Request Realistic flipper physics

smbhax

Active member
Apr 24, 2012
1,803
5
Now that we have fairly consistent lighting, the only major thing the game lacks as a true pinball simulation--aside, perhaps, from ball spin, which Bobby King has at least mentioned in passing as something they have looked / will look into (then again he mentions a lot of things in passing, so who knows)--are flippers that handle in a realistic manner; currently, TPA's flippers can only flip one way: all the way. But anyone who's spent a little time on an actual pinball machine knows that with real flippers, a very quick press will cause them to flip only slightly; this allows tap passes from one flipper to the other, for starters, and probably is what allows for more advanced things like live catches, which simply can't be done with TPA's simplistic flipper operation.

FarSight have mentioned--I think it was on this forum--that they've considered making TPA's flippers pressure sensitive to simulate less than full flips, but that isn't how real pinball flippers work--they aren't pressure sensitive, they are *time* sensitive: whether or not a flip is partial or full depends on how long you held the flipper button down.

I'm sure FarSight, who have disassembled numerous pinball machines to date, know how real pinball flippers are constructed and operate, with their dual coils and all that (for instance: http://stevekulpa.net/pinball/bally_flipper1.htm ), but they seem strangely unconcerned with attempting to model their behavior; I suppose this is the hardcore view that doesn't give additional bullet items to the product description, but I'd certainly like to see them get realistic behavior on the flippers before they work on bells and whistles they've apparently got in the pipeline like challenge modes, head to head modes, UI overhauls, and so forth.

To put it simply, if we can't live catch, the simulation is incomplete.

And beyond that, it would be nice if they paid a little more attention to how flippers from different eras and different manufacturers behave: different flippers have different weights and voltages and so forth that make, say, 1960 Gottlieb flippers feel (and function) differently than 1973 Gottlieb flippers which flip differently than 1980 Williams flippers--and yet, in TPA, the flippers--short, lighting, or full length--all feel essentially the same. Even the better Visual Pinball tables, with their relatively simple flipper physics properties, have this differing feel down pretty well; it would be nice if it was given some attention in TPA.
 

Carl Spiby

New member
Feb 28, 2012
1,756
0
The only way to get realistic physics would be to model the entire flipper assembly, something which mobile devices will not have the CPU power for.

I mention mobile devices as I believe all the versions are built from the same source, they would have to start forking which increases their workload.
 

brakel

New member
Apr 27, 2012
2,305
1
I think they could cheat a little though and make something a little more realistic without modeling the physics of the entire flipper and assembly. They could program to providing certain features like live catching instead of capturing every ounce of flipper simulation. It therefore wouldn't be completely simulated but could provide some more of the experience while being capable of running on the mobile platforms.
 

DrainoBraino

New member
Apr 11, 2012
634
0
The flipper and ball physics definitely need upgraded.

Flippers have always felt wonky to me. Slow, not snappy enough. As mentioned many times in the past, go back and play Pro Pinball, the flippers are so much better, and that was nearly 20 years ago.

Ball physics are strange too. It feels like the ball is stuck to the table. And it's floaty. The early releases are particularly floaty.
 

NeedleBanger

New member
Dec 10, 2013
40
0
Whoa good idea man!! Or throw the option in there for the pros at least. I just became a self-proclaimed intermediate pinballer myself. This game is the ****!!! Easily the best game on PS4 Store (a Slinky would win 2nd place).
 

Exo

New member
Nov 10, 2013
30
0
+1 to smbhax's post (you forgot to mention cradle separations ;) ) and Brakel has pretty much hit the nail on the head.
There are plenty of solutions to this problem which don't put any noticeable extra strain on the cpu.
Maybe Farsight will work on it this year since not a lot happened to the physics last year (and hopefully they'll upgrade the old tables to the newer physics as well).
 
Last edited:

Jeff Strong

Moderator
Staff member
Feb 19, 2012
8,144
2
If the Pro Pinball guys could pull off great flipper physics in the 90's, there's no reason it couldn't be done on mobile since today's mobile CPUs are much more powerful than what we had 15 years ago in desktops.
 

Warped Trekker

New member
Dec 26, 2013
38
0
If the Pro Pinball guys could pull off great flipper physics in the 90's, there's no reason it couldn't be done on mobile since today's mobile CPUs are much more powerful than what we had 15 years ago in desktops.

It appears though they have no interest nor the manpower to split the mobile version from the main console version. This has to be done to have proper ports of each. Look at the main menu and how they showed how it was made in that youtube video. They try to make one menu across devices and resolutions. That simply is not the best way. That's the same as businesses still using old TV commercials formatted for 4:3 aspect ratio, where in today's time, majority of all household tv's are 16:9. Broadcast standard is 16:9. Make a new commercial. You basically have to spend money to do things right the first time. Problem is penny pinching gets you nowhere.
 
Last edited:

brakel

New member
Apr 27, 2012
2,305
1
It appears though they have no interest nor the manpower to split the mobile version from the main console version. This has to be done to have proper ports of each. Look at the main menu and how they showed how it was made in that youtube video. They try to make one menu across devices and resolutions. That simply is not the best way. That's the same as businesses still using old TV commercials formatted for 4:3 aspect ratio, where in today's time, majority of all household tv's are 16:9. Broadcast standard is 16:9. Make a new commercial. You basically have to spend money to do things right the first time. Problem is penny pinching gets you nowhere.

But Jeff was saying that they could do it on the mobiles also. So it wouldn't be split.
 

ER777

New member
Sep 8, 2012
797
0
If the Pro Pinball guys could pull off great flipper physics in the 90's, there's no reason it couldn't be done on mobile since today's mobile CPUs are much more powerful than what we had 15 years ago in desktops.

This is very true. The CPU performing the calculations isn't the hard part, its programming them in the first place that's tough.. but doable. Somebody should get FS to watch that physics demonstration video for Unity that was posted in the other games section. That's what FS should be aiming for.
 

Alex Atkin UK

New member
Sep 26, 2012
300
0
I think a lot of the little things are still way off realistic on PA. Personally I have had very little play time on real pinball machines so its harder to spot exactly what is wrong. However, I have been watching YouTube videos of No Good Gofers and immediately noticed the following.

The lighting seems like blobs of light not actually coming from bulbs. Also the table artwork seems to be "simulating" decal burn from the bulbs which makes it look even more bloomy. Come on Farsight, nobody wants to own a worn out pinball machine, this should be simulating a well looked after table with pristine high resolution artwork.

The ball feels glued to the table, real pinball is a lot more erratic and fluid. Things like how the ball speeds up, its not so linear - although I suspect ball spin has a lot of influence here. This is bizarre, you would think ball physics simulation is right at the top of the list when trying to replicate real pinball tables.

Bumpers and slingshots (particularly the latter) need good animations if its ever going to feel realistic. Combined with the bad ball physics this feels very wrong right now.
 

superballs

Active member
Apr 12, 2012
2,653
2
This is very true. The CPU performing the calculations isn't the hard part, its programming them in the first place that's tough.. but doable. Somebody should get FS to watch that physics demonstration video for Unity that was posted in the other games section. That's what FS should be aiming for.

A lot of people need to disable the complex flippers in the unity 3d pinball sim right now as they are actually quite a drain on performance. I think I saw a 10% drop in FPS on a lot of reports over at VP forums.
 

Exo

New member
Nov 10, 2013
30
0
We are not talking about actual 100% emulation of physics here as that can indeed be quite cpu-intensive but rather about the fact that there are a bunch of simple workarounds that produce the same results.
Even if we were talking about real physics emulation, it only needs a dual-core 2,6 cpu to run at a 500-1000+kHz refresh rate and an i3 cpu (4 years old by now) to run at full-speed.
The slowdowns and bad performance are mainly caused by bad writing on the graphics part and not the actual physics engine anyways (also its a pre-alpha release so what do you expect in terms of coding optimization).

Even if Farsight went that route and included pin-sim physics accuracy in their game it could always be toggled via an option but that's probably not gonna happen due to incompatibility with some mobile devices.
It's a shame that TPA is held back by limitations of inferior devices just because they decide to release the same code without alterations on all platforms but that's understandable from a business point of view I guess.
Maybe they'll at least be able to create some workarounds for the essential pinball moves this year.

P.S.: Here is some footage of the unity vp with an i3 processor : http://www.youtube.com/watch?v=ZjKnKj-AFIM and here are some performance specs from Core 2 @ 2.66, 2GB DDR, 512MB GeForce 8500 GT on the pre-alpha version: TAF - 10 - 50 fps, 500 - 1000 physics | BK2K - 35 - 62 fps, 800 - 1000 physics |Dr.D - 32 - 52 fps, 700 - 1000 physics; FPS drops are mainly caused by the fact that the pixel light count has been increased from 8 to 200 which is hilariously high.
 
Last edited:

ER777

New member
Sep 8, 2012
797
0
The slowdowns and bad performance are mainly caused by bad writing on the graphics part and not the actual physics engine anyways (also its a pre-alpha release so what do you expect in terms of coding optimization).

This would be my belief as well. The mathematics needed to apply a force curve as shown in that video shouldn't strain a modern CPU at all. Of course if not optimized and implemented well any new code can slow things down, even if that code is doing the simplest operations. I think they can get what was shown in that video running smoothly with time.
 

Warped Trekker

New member
Dec 26, 2013
38
0
A lot of people need to disable the complex flippers in the unity 3d pinball sim right now as they are actually quite a drain on performance. I think I saw a 10% drop in FPS on a lot of reports over at VP forums.
You can't compare emulated games vs games created directly for the hardware. Of course emulated games require beefy PC's. Most people running these emulated games are probably trying to run on minimum PC specs that are old.
 

superballs

Active member
Apr 12, 2012
2,653
2
You can't compare emulated games vs games created directly for the hardware. Of course emulated games require beefy PC's. Most people running these emulated games are probably trying to run on minimum PC specs that are old.

Wait how is one game emulated ane one not?
 

Exo

New member
Nov 10, 2013
30
0
The emulation of the PCBs isn't actually the problem here. Both TPA and VP emulate the roms without needing big resources to do it because the PCBs are rather simple in their structure. Warped Trekker is likely generalizing his view towards MAME and the emulation of "newer" consoles like the PS2 or Wii which are quite taxing on the CPU.

Even the emulation of physics isn't straining as mentioned previously as long as you don't aim at emulating it on an atomic level (see PhysX demonstrations like this http://www.youtube.com/watch?v=JcgkAMr9r5o which even a gtx580 released in 2010 can calculate at ease while being a billion times more complex than any pinball sim can ever be).

The important bottleneck is the graphics engine which Farsight has already covered so there is really nothing stopping the development of improved physics besides Farsight not having the money and/or manpower to do it (or they just don't care which hopefully isn't true).
 

smbhax

Active member
Apr 24, 2012
1,803
5
+1 to smbhax's post (you forgot to mention cradle separations ;) )

Yeah I wasn't trying to make an exhaustive list of flipper skills by any means ;). Actually though cradle separations are one thing I can do with some reliability in TPA; they don't seem to absolutely require a partial flip, although I'm sure that would make them more controlled and dependable.
 

Exo

New member
Nov 10, 2013
30
0
Cradle separations (if they do happen by accident) are nothing like their real life counterparts.
I've had a bunch of them happen at random but they're incredibly uncontrolled and all over the place.
Normally one of the balls just jumps up while the other one stays put on the flipper and even if something close to that happens in TPA the ball doesn't bounce off of the one on the flipper with enough force due to the physics being highly inaccurate in that aspect.
 
Last edited:

Members online

No members online now.

Members online

No members online now.
Top