357 replies to this topic

[Noah Fentz]

I would place being able to make shot accuracy way way ahead of speed on the priority list, otherwise we are just back to VP9.  I can get used to a little lag and anticipate it, I can't anticipate flippers that don't shoot consistently and accurately.

Well, if people don't even play it, because it doesn't feel 'right', we're losing more than we're gaining, aren't we?

Besides, I wasn't suggesting speeding them up to be more realistic while sacrificing accuracy. My thought is speed AND accuracy far outweigh advanced flipper tricks, in my opinion. Those could then be honed based on the more realistic flippers.

Sent from my SM-G930T using Tapatalk

[Ben Logan]

Would it be possible to add a flipper rotation speed field to the flipper physics section (if we reached consensus that we wanted such a field, of course)?

This way if an author felt current VPX flipper rotation speed to be optimally accurate, he or she could still work with those settings. An end user could make flippers snappier post-download by changing the value, if that were desired. I'm thinking this might make everyone happy.

Then again, maybe flipper rotation speed is calculated deeper within the engine and can't be isolated easily enough to place in a simple flipper physics field (as with coil ramp up, return speed, etc.) ?

[toxie]

I also do not remember the exact details, will look into it later.

 

As said earlier: Could some of the VPX wizards chime in here?

 

@Noah: Did you try the trick by gtxjoe, just to see if that would be good enough for your needs, without having to resort to the flipper speed setting? Still did not have the time to look into it yet.

[mfuegemann]

Hi,

 

for me a speed setting option is vital for VPX too. Maybe it should be used in desperate cases only, though. At the Moment I have real problems to get the upper flipper shot work for "The Web" in VPX because of the missing speed setting, as all other parameters are not improving the behaviour.

 

Michael

[Noah Fentz]

I also do not remember the exact details, will look into it later.

 

As said earlier: Could some of the VPX wizards chime in here?

 

@Noah: Did you try the trick by gtxjoe, just to see if that would be good enough for your needs, without having to resort to the flipper speed setting? Still did not have the time to look into it yet.

 

I don't have any 'needs', and I would appreciate it if I could provide insight without it being seen as some sort of self-interest agenda.

 

Due to the slow flipper speeds, I prefer VP9 over VPX. I strongly believe, of all that can be as close to perfect as possible, the flippers are at the top of the list. I don't play TPA anymore, either, because the slow flipper speeds created more frustration than enjoyment of the game overall.

 

I play a TON of real pinball. I have literally hundreds at my disposal pretty much any time I want to play. The speed difference between VPX flippers and real is so far apart, many people immediately have a negative feeling about it. I think a lot of us, myself included over the years, fall into the 'feel' of VP being almost a comfort zone and forget we're simulating a real thing.

 

Again, I'm not trying to be negative in any way. Just sharing my personal observation, along with those of many others. I think you guys have done a killer job with VPX!

[toxie]

Still, could you please test gtxjoes suggestion? After all its super-simple to add.

 

I want to collect as much info as possible first, before we add yet another parameter to the already flooded-with-stuff flipper options.

[toxie]

@Noah & mfuegemann: Also, i just played around with the VPX flippers again: Setting coil ramp up to 0 already makes them fast. Then one can increase the strength: Even faster. And if that is still not enough: Decrease mass -> Almost instant move to full rotation.
Can you guys also test this on your end (using latest VPX 10.2 beta, please, so that we test the same stuff)? Is that good enough? Other unwanted sideeffects then?

 

EDIT: The last question i can also already answer myself: Yes.     (really high strength makes weird ball interactions)    Still, please test.

EDIT2: Actually its not weird, and makes perfect sense like-is: Due to the weak coil effect, a ultra high strength will make the flipper bounce back a bit, but then immediately the extremely high strength will shoot the ball away again. This can be solved by tweaking the EOS torque accordingly (=lower value).  So, please test. 

Edited by toxie, 14 October 2016 - 11:25 AM.

[nFozzy]

Lowering mass will move shot trajectory up the flipper. < 0.8 mass, the flippers start to glitch out when balls hit them in the resting position.

[BorgDog]

I still think the bulk of the issue outside people see is not how fast the flipper moves, it's the time before it starts moving.  

 

I sold my Vulcan machine last night, and the guy that bought it was checking out my vp cab and of course fired up the table that, at least on my machine, is about the worst for flipper lag, AFM (not sure whose version).  I found out later that he happens to be the guy currently ranked #5 in the world pinball rankings but it was readily apparent watching him play that the time from button press to flipper starting to move was the biggest issue, and yes he noticed it immediately.  Unfortunately did not have time to play with settings while he was here, but I will definitely give gtxjoe's em flipper method into the script a try at some point, as well as some other flipper settings.  i definitely do not see nearly as much flipper lag in em tables as I do on vpm tables.

 

The guy said he also used to play vp several years ago, and was very impressed with how far it has come, just simple things like rubbers, posts etc as a single object instead of stacks of walls, but in general just how nice it looked as well.  He said he might just have to load it up again and check it out some more.  Nice guy enjoyed chatting with him.  It was also readily apparent watching him test out Vulcan for a few minutes that he is a way way better pinball player than I. 

[toxie]

Lowering mass will move shot trajectory up the flipper. < 0.8 mass, the flippers start to glitch out when balls hit them in the resting position.

 

The latter should not be the case anymore in 10.2. Could you please retry that one? Otherwise please post a repro setting, so i could have a look.

[toxie]

Just did some experiments.. Actually the VP<->VPM turnaround time is pretty drastic.. Most of the time the flipper could already be in the up-position, but instead it just got the initial impulse to start going up..  :/

 

So my suggested simple hack (a timer that resets invalid flipper movement) won't do it.  :/

[Ben Logan]

@Noah & mfuegemann: Also, i just played around with the VPX flippers again: Setting coil ramp up to 0 already makes them fast. Then one can increase the strength: Even faster. And if that is still not enough: Decrease mass -> Almost instant move to full rotation.
Can you guys also test this on your end (using latest VPX 10.2 beta, please, so that we test the same stuff)? Is that good enough? Other unwanted sideeffects then?

 

EDIT: The last question i can also already answer myself: Yes.     (really high strength makes weird ball interactions)    Still, please test.

EDIT2: Actually its not weird, and makes perfect sense like-is: Due to the weak coil effect, a ultra high strength will make the flipper bounce back a bit, but then immediately the extremely high strength will shoot the ball away again. This can be solved by tweaking the EOS torque accordingly (=lower value).  So, please test. 

 

With regard to increasing flipper strength as a means of producing faster flipper rotation...

 

I find myself lowering flipper strength in VP to simulate real world pinball. Seems like authors often choose flipper strength of 2300 or higher for 80s and 90s solid states. Powerful, but mismatches real games on location around here. From my perspective, this is more "modern new-in-box Stern" strength - and fresh out of the box at that!

 

At the risk of sounding like a snob, 2300 or higher often makes ramps too easy to shoot (unless they're really steep). You lose the sense that you're interacting with gravity. If an inaccurate shot can bounce into the side of the ramp and still ricochet its way up and through and out the ramp, flipper strength is too high, imo. With lower flipper strength, you know you nailed it when you travel a ramp. Fist bump moment (which is sad, because I usually play alone!). Or you don't nail it, and gravity sends the ball back down the ramp and SDTM! High stakes = excitement.

 

Haven't experimented with mass. But, I'll try it and report back. I always set Coil Ramp Up to zero. High coil ramp can make flippers feel nice and smooth (especially on EMs), but the cost of introducing latency on the initial flip isn't worth it, imo. 

 

I agree with BorgDog that "the bulk of the issue outside people see is not how fast the flipper moves, it's the time before it starts moving."  But, since even the fastest refresh-rate monitor and GPU constrains us with hardware limitations, it'd be nice to have the option to bump flipper rotation speed up to compensate for unavoidable digital latency.

 

In the end snappiness may be experienced as a function of total time from flipper keydown to full extension. "Making up" some of the time lost in latency between keydown and onset of flipper movement (the time before it starts moving) with faster rotation speed could make for an effective means of compensation. 

 

That said, I also agree that BorgDog is correct in suggesting that if we introduce a bunch of wonkiness into ball trajectory / lose the ability to reproduce shots consistently (a common complaint about Future Pinball) in pursuit of snappier flippers, we'd be doing ourselves a disservice!

 

So far gtx-joe's solution is yielding great results. No negative affect on ball trajectory. If you like to practice live / drop catches, this is your chance! I've tested for hours on VPX Knorr and Clark's Road Show, UW's Bad Cats, Ninuzzu's Super Mario Bros. I love it! Without gtx-joe's code, I could never hit Red from the upper left flipper after ball releases from lock, for example. You've gotta be on your toes for that one, as on the real table the ball travels only about three inches of space before it's time to flip for the jackpot. With gtx-joe's code, it's totally possible. Without it, you have to listen for the sound of the ball release, and "guesstimate" a flip way early to compensate for latency (who does this in the real world?!). 

 

Thanks again everyone for brainstorming about this. It's really fun and exciting.

Edited by Ben Logan, 14 October 2016 - 05:52 PM.

[Noah Fentz]

Can someone please link me to gtxjoe's solution.

 

As for the latency, once I speed up the flippers on VP9 tables, the perceptible reaction is instantaneous. Even the most scrutinizing pinhead has been impressed with our units. Of course, the VirtuaPin PC overkills the required specs, and I optimize for VP like no tomorrow, so that may help in the latency department. Slow flippers are definitely the culprit in my experience, and I've been at this for 6 years and have received feedback from literally thousands of people.

 

I think we're all seeing different results, so ymmv.

[Ben Logan]

Here you go, Noah. Works a treat! 

 

 

 

 

 

Here was jimmyfingers discussion on possible USB polling increases. http://www.vpforums....showtopic=29207

 

The other thing that has been done in vPinmame table scripts to improve flipper response time is to rotate the flipper immediately when the flipper button is pressed, instead of waiting for the flipper button press to be passed to vpinmame and then vpinmame to activate the rom flipper solenoid and then VP rotates the flipper.  This can introduce as much as 30msec lag.  EM tables won't have this delay as vpinmame is not used

 

Here are some example timings of the delay

38582.59 sec: Flipper button pressed

38582.62: Flipper Solenoid triggered 27.34375 msec later

38582.89 sec: Flipper button pressed

38582.91: Flipper Solenoid triggered 11.71875 msec later

38583.2 sec: Flipper button pressed

38583.21: Flipper Solenoid triggered 15.625 msec later

38583.49 sec: Flipper button pressed

38583.51: Flipper Solenoid triggered 15.625 msec later

38583.83 sec: Flipper button pressed

38583.85: Flipper Solenoid triggered 15.625 msec later

38584.24 sec: Flipper button pressed

38584.27: Flipper Solenoid triggered 27.34375 msec later

38584.55 sec: Flipper button pressed

38584.57: Flipper Solenoid triggered 15.625 msec later

38584.85 sec: Flipper button pressed

38584.87: Flipper Solenoid triggered 15.625 msec later

 

 

You can do this by adding this to the top of the  keydown routine

	if keycode = LeftFlipperKey Then leftflipper.RotateToEnd
	if keycode = RightFlipperKey Then rightflipper.RotateToEnd

and this to the top of the keyup routine

	if keycode = LeftFlipperKey Then leftflipper.RotateToStart
	if keycode = RightFlipperKey Then rightflipper.RotateToStart

The side effect though is that the flippers will always be active even when the game is not started or is tilted

 

[Noah Fentz]

So, in a nutshell, it's bypassing the VPM sol calls.

 

Makes sense.

 

Is there any way to poll the software to disable the flippers?

[toxie]

Not to my knowledge. Maybe DJRobX can come up with something. He said he could reduce some of the lag maybe.

 

My current workarounds to 'correct' wrong flips is far from perfect for now. But maybe i can get it to a point where most cases can be handled correctly.

Cause whatever optimizations will be done in VPM, there will always be some kind of delay in the end that remains.

 

Old machines at least had some kind of mechanism to switch of the flipper circuits (during bonus count, etc) that might be able to get easily, but the newer machines flipper circuits were handled by the CPU from what i know, which means that this will not be instantaneous to get.

[Drybonz]

Hey Toxie...  out of curiosity, are the fixes, such as gtxjoe's, for instance, something that can be integrated into VPX, or does it need to be scripted into every table?

[fuzzel]

gtxjoe's fixes can't be integrated into VPX. The flipper handling/controlling is script based. Even if VPM wants to move the flippers it goes through COM/VBS.

[Knorr]

@Noah & mfuegemann: Also, i just played around with the VPX flippers again: Setting coil ramp up to 0 already makes them fast. Then one can increase the strength: Even faster. And if that is still not enough: Decrease mass -> Almost instant move to full rotation.
Can you guys also test this on your end (using latest VPX 10.2 beta, please, so that we test the same stuff)? Is that good enough? Other unwanted sideeffects then?
 
EDIT: The last question i can also already answer myself: Yes.     (really high strength makes weird ball interactions)    Still, please test.
EDIT2: Actually its not weird, and makes perfect sense like-is: Due to the weak coil effect, a ultra high strength will make the flipper bounce back a bit, but then immediately the extremely high strength will shoot the ball away again. This can be solved by tweaking the EOS torque accordingly (=lower value).  So, please test. 

 
With regard to increasing flipper strength as a means of producing faster flipper rotation...
 
I find myself lowering flipper strength in VP to simulate real world pinball. Seems like authors often choose flipper strength of 2300 or higher for 80s and 90s solid states. Powerful, but mismatches real games on location around here. From my perspective, this is more "modern new-in-box Stern" strength - and fresh out of the box at that!
 
At the risk of sounding like a snob, 2300 or higher often makes ramps too easy to shoot (unless they're really steep). You lose the sense that you're interacting with gravity. If an inaccurate shot can bounce into the side of the ramp and still ricochet its way up and through and out the ramp, flipper strength is too high, imo. With lower flipper strength, you know you nailed it when you travel a ramp. Fist bump moment (which is sad, because I usually play alone!). Or you don't nail it, and gravity sends the ball back down the ramp and SDTM! High stakes = excitement.
 
Haven't experimented with mass. But, I'll try it and report back. I always set Coil Ramp Up to zero. High coil ramp can make flippers feel nice and smooth (especially on EMs), but the cost of introducing latency on the initial flip isn't worth it, imo. 
 
I agree with BorgDog that "the bulk of the issue outside people see is not how fast the flipper moves, it's the time before it starts moving."  But, since even the fastest refresh-rate monitor and GPU constrains us with hardware limitations, it'd be nice to have the option to bump flipper rotation speed up to compensate for unavoidable digital latency.
 
In the end snappiness may be experienced as a function of total time from flipper keydown to full extension. "Making up" some of the time lost in latency between keydown and onset of flipper movement (the time before it starts moving) with faster rotation speed could make for an effective means of compensation. 
 
That said, I also agree that BorgDog is correct in suggesting that if we introduce a bunch of wonkiness into ball trajectory / lose the ability to reproduce shots consistently (a common complaint about Future Pinball) in pursuit of snappier flippers, we'd be doing ourselves a disservice!
 
So far gtx-joe's solution is yielding great results. No negative affect on ball trajectory. If you like to practice live / drop catches, this is your chance! I've tested for hours on VPX Knorr and Clark's Road Show, UW's Bad Cats, Ninuzzu's Super Mario Bros. I love it! Without gtx-joe's code, I could never hit Red from the upper left flipper after ball releases from lock, for example. You've gotta be on your toes for that one, as on the real table the ball travels only about three inches of space before it's time to flip for the jackpot. With gtx-joe's code, it's totally possible. Without it, you have to listen for the sound of the ball release, and "guesstimate" a flip way early to compensate for latency (who does this in the real world?!). 
 
Thanks again everyone for brainstorming about this. It's really fun and exciting.

Really? Red Jackpot is easy to hit, you only must flip the ball when it is at the orange traffic sign... Honestly I think it's to easy to hit...

[Ben Logan]

Well, my reflexes are pretty slow!

But yeah -- on my gtx750ti I had everything maxed out graphics wise because your table looks so beautiful. I couldn't for the life of me hit that Red shot! Not a design flaw in your and Clark's recreation, of course, Knorr. It's perfect! But my five year old machine and slow refresh rate Costco TV wouldn't let me nail that jackpot (or the upper ramp in Tommy, for another example) without guess-flipping way early. With GTX-joe code, it's much easier for me. Feels natural.