Measuring Controller Lag in Console Gaming

Terry Olaes

I Used to be the [H] News Guy
2FA
Joined
Nov 27, 2006
Messages
4,646
Readers that love the minutiae & methodology of our reviews will enjoy this EuroGamer article. They take a look at devising methodology to measure controller input latency in console gaming, a real concern in the wireless controller world. No one likes a laggy game that feels unresponsive.

"Our mantra of '60FPS 60FPS 60FPS!' would all be for nothing if we had horrible input lag," says Infinity Ward's Drew McCoy. "It is extremely helpful being able to see the physical, measurable, result of what is going on in our game - especially if things change or if someone in the office complains that things 'don’t feel right'. If anyone cares about the end user experience of their game, they should be heavily invested in their input latency."
 
Wired controller FTW!

Can someone with more of an EE background please explain to me why wireless controllers experience more lag than wired controllers?

I mean...light travels faster than electricty, right?
 
Can someone with more of an EE background please explain to me why wireless controllers experience more lag than wired controllers?

I mean...light travels faster than electricty, right?

Electricity moves at the speed of light. Radio waves don't.
 
Electricity moves at the speed of light. Radio waves don't.

Radio waves ARE light - just a different part of the spectrum. Light IS faster than electricity, but with wireless controllers there is additional A/D processing on both ends, vice the direct communications with wired controllers.
 
Radio waves ARE light - just a different part of the spectrum.

I wouldn't say they're light, but more that radio waves and light are both EM waves. As for difference in wired/wireless signal speeds, it's perfectly negligible for gaming purposes and calculating lag.
 
I thought radio was slower also but i was wrong .

Radio waves / light travel 186,000 miles per second. Both are electromagnetic radiation.

Electricity travels 96% the speed of light since there are electrons involved. Electrons are matter/mass and there for cant do it.

So in theory radio could be a better signal to use for controllers. As long as you could negate the time it takes to process the signal. IMO the current controllers are plenty for console gamers. Its not like they need it any faster for the games they are playing. If any game has lag its bad programming in the software.

Now instagib on Quake live is another story.
 
Electricity travels 96% the speed of light since there are electrons involved. Electrons are matter/mass and there for cant do it.

IAAEE.

Electrons only travel in the order of a few millimeters per second in typical electric fields you might find in your household. It is the electric field which travels at the speed of light.
 
Can someone with more of an EE background please explain to me why wireless controllers experience more lag than wired controllers?

I mean...light travels faster than electricty, right?

IAAEE.

Wires are typically an insulated from background noise. Higher SNR means you are more reliability able to transmit data through the medium, won't have to rerequest data due to error, and implement extra layers of logic to do extra things like error correction.

In an ideal environment though, wireless should be similarly fast as wired.
 
IAAEE.

Electrons only travel in the order of a few millimeters per second in typical electric fields you might find in your household. It is the electric field which travels at the speed of light.

OK, so that makes the speed of a signal from a wired controller (near) the speed of light correct?
 
I'll add that metal has a higher electrical permittivity than air, which also contributes to the factors I mentioned.
 
I'll add that metal has a higher electrical permittivity than air.

What does this mean? I know golfers get struck by lighting more than they should. But why exactly? Does it have to do with metal being a better source for a ground?
 
What does this mean? I know golfers get struck by lighting more than they should. But why exactly? Does it have to do with metal being a better source for a ground?

well, i know that, but why is what im asking? What is it about the metal that causes that?
 
Radio waves are an electromagnetic wave. Light is an electromagnetic wave. Not surprisingly, what we refer to as "electricity" is also an electromagnetic wave.

All of these things travel at the same velocity through a vaccum. Air slows the radio waves down, and copper slows the electric field down. I'm not sure which medium slows its waves down more, but in any case, since we're talking about distances of 10-15ft, the difference is most decidedly negligible.

The answer therefore must lie in extra processing involved with wireless communications. This tells me the problem of input lag is solvable; better processing technology could potentially reduce it to unnoticeable proportions.

And yes, I just answered my own question. I was hoping someone would explain the details of the extra processing, not tell me radio waves are slower than light. :p
 
IAAEE.

Wires are typically an insulated from background noise. Higher SNR means you are more reliability able to transmit data through the medium, won't have to rerequest data due to error, and implement extra layers of logic to do extra things like error correction.

In an ideal environment though, wireless should be similarly fast as wired.

Ah - good answer. Thanks.

well, i know that, but why is what im asking? What is it about the metal that causes that?

In a metallic bond electrons are free to move among atoms and therefore transmit an electric field much more efficiently. Although my education in chemistry does not go very far.
 
The answer therefore must lie in extra processing involved with wireless communications. This tells me the problem of input lag is solvable; better processing technology could potentially reduce it to unnoticeable proportions.
It's extra processing, combined with the extra noise introduced in over-the-air transmission of electrical fields.
 
This is the reason I don't play many video games anymore. As more technology and electronics are added between my thumbs and the light and sound coming back, control only becomes more frustrating -- and for me control is everything.

People will argue that you cannot perceive the difference, but that is just them. 5 to 10 ms of latency is very noticeable. Considering that most consoles + television sets of today add up at least 40 ms of latency, I consider that unplayable.

I think a lot of people are getting the wrong idea when thinking about the the speed energy rather than the electronic switching in between. The experiments in the article have nothing to do with the velocity of various forms of radiation as that difference there is truly not perceivable within a console and television set. What drags the system down is processing time. Digital processing takes time and translating signals at each input and output of an electronic chip takes time. I'll try to provide a rough example:

  1. Controller button depressed
  2. input muxed via a simple chip for passing down either a cord or into a radio stream
  3. interpreted at the console and demuxed
  4. translated via chip and passed via IRQ
  5. interpreted by the game logic
  6. applied where valid (any amount of innate game based latency may apply here)
  7. new game state queued for a frame of output (update frequency limited)
  8. graphics drawn for that frame to a GPU RAMDAC
  9. one of many video encoding schemes applied (muxed or encoded for HDMI)
  10. passed to a television
  11. demuxed or decoded
  12. adapted for current resolution
  13. frame information stored via Overdrive technology or similar anti ghosting technology
  14. image correction technology applied
  15. LCD activated for current frame
  16. liquid moves, frequencies of light is blocked for colour
  17. image is seen.

The only way to reduce this latency is to go with simpler, shorter forms of data transmission. One solution to avoid electronics and digital processing is to use the energy created at the source (button) and gate it using prisms for the next system rather than translate it to a different signal altogether for each passing step. Some theory behind this potential solution can be found here: http://scitation.aip.org/journals/doc/APPLAB-ft/vol_94/iss_10/101110_1.html
 
Console Gamers don't need to worry about controller lag as much as the other factors going on.

Wireless Controllers:
Xbox360 = normal radio with dedicated channels. Each controller has its own frequency, so I doubt it uses any collision avoidance tech. No encryption or compression. VERY LITTLE LAG.
Playstation 3 = Bluetooth. This means there is data encryption and possibly compression also, all of which take time to process. I'm not entirely sure if all controllers on one system use one frequency channel, but if they do, collision avoidance would have to be implemented just like wifi.

Wifi:
The absolute biggest source of lag on a console gamer would be using wifi instead of wired ethernet. WiFi uses collision avoidance (CSMA/CA) which means the wireless devices must transmit a jam signal first to tell every other wireless device that its going to transmit...then it transmits. Only way to limit this lag is to not use wireless at all.

HDTV:
Depending on the input and the TV, there are all sorts of lag inducing image processing going on. If your hooked up via Component or VGA, there is always additional Analog to Digital conversion happening (unless you have a monster CRT HDTV). Many TV's have image processing to brighten or change the color also. And if you have one of those new 120hz, 240hz, etc. TV's, you are also introducing a huge amount of lag because the input is only 60hz, and the TV has to process the image again to artificially double, triple, etc. the frames per second. Ways to decrease this lag include hooking up via HDMI and turning off any image processing that is going on.
 
The only way to reduce this latency is to go with simpler, shorter forms of data transmission.
There are other ways to reduce the latency in digital electronics. Increasing the clock speed, for one, will reduce the latency of a logical pipeline. There are other methods as well.


velusip said:
One solution to avoid electronics and digital processing is to use the energy created at the source (button) and gate it using prisms for the next system rather than translate it to a different signal altogether for each passing step.
The paper you linked to only discusses optical logic gates, which at best would only be a replacement for electrical logic gates, and has nothing to do with a fundamental change in how controllers would function.

I'll also point out that the energy in current control systems ultimately comes from the source anyway, so that wouldn't be any different.
 
It's been a while since I've read a thread where I learned so much. Thanks guys! :D
 
I wouldn't say they're light, but more that radio waves and light are both EM waves. As for difference in wired/wireless signal speeds, it's perfectly negligible for gaming purposes and calculating lag.

I don't know what physics class you've been taking, but light as a general term in physics refers to the entire EM spectrum. Therefore, radio waves are light, gamma rays are light, microwaves are light.
 
I don't know what physics class you've been taking, but light as a general term in physics refers to the entire EM spectrum. Therefore, radio waves are light, gamma rays are light, microwaves are light.

Light is better used for the visible spectrum. Yes you can call the other EM waves light, but it just introduces unnecessary confusion in a setting where the speed of EM waves is thought to be smaller in air than in cables (see previous posts).
 
This is the reason I don't play many video games anymore. As more technology and electronics are added between my thumbs and the light and sound coming back, control only becomes more frustrating -- and for me control is everything.

People will argue that you cannot perceive the difference, but that is just them. 5 to 10 ms of latency is very noticeable. Considering that most consoles + television sets of today add up at least 40 ms of latency, I consider that unplayable.

I agree.
There is something to be said about playing a game of MS pac man on a CRT monitor with a 4 way on/off switch type switch setup for control. Very accurate. I think this is why i cant stand to play a modern game, its like swimming in mud. But even with all those things you listed (which i agree with) another big problem is the tendency of developers to add weight and physics to everything.

COD4 for example. The recoil on the guns and explosions is ridiculous. Ive been around both, its not even near that bad. It just muddies up the game to the point i get frustrated and just turn it off.
 
Light is better used for the visible spectrum. Yes you can call the other EM waves light, but it just introduces unnecessary confusion in a setting where the speed of EM waves is thought to be smaller in air than in cables (see previous posts).

Actually, the only time I've read something similar along those lines was in a first year biology textbook where it says it'll refer to light as the visible spectrum for whatever purpose.

And with regards to radio waves, they travel at the speed of light in a vacuum. When traveling through a medium such as air, the speed is actually C = 1/sqrt(Permeability * Permittivity). But we consider air as having close to "normal" permeability and permittivity, radio waves travel VERY close to the speed of light with minimal deviation. So yes, they essentially travel at the speed of light.
 
Light is better used for the visible spectrum. Yes you can call the other EM waves light, but it just introduces unnecessary confusion in a setting where the speed of EM waves is thought to be smaller in air than in cables (see previous posts).

There is no confusion to be had if you understand precisely the definitions. It's not my fault if people don't know their physics precisely, they should seek to correct their misunderstandings.

And I disagree that light is better used for the visible spectrum. It is unfortunate that everyday talk refers to light as visible light, but from a physics perspective (which is what this thread seems to have taken on), light means strictly entirety of the EM spectrum. For one to specify the light we see, it would be prudent to specify "visible light". In any case, I guess it's neither here nor there for this discussion, but I felt compelled to make this distinction.
 
There is no confusion to be had if you understand precisely the definitions.
The term light "light" is not precisely defined in physics. Some people will use it to refer to any wavelength of EM radiation, but certainly not all. You act as if there is a rigorous definition for the term, when there is none that is universally accepted.
 
So I think what this is saying is its a safer bet to just get a wired form of input.

Am I right?
 
One thing I neglected to point out is how ironic it is for Infinity Ward to paying so much attention to input lag and such, when they have a glaring problem in the single player for Call of Duty:

BULLET LAG. My god... I can't tell if they're trying to make it a "design feature" by trying to fake bullet travel times, but I think it's horrible. I know for a fact this problem existed in Call of Duty 2 in single player, and it's also present for Call of Duty 4 single player. There have been times where I just flat out stopped playing the game because it was so frustrating to have my shots constantly go behind the enemies I was shooting at.
 
I used my guitar processor with an echo setting of 2,5,10,20,30,40ms all the way up to 200 ms to see if I could hear what I couldnt see. Normally you would never use delay settings that low because you wont hear the slap back. I know this is apples and oranges but I wanted to see if hearing input lag messed with my playing (I muted the original pluck of the string and only the delay signal could be heard).

Once I got above 10 ms on the guitar (real guitar BTW) I could feel that playing single note scales started to feel a bit strange. At 40 milliseconds it began to mess with my timing playing big time. No scientific conclusion here or anything. Just seeing what the difference was between trying to feel input lag in a game and feel/hear it on guitar. It depends on the medium in the end I guess. With games a 50ms lag maybe ok but when playing a musical instrument it is a deal breaker.....

I imagine that all the games we love to play as individuals would feel quite different if it were possible have 0ms of lag considering how obvious 10ms is on a real guitar vs 50ms in a game. I cant feel this lag in games with my current display (could with one I returned recently) but damn was it obvious with a real musical instrument. Ears are faster than eyes ?
 
This is the reason I don't play many video games anymore. As more technology and electronics are added between my thumbs and the light and sound coming back, control only becomes more frustrating -- and for me control is everything.

People will argue that you cannot perceive the difference, but that is just them. 5 to 10 ms of latency is very noticeable. Considering that most consoles + television sets of today add up at least 40 ms of latency, I consider that unplayable.


I read this without reading who posted it and thought, "That sounds like vel." Wow.

I'm in agreement at least as far as the 5-10 ms of latency goes. Between the processing, the monitor, your own brain/eyes, and the controller.. there's a ridiculous amount of lag. I'm always put in shock that can't feel the difference between a 30 ping and a 50+ping. They aren't very well tuned with their electronic appendages..
 
Back
Top