to achieve.  That is going to require a non-rigid body model.

[batman444mab]

This topic is very interresting.

As i disassembled my old rusty Hollywood Heat playfield, i can drop a dotted ball on a rubber post ( used rubber ), film it at 1080p 60fps and make it slowmo to study the right behavior. If anyone need this to help, just tell me

[Thalamus]

The more info we have, the better @batman44mab - so, please, if you've got the time we would like to see.

[batman444mab]

@Thalamus

Allright, here is a quick video on rubber post on my ultra savaged HH saved from a basement. the cab was plenty of moisture etc...

My camera isn't fast enough to get slowmo over 60fps over a 25fps footage.

I slowed down a 60fps video to 40% on a regular 25fps one.

https://youtu.be/bB8n-w7OwvY

 

New rubber should have way more friction

Edited by batman444mab, 27 April 2020 - 03:50 PM.

[Ben Logan]

Thanks so much Batman and Schreibi for your videos! They make great pieces for comparison. 

[Schlabber34]

The ball seems to roll more and slip less after it hit the rubber! There is no noticeable slowdown from sliding and no curveball. My feeling that the ball starts rotating more easily could be right.

Gesendet von meinem SM-G955F mit Tapatalk

[Thalamus]

@batman44mab : Thank you for making those !

[rothbauerw]

Keep in mind, these are low speed impacts.  I'm sure you'll see more sliding with high speed impacts.

[Schlabber34]

Yes, but the VPX videos were also taken from low to mid speed impacts.