357 replies to this topic

[toxie]

I actually don't think that one would notice the difference with DOF and sound in that case (e.g. wiring that stuff to solenoid still). But thats up for experiments.

But also every lag decreasing changes you have are highly appreciated. 

 

If i find some time i'll try the timer solution today. I'm pretty convinced that this should work almost flawlessly.

[Brer Frog]

Drybonz -

 

Thanks for the information. I remember seeing your setup when I was looking for ways to rotate my desktop monitor in order to play FS tables.

 

Then I started thinking why not just build a mini-pin cab & use a 27" playfield monitor. Your use of a 27" got my attention & I'll look into the Acer monitor.

 

Thanks.

[StevOz]

Great topic, I went ahead and experimented with all the options and some with a finally rewarding result here. Firstly my PC specs Core i3 Haswell running @ 3.4Ghz with a nvidia 660, it runs VPX tables silky smooth at 60 FPS with VP settings mostly maxed and vsync set at 1, that said it cannot maintain over 120fps rates.

 

To cut to the chase I changed VP vsync to 2, adaptive, noticed that the flippers indeed had less lag, like so much it was almost light and day, though this introduced slight ball stutters during the game.

 

So I set the VP vsync to 0, then within the nvidia control panel I adjusted it to adaptive, same result then I noticed the other option...Adaptive ( half refresh rate ) after setting this option I had the same great flipper response without the slight ball stutters.

 

So in conclusion this was a solution for me and maybe some others...pics included or it didn't happen.

 

 

Test table South Park during 5 ball multiball...

 

 

Sweeeeet!

 

[Noah Fentz]

I didn't have time to catch up on the entire topic, but this is something I've been pointing out for years, and it's not button latency. In fact, it has nothing at all to do with button input. The problem lies in the actual speed of the flipper rotation itself.

Any doubters, feel free to download any 921+ table and adjust the flipper speed from the (roughly) .455 to .955. I'm not sure how the .455-ish value became the commonly accepted speed by authors, but when I did shows, people would make the same comment about going from real to virtual. Once I started using .955, those comments were rare to non-existent.

I was chatting with Dennis Nordman one year, and he asked if we had WH2O, so I downloaded it. Without thinking about it, I didn't adjust the flippers. He wasn't enjoying it at all and commented about 'lag'. I exited the table, adjusted the flippers, and when he played again, the biggest smile appeared on his face. He had a blast and commented the difference was night and day and felt right after that.

Sent from my SM-G930T using Tapatalk

[Knorr]

Paul is right. I think its almost a year back when i talked to toxie/fuzzel if its possible to change/reduce the strenght/mass parameter in relation to the rotation. When i adjust the flipper side by side with my other pin i end up with Mass 0.7 and strenght 10000...

[chepas]

I haven't really got any fix and really never tried to fix.

 

When I play a game, I think it's fine, but that is basically down to me being used to it at this point.

 

When I let someone else play they are around 0.2ms out when pushing the flipper, they always push late on my games, I know these guys can play real pins ok, so it's not their reactions.

[Kernel]

@Toxie: Do you think this vpinmame lag con be linked with the slow sample speed? http://www.vpforums....topic=35874&hl=

 

@gtxjoe, great suggestion!! Thanks!

 

 

I didn't have time to catch up on the entire topic, but this is something I've been pointing out for years, and it's not button latency. In fact, it has nothing at all to do with button input. The problem lies in the actual speed of the flipper rotation itself.

Any doubters, feel free to download any 921+ table and adjust the flipper speed from the (roughly) .455 to .955. I'm not sure how the .455-ish value became the commonly accepted speed by authors, but when I did shows, people would make the same comment about going from real to virtual. Once I started using .955, those comments were rare to non-existent.

I was chatting with Dennis Nordman one year, and he asked if we had WH2O, so I downloaded it. Without thinking about it, I didn't adjust the flippers. He wasn't enjoying it at all and commented about 'lag'. I exited the table, adjusted the flippers, and when he played again, the biggest smile appeared on his face. He had a blast and commented the difference was night and day and felt right after that.

Sent from my SM-G930T using Tapatalk

I too feel the problem is also here. Changing flipper speed from .455 to .955. is also valid for VPX tables?

[toxie]

@Noah, etc: Of course flipper rotation speed is one factor of all this (see the initial post that started the thread), but one can only increase it so much. Then at some point it becomes unrealistic. So increasing it beyond that limit will of course also "decrease lag" in some sense, but at the price of less realistic ball<->flipper behavior.

 

Thats why we should decrease all the other problems first, as best as possible, and then the remaining input lag could still be compensated by a bit of unrealistic behavior.

 

@Kernel: In some sense, yes, this is the part that DJRobX wants to tackle. It will cost some performance, lets see how much though.

Edited by toxie, 13 October 2016 - 02:37 PM.

[vampirolatino2]

I like all this debate. Good findings and minds working together.

[Ben Logan]

Great topic, I went ahead and experimented with all the options and some with a finally rewarding result here. Firstly my PC specs Core i3 Haswell running @ 3.4Ghz with a nvidia 660, it runs VPX tables silky smooth at 60 FPS with VP settings mostly maxed and vsync set at 1, that said it cannot maintain over 120fps rates.

 

To cut to the chase I changed VP vsync to 2, adaptive, noticed that the flippers indeed had less lag, like so much it was almost light and day, though this introduced slight ball stutters during the game.

 

So I set the VP vsync to 0, then within the nvidia control panel I adjusted it to adaptive, same result then I noticed the other option...Adaptive ( half refresh rate ) after setting this option I had the same great flipper response without the slight ball stutters.

 

So in conclusion this was a solution for me and maybe some others...pics included or it didn't happen.

 

 

 

 

Another great idea. Can't wait to give this a try. Thanks for the screenshot, StevOz! 

 

I'm having so much success with gtxjoe's code and video settings tweaks, VP feels like a whole new experience. Low voltage flips ("flipper tricks" stuff) even benefits hugely here. Formerly, I found it really challenging on my machine to let go of the leaf switch, and then press again quickly enough to activate a low voltage flip (whereas on a real machine, especially a Williams solid-state, this is totally easy). With these new settings, even with a high return rate on a flipper, low voltage flips are easy to grab. The flippers on Uncle Willy's new Bad Cats release react wonderfully. He's set those up just perfectly. 

 

 

@Noah, etc: Of course flipper rotation speed is one factor of all this (see the initial post that started the thread), but one can only increase it so much. Then at some point it becomes unrealistic. So increasing it beyond that limit will of course also "decrease lag" in some sense, but at the price of less realistic ball<->flipper behavior.

 

Thats why we should decrease all the other problems first, as best as possible, and then the remaining input lag could still be compensated by a bit of unrealistic behavior.

 

 

I hadn't thought about the notion that increasing animation flipper speed necessarily changes ball behavior. Yikes! That's an important point. In fact, my initial idea was clearly bonkers in light of this idea. 

 

I'm thankful to Noah for pointing out that flipper rotation speed is one more factor available for possible tweaking. I agree with Toxie that designating flipper rotation speed tweaks as a last resort is a sensible order of operation. 

[Noah Fentz]

@Noah, etc: Of course flipper rotation speed is one factor of all this (see the initial post that started the thread), but one can only increase it so much. Then at some point it becomes unrealistic. So increasing it beyond that limit will of course also "decrease lag" in some sense, but at the price of less realistic ball<->flipper behavior.
 
Thats why we should decrease all the other problems first, as best as possible, and then the remaining input lag could still be compensated by a bit of unrealistic behavior.
 
@Kernel: In some sense, yes, this is the part that DJRobX wants to tackle. It will cost some performance, lets see how much though.

Hi toxie,

I wasn't suggesting it was A factor. I'm saying it's THE factor. I have analyzed, dissected, researched, and gathered public opinion. The single biggest concern of callers asking about the VirtuaPin is flipper lag. They've played other makers' machines, and flipper lag seems to be a common perception when the flipper speed is too slow. I don't even show VPX tables on demos because of this issue. It's always interpreted as laggy.

Here's a video I made quite some time ago showing how the increase to .955 made the flippers match real pinball flipper speed. Imagine what .455 would look like in comparison ...

https://youtu.be/JV0ou6GtyOs

VPX feels more like .455 all over again.

Please don't take this as complaining or bitching. Just trying to help.

Thank you.

Sent from my SM-G930T using Tapatalk

[Ben Logan]

Cool, Noah. Great conversation. Exciting possibilities.

Just watched the video. Didn't realize it was a loop. Thought player was nailing that ramp shot over and over!

Video is pretty compelling. VP flipper looks just as fast as the real thing, and doesn't look artificially sped up to my eyes.

Wondering how a jump to .955 would affect ball physics on current tables. Would faster flipper rotation speed necessarily mean increased flipper strength? Or are these separate factors in the digital realm?

Edited by Ben Logan, 13 October 2016 - 07:51 PM.

[Noah Fentz]

Personally, I'd like to see the return of the Flipper Speed setting. It's sorely missed in VPX, in my opinion.

 

In fact, it was the first thing I looked for when I downloaded my first VPX table.

[toxie]

Hmmm.. The VP9 video indeed looks 1:1 the same..

But are all the flipper tricks possible with this kind of setting?

 

We need some more of the VPX table wizards in this thread.. I don't know the state of the art settings for VPX flippers..

[fuzzel]

The flipper speed property was removed when we merged the physmod code into VPX. So the calculation of the rotation speed (flipper speed) is physicaly correct but it introduces some trouble if you just want to tweak it. I didn't check the values and how they work with different flipper physic settings but adding a factor to the final speed value could speed things up (animation wise). Though I don't know what impact that factor might have on the rest

[gtxjoe]

I assume flipper mass and flipper strength ultimately result in flipper speed. Is that an incorrect assumption?

[fuzzel]

Flipper mass is definitely one factor but I'm a bit uncertain about the speed. I don't have the code in front of me so will check that later.

[Noah Fentz]

I would think it best to have the speed accurately represented, first and foremost.

 

The notion of lag is very widespread, and it kills it for a lot of people before they'd even get to flipper tricks.

[BorgDog]

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.

[Ben Logan]

I love that you guys are looking into this. Super exciting. nFozzy's flipper tricks code was an absolute game changer from my perspective. You guys' inquiry here regarding flipper rotation speed could potentially lead to an even bigger improvement in realism. Thanks a million!

Edit: A wise cautionary note from BorgDog above. Good mix of opinions here.

Edited by Ben Logan, 14 October 2016 - 01:11 AM.