Over my noodle but very interesting

Damn...thats how you present a case...well done

Random delay... UGH!!

I prefer analog o-scopes, but that's one good use for a digital scope. Sounds like the software takes a while to get around to checking the flipper button status. Maybe it could be patched to check more frequently? Maybe redesigned to use an interrupt? I don't know how an AFMr flipper works, but there must be a fix.

Yep, I noticed that the first time I played a remake which was the launch of AFMr at the Dallas AMOA last March.

I was playing with a couple of top 20 players and they both commented on it too. It really throws your shots off, especially if you're used to the original WPC95 machine.

rd

Great analysis and approach

Not sure I'm buying that 1.5ms difference is that perceivable. You can watch tv with the sound delayed by that much and not notice. Your brain just stitches it together. Famous experiment where they delay a flash by 10ms or so when you click a button. After your brain adjusts to they remove the delay and your brain believes the flash happens BEFORE you press the button.

I agree they will play and feel different, as will any other classic AFM will from each other. 20+ years adds character to a game.

OP, did you measure the angle of the flippers? They are adjustable in the remakes. I know an operator that has both an MMR & AFMR on his route and he has adjusted the angles of the flippers on both and they now play and feel more like the original.

So what causes the delay?!?

Adding software and networking is adding delay.

You should try this test on a new stern and see what they respond at.

Oh, and it's all in the spirit of making MORE MONEY!!!!!!!!

It's enough to be noticeable. If it was a "fixed" delay you would eventually compensate for it after a few games ... if the delay is changing all the time, that's not really ideal.

rd

The additional layer of emulation causes the delay. They should cheat and stop running flipper input through the emulation, but instead program the hypervisor to watch the flipper buttons and activate the coils when the switches close. Can still feed input to the emulator so lane change works, but ignore the flipper output and leave that for the host program.

If they're already cheating and doing it that way, then the beagle implementation is slower than an 8-bit M6809 CPU

I wish I had something technical to say. But I don't. I love my MMRLE and my AFMRLE.

Thank you for this!!!! I've had so many scrubs telling me "same mech" "it's the same" "mine plays great" etc. Nope. No control over the shots! And no way to adjust!!! Interesting findings for sure.

I had an UltraPin back in 2008. Even though everyone said it was a perfect representation to then original games I found if I played it too much it threw my game off on a real pin.

While everyone picks on the old school electronics it is nice to see a hard documenting of the delay introduced,

Amazing thread. Thanks for doing what all of us are too lazy or too stupid to do.

I hope they can fix it somehow, even if it's just for future titles with a hardware change.

I would love if you could do this for jjp and stern games from whitestar/Sam versus spike 1/2

I had always heard a fake Rolex kept time just as good as a real one.

All of this is way over my head, but would pops and slings etc. respond the same way?

I've never noticed any delay on my AFMr. I don't have much experience on the original, but had a guy over who has an original (and had it for years), says it played exactly the same only faster. I guess my brain can't detect a 1.5ms-4.5 ms variance. Some of you guys should be fighter pilots. Being able to detect a 0.0045 second difference in something is pretty impressive.

Really? 1ms isn't enough time for the sound of the flipper button being pressed by your hand to reach your ears... And this is going to throw your game off?

Rolex doesn't use a processor.

Time lag will make a difference.
That's why I don't like virtual pinball.

As I said, you would soon modify your game to adjust to the delay if it was a fixed delay, over the space of a few balls ... no different to how you adjust to different flipper angles etc.

However, the OP explains that the delay changes every time you press the flipper button. This will certainly throw your game off.

rd

Science for the win

That's what MMr reminded me of when I played it.

Don't really understand why they made a new platform for the remakes to begin with. I get they wanted to reduce the number of boards/components etc. but this is the tradeoff of a new platform.

My opinion is they should have found a happy medium kind of like Altek with updated interchangeable boards for the new game but also as replacements for the original as people still need boards for originals. Don't see why that could not have been done and bet they would sell plenty of boards.

Well, not really. Proven that there is a delay, yes.

I think it is a pretty big leap from there to assume that a 1.5 to 4.5ms delay on the flipper coil firing makes the machine play differently. Google around on human response times, the internet "science" on that would seem to indicate otherwise. The op has an opinion, and set out to prove it, hasn't quite gotten a scientific proof yet although the info so far is interesting.

This is so much faster than your reflexes. I mean a single frame of a movie is 33ms or longer. I might be getting old but I can't see each frame at speed.

It's like your game being off in Denver cause the air is thinner and the ball is uncontrollably faster. Yea it's measurable but not really perceivable. Or 50Hz AC throwing off your game cause I grew up playing 60Hz AC power.

But what do I know. Just sounds far fetched if you really think about it. I'll stop.

I'd bet it's due to flipper angle or perhaps the

Anyone want to hear a joke?

2017-09-30-20-23-54- (resized).jpg

Says a person with a machine that does not play like the original

Because of Software / Firmware interfaces I guess it is downhill from now on

I LOVE that the OP used science to try to prove this but, I think all he's proving is that most people don't have any idea just how fast 1 millisecond is.

A bee's wing flaps at 5 ms and everyone in this thread is trying to say that that they can "feel" 1.5-4.5 ms ? Can you grab the wing of a bee when it's at the top of its flap Danielson?

The human eye blinks at a rate of 100-400 ms. That is 100X LONGER than this delay

but maybe this does explain why I'm completely underwhelmed with MM since all I've ever played is MMr

I guess that means no.

Every game I play, plays slightly different even if it's the same game. I'm sure other realize that there are countless other factors that weigh on how a game plays. Not sure 4.5ms would be one of them. regardless, its an interesting bit of info.

I applaud the science here and appreciate the effort to understand the differnces.

Each game in my collection feels different, just like my guitars feel different from each other. I do wonder what humans can actually perceive with such short latencies, and I wouldn’t be shocked if some of the best players in the world can feel the difference.

Marc

Way above my pay grade, but I'd venture to say Stern and CGC use different flipper components than B/W...clear coated PF's will make the remakes "feel" different too.

It is well documented that humans perceive anything under 200ms as instantaneous. So, at under 10ms, absolutely, undeniably undetectable difference.
Interesting experiment though.

CGC uses B/W flipper components.

Stern uses Stern flipper components.

LTG : )

I'm not at all convinced that a 3ms delay would be even remotely noticeable. Try this online test just to get an idea of how fast that is.

https://www.humanbenchmark.com/tests/reactiontime

Is it all coils because of the emulated code and type of boards used compared to original? Or is it just the flipper coils....?

I've been doing some googling on this as well, numerous articles and websites document the average human reaction time is between 150-250ms.

4.5ms? That's 40 times faster! Sure, it's measurable with some test equipment, but not in playing pinball...

I'm willing to bet that a 1.5-4.5 ms delay is absolutely perceptible. Anyone who is into gaming will absolutely notice any latency introduced in flat screen TVs versus CRT. Also, even though a movie is 29 fps, games that run at 60 fps second is absolutely perceptible. Even games that run at locked 60 fps that drop a single frame every couple milliseconds is perceptible to those who really have the feeling of a game down.

***

Anyway, I honestly think that it's very impressive for William's to have been able to have the flippers "CPU Controlled" without introducing any lag. That's damn impressive for the 90s. Also interesting, I didn't know that AFMr and MMr are essentially emulated. Kinda cool, I guess.

How much?

I'd be curious to know the difference, in ms, of the following on AFM:

From a rolling inlane to the right flipper, a shot to the left orbit
and
From a rolling inlane to the right flipper, a shot to the middle of the saucer

The answer to this question will help me decide how much 3 ms matters.

Coming from the video game world for many years I have seen emulations. The generic emulators from a long time ago were awefull. They have gotten progressively better as the years past. That said now the real manufactures put out emulation of there own original product. Hence "Namco" Technology has gotten so advanced that the word emulation is not as dirty as it used to be. I would not claim to be a Pinball expert by no stretch of the imagination. However I do think that when dealing with such slight milliseconds difference it would be very hard for any human to detect. Not to say it is impossible just unlikely. I do find this thread very interesting. Perhaps someone from CGC could pop in and make a statement. I would love to see how they would respond to such a thread. Thanks for the science lesson. However I feel I cant comment until the experts from both sides have an opportunity to speak.

That's the thing, I don't think a human can quantify "how much" it can be perceived, but only that it "feels" different.

If you spend any significant time on a game that drops a single frame every couple milliseconds (or has an unstable frame time), you'll start to notice--even if you can't quantify the difference.

I play online shooters and input lag is a serious issue that I don't take lightly and don't buy certain tv's because of it.
I went to York show with serious thought of getting one of the remakes. Guess why I dismissed both immediately. Yes that millisec of delay in the flippers. Imo I feel it slightly in Williams also but have no scientific proof besides for what my brain tells me. I perceive no lag in Stern flippers,... cort , dlp, and plasma tvs. Fortunately I can buy a new stern, however all new tv tech sucks when it comes to input lag even when game mode is implemented. I strongly agree with op.

Here is a link to a simple reaction test. It shows reaction times.
It is kinda amazing how repeatable your times are. Random 1 to 4.5 milliseconds may actually feel different to the advanced player. I doubt I will ever notice.
https://www.humanbenchmark.com/tests/reactiontime

The question really has nothing to do with how fast people can react to the shot. The bigger question is how long it takes for the ball to roll down the flipper for the shot to change. If the moment I'm flipping is exactly the same, and my shot isn't going to the same place, that's where you would perceive a problem.

The lock shot is a tight shot on Attack from Mars. Is it 4ms tight??? That I don't know.. and you can't really prove that.

My theory is still that the way the coils get power affect the timing of the shot. It feels like the ball stays on the flipper for longer, which causes my shots to be pushed more sideways than I would like, and this changes the angle on every shot. Elwin's theory was that it's lag like you would have with video pinball.

Who has a super high speed camera, something with enough frames to actually measure the time it takes to complete the flipper stroke on an original and a remake, and can do that test? Maybe combine all the factors and you have a less precise game.

Awesome thread and I love that you’re looking at the science of this. Sorry for the long post, so skip it unless you’re interested in the science behind this.

1. Some people have mentioned REACTION time of humans being on the order of 200ms. This is not what the OP is trying to demonstrate by measuring the flipper delay and reaction time has absolutely nothing to do with controlled pinball flips. I say controlled flips, because of course reaction time would play a small part in determining the trajectory of the ball once it rebounds off a sling or post or other object and heads off in a different possibly unknown direction at a high speed. Reaction time is measured as response to an unknown timed stimulus. For example, in the link above (Post #38), the user is asked to tap the screen when the green color appears. But the timing of the green is unknown so you are just “reacting” and this is why it takes 200ms or more to tap the screen. In reality, the subject of hitting a flipper when a ball reaches a certain spot on the flipper is TIMING and NOT REACTION TIME. This is similar to what the OP states in post #45.

2. Neurophysiology research doesn’t always agree on what that the limit of human TIMING is, but there is at least one paper (google Hore and Watts) that indicates that the neurophysiological timing is on the order of 1ms for skilled major league pitchers that have to TIME the release of the ball of their fingertips in order to hit a target 90 feet away at 100mph. It’s interesting to note that other papers demonstrate on the order of 9ms for average humans.

3. But, it’s not the timing delay in the remakes that would potentially be a problem to the top skilled players, because they would just include this delay in their brain when doing the internal “calculations” and adjust their timing such that they flip at the nearly exact moment they want to... well, within 1ms if they’re as skilled as some major league pitchers... and there’s no reason to believe they aren’t just as skilled in their TIMING. The real problem would be that if the delay in the flipper response introduced by the circuitry isn’t consistent. The OP states that there is a standard deviation on the order of 1.5ms and this is exactly the problem if you are a person that has a consistent TIMING response on the order of 1ms. It means that for Keith Elwin who plans to hit the flipper within 1ms of when he wants to, he is really hitting it within about 2.5-3ms of when he wants to.

4. Does any of this make a difference??? Well, I haven’t calculated the ball speed down the flipper (from a cradled position) to see how much distance it travels in 2.5ms. That is the next exercise...

As others have already said (in admittedly much shorter posts), I doubt I’ll ever notice with my average 9ms timing brain... but the timballs of the world may notice!!

Edit: Doh timballs... you posted yours a few seconds before I finished typing. Must be your advanced timing skills.

And I even referenced you while writing my research paper post. Ha.