Help me understand, its crt monitor = no input lag?

[abispac]

The more i read the more i get confused. Why, because some say, a pi4 its better than mister, some say, mister is better than pi4 and some say groovymame its better, other said, jrock fro williams games its the same as the origina hardware, and they all claim to be better with inputlag on crt monitors. Some said lcd are not good, and so on.  So i was wondering, if just having playing old school  arcade games, with vanila mame or mame core or groovymame, if the crt helps to have almost no input lag.  To be honest, i cant tell if my setups have input lag or not. As some said robotron has plenty of input lag in mame but not on jrock, and some said, on a pi4 theres no input lag either, but, how can a smallish pi4 have better management at input lag that an full powerfull mame pc with groovymame and emudrivers, and of course an amd card... can someone help me understand this please?
And what do thet old schools like me think? Its input lag real? Does it bother you?
hanks for your time.

[bobbyb13]

CRTs can draw and display an image faster than anything else.

Good control interfaces, like ones from groovygamegear, ultimarc, or a homebrew thing like Kade on an arduino introduce no lag relative to original hardware.

After that, for classic golden era games, whether you are using a PC, RPi, or MiSTer doesn't really matter as long as it is set up properly.

RpI is easy to set up.
A PC can have more power to run games at original frame rates than an Rpi, but it might not matter for the games you play.
The MiSTer was originally for emualting old computers and consoles but is getting better for arcade cores all the time.

[lilshawn]

you are confusing

input lag

and

display lag

input lag is the time it takes for your button press to be turned into an action on the screen

display lag is the time it takes for a display device to receive an updated frame and change the pixels on that device to that updated frame...

input lag is determined by how fast the button is polled for its state (open or closed), and how that information is coded (if need be) and how it's sent to the computing device for processing. for a usb device that means polling the buttons...building a packet with switch data on it and sending it over the USB chip to the computer USB host chip where it makes its way through other various chips and their associated buses before it gets to a CPU place where it can be decoded and used to control the program. this all takes time.

display lag is determined by how long it takes for a display device to receive frame data and to render that frame on the output device...

an LCD has to receive a whole frame in it's buffer before it changes the entire LCD to reflect that data. commonly called "panel updates" or "whole frame updates" there may be some processing that needs to happen to the frame to make it fit or to find the closest pixels to turn on or off or partially turn on... that takes time. all this is independent of the time it takes for LCD itself to actually CHANGE the pixels and the orientation of the LCD crystals... don't believe the 1ms claims on the box... those numbers are gray-gray measurements... a color-black or black-color pixel change can be much more and the whole process can delay the frame by 20ms easily.

a CRT renders frames as they are being sent to the monitor pixel by pixel, line by line as they come. there is virtually no input lag because the beam is being controlled directly by the data coming in, as it happens. we don't have to process it at all.

of course, all of this is going to depend on how things are controlled, how they are connected...how they are processed. emulation isn't perfect and is introducing its own kind of "lag" and nothing beats dedicated hardware for reducing input to display times...(this is why speedrunners run original hardware on CRT screens if at all possible) and there is always some lag everywhere. even if it's microseconds here and there, but it's getting better and better as computers improve.

honestly, very few people benefit from CRT when it comes to "lag" and most people don't notice a couple MS here or there with using USB encoders

[abispac]

CRTs can draw and display an image faster than anything else.

Good control interfaces, like ones from groovygamegear, ultimarc, or a homebrew thing like Kade on an arduino introduce no lag relative to original hardware.

After that, for classic golden era games, whether you are using a PC, RPi, or MiSTer doesn't really matter as long as it is set up properly.

RpI is easy to set up.
A PC can have more power to run games at original frame rates than an Rpi, but it might not matter for the games you play.
The MiSTer was originally for emualting old computers and consoles but is getting better for arcade cores all the time.

What do you think of the 0 delay usb encoder? its that a good one?

[abispac]

you are confusing

input lag

and

display lag

input lag is the time it takes for your button press to be turned into an action on the screen

display lag is the time it takes for a display device to receive an updated frame and change the pixels on that device to that updated frame...

input lag is determined by how fast the button is polled for its state (open or closed), and how that information is coded (if need be) and how it's sent to the computing device for processing. for a usb device that means polling the buttons...building a packet with switch data on it and sending it over the USB chip to the computer USB host chip where it makes its way through other various chips and their associated buses before it gets to a CPU place where it can be decoded and used to control the program. this all takes time.

display lag is determined by how long it takes for a display device to receive frame data and to render that frame on the output device...

an LCD has to receive a whole frame in it's buffer before it changes the entire LCD to reflect that data. commonly called "panel updates" or "whole frame updates" there may be some processing that needs to happen to the frame to make it fit or to find the closest pixels to turn on or off or partially turn on... that takes time. all this is independent of the time it takes for LCD itself to actually CHANGE the pixels and the orientation of the LCD crystals... don't believe the 1ms claims on the box... those numbers are gray-gray measurements... a color-black or black-color pixel change can be much more and the whole process can delay the frame by 20ms easily.

a CRT renders frames as they are being sent to the monitor pixel by pixel, line by line as they come. there is virtually no input lag because the beam is being controlled directly by the data coming in, as it happens. we don't have to process it at all.

of course, all of this is going to depend on how things are controlled, how they are connected...how they are processed. emulation isn't perfect and is introducing its own kind of "lag" and nothing beats dedicated hardware for reducing input to display times...(this is why speedrunners run original hardware on CRT screens if at all possible) and there is always some lag everywhere. even if it's microseconds here and there, but it's getting better and better as computers improve.

honestly, very few people benefit from CRT when it comes to "lag" and most people don't notice a couple MS here or there with using USB encoders

So what do you think, regular mortals that are not aware of this info , complain about? Like on my robotron group, they all 100% sure that jrock its better than anything, like is the real hardware. Some say pi4 does a good job and some say mister is before jrock. I really would like to test a real robotron hardware to see if it really has an inpact on gameplay....    To be honest, on my end , i think a few milli seconds really dont inpact much on gameplay, on most old school games like from neo geo back. Even my favorites feel just great, circus charlie,mikie,gaplus,kicker, they all feel normal to me. Robotron in the other hand, i did play it when i was young but not anymore on real hardware, and in my country, i bet they all dead, the only hope for robotron for me, would be buying some day a working board. if i dont die first.

[lilshawn]

So what do you think, regular mortals that are not aware of this info , complain about?

really, a combination of all the delays from start to finish. a millisecond or two in the input, a millisecond or two in the processing... hell, a driver that is poorly written could add more delay given identical systems. all of this add up to noticable delays. a pile more more in the display and before you know it, you have 60 or 70 some odd ms of delay from button press till something happens. at this point you are looking at an image 4 to 5 frames behind. if a guy can shave 3/4of that off, it's becomes imperceivable. a CRT is a super easy way of doing it. since the lag there is virtually zero.

you are going to have varying amounts of opinion as to what amount of lag becomes noticeable.

some say up to 100ms of delay is acceptable... while other say literally any is unacceptable.

[abispac]

Yeah, i think most of us that are not really tech savy, we blame a conbination of everything but really we dont know what cause it. Againg , on my end, ive never notice lag anywhere, and im talking since the beginning of mame, i like to think of it as that maybe if it really theres a lag either on input or video, i can get use to it and play with it with no problem, lets say, to just give an example, lets say such lag is so bad, to a 1 second rate, my brai know i have to press the button 1 second before making a jump or whatever, so i get used to it and i can play along with such lag. Of course there is no 1 second lag, but is just an example so i can explain how i deal with it. hope it makes sense.

[Gilrock]

Robotron is my #1 game.  I've got it on a Pi4 with an LCD screen.  I've got it on a Mame PC running Windows 10 with a big TV screen.  I've got it on Mister but that setup doesn't have a dual joystick.  And I've got my scratch build with a JROK, CRT, and original refurbished joysticks.  After playing on the JROK I won't play it anywhere else.  I've gone to ZapCon and played on original machines and it felt the same as my JROK setup.  In fact the joysticks are highly important.  Even on an original machine if the joysticks are worn out or not adjusted properly it will affect your ability to score high.  I'm outta practice but a couple years ago I was getting a million minimum on my bad games and my high score was 5 million.  That's with difficulty 5, 3 men start, new man every 25,000.   If its not an original or a JROK and not those marathon settings I wouldn't even pay attention to someone high score.

[PL1]

What do you think of the 0 delay usb encoder? its that a good one?

The "Zero Delay" branding implies great performance, but it doesn't exactly deliver.   

AFAIK the "Zero Delay" encoders are at best a mixed-bag.
- The boards have been copied by multiple companies.
- Some of those companies use different firmwares with different amounts of lag.

There was a site that someone linked to a while back that tested hundreds of encoders for lag, but I haven't been able to find the link.   


Scott

[abispac]

What do you think of the 0 delay usb encoder? its that a good one?

The "Zero Delay" branding implies great performance, but it doesn't exactly deliver.   

AFAIK the "Zero Delay" encoders are at best a mixed-bag.
- The boards have been copied by multiple companies.
- Some of those companies use different firmwares with different amounts of lag.

There was a site that someone linked to a while back that tested hundreds of encoders for lag, but I haven't been able to find the link.   


Scott

Back in the day when i was single and had extra income, yeah i used to buy groovyarcade products, nowdays i choose the 0 delay for its price...some day ill go back to the real quality...