Playing rhythm games, I can easily feel ~3ms of input lag. However it's pretty easy to adapt even to 100-200ms of lag, so I'd say most people are in the audiophile placebo territory. I can hit a 5-8ms window reliably even with a ton of lag if I spend 20 mins getting used to how it feels (running a game at 10fps under wine on a 10 year old chromebook with bad wireless gear). Network lag is far more significant than any input lag due to input processing or display pipelines.
Are you mixing up jitter and lag? Even if you hit perfect 120fps synced to refresh, your input to display lag goes up to 8ms. And that's ignoring the input device itself, processing time, etc. which also stack up. 3ms is not really doable with typical modern hardware.
Playing rhythm games, I can easily feel ~3ms of input lag.
Hrmmm. I’m not sure this sentence even makes sense. I’m not sure what the definition of 0ms input lag even is. And I say this as someone who has deeply profiled true end-to-end input/output systems. 0ms relative to what exactly?
If you spend 20 minutes getting used to something you could maybe tell the difference if 3ms of input lag was artificially injected. Maybe. But that’s an exceedingly small value and if I had a toggle switch that turned it on/off I’d bet against you being able to tell when it’s on.
It does depend on the context. In a game like Guitar Hero 3ms is going to be exceedingly difficult to distinguish imho.
Humans are incredibly sensitive to audio rhythms/lag, and we’re talking about a rhythm game. Visual processing isn’t likely to notice, but this is probably between hearing/feeling the keystroke and hearing the reaction. 3ms is entirely believable; detection is based on a frequency domain transformation of the overlaid waveforms, not tuning as such.
Yeah I don’t buy that. It’s a long pipeline and there’s almost definitely more than 3 milliseconds of jitter. I’d wager that success/failure sounds are quantized to audio buffers that are a good bit longer than 3ms.
This is a deep and interesting enough topic that you could publish a pretty good and valuable paper on what the JND is for poor, average, elite, and “professional” players is.
If the line is 3ms for pro I will gladly bet the over!
Have you ever had the experience of having a malfunctioning phone system play back your own voice to you on a slight delay? I have. It's maddening to the point where I find it difficult to continue the phone conversation, and I'd bet that's on the order of 3-5 ms audio delay.
You can set up a double bound really easily for this. Play on speaker, and have someone move the speaker a couple of metres. See if you really can tell whether it has been moved.
After reading some of the science on auditory perception it seems you are wrong. If you perceive a signal in the left ear and then get a louder signal on the right ear a few milliseconds later you will perceive this as one sound coming from the right. The precedence effect breaks down.
Edit: I can also tune a guitar by plucking two strings at once, but I can't by plucking them one at a time.
This really has nothing to do with noticing the absence or presence of a 3ms latency difference in a situation where your baseline is already dozens of ms of latency in the best case.
Adapting isn't the issue though? It just feels much nicer to play on my 144Hz monitor than it does on my 60Hz monitor. This is first-world problems territory at most. It's not just gaming either. Doing anything on my 144Hz feels so much nicer.
Do you notice the difference in audio/video sync between sitting in the front or the back of the cinema?
rhythm games are literally scoring your timings, a cinema offers no such equivalent of a scored feedback loop.
No, I wouldn't notice 3ms on the lips of an actor. Drummers are known to be able to perceive 1-3ms differences in timing; again, a tactile action.
If a human is involved in the action their sense of rhythm and prediction takes over to allow for manipulation within those timescales. It's not uncommon, either.
I do absolutely not doubt a 3ms difference in timing being perceptible with two rythmic signals. I absolutely do doubt a 3ms latency though.
You are claiming that someone can tell the difference between playing the game oneeter or two metres from the speakers.
Put another way, on a large stage the sound from the bands monitors takes ~10ms to reach the drummer.
I don't have data, and neither does anyone else here apparently, but in the absence of evidence it seems exceedingly likely that any perceived differences due to 3ms latency (difference ?) are placebo. We know that placebo is real and large so unless someone has an experiment that tries to account for this (e.g. properly blinded) my prior is that you can't tell when playing a rythm game whether the speaker is 1m or 2m from you.
There's also a difference in volume and freq response due to the head being in the way and also the shape of the ear that helps you determine the direction.
I don't think we could tell direction based on the distance and less than 1ms difference alone.
Comments
Top-level speedrunners of action games are definitely able to feel input lag changes on the order of 10-20ms.
For everyone else, you'd really want to prove it with a high-speed camera or something.
Playing rhythm games, I can easily feel ~3ms of input lag. However it's pretty easy to adapt even to 100-200ms of lag, so I'd say most people are in the audiophile placebo territory. I can hit a 5-8ms window reliably even with a ton of lag if I spend 20 mins getting used to how it feels (running a game at 10fps under wine on a 10 year old chromebook with bad wireless gear). Network lag is far more significant than any input lag due to input processing or display pipelines.
Are you mixing up jitter and lag? Even if you hit perfect 120fps synced to refresh, your input to display lag goes up to 8ms. And that's ignoring the input device itself, processing time, etc. which also stack up. 3ms is not really doable with typical modern hardware.
Hrmmm. I’m not sure this sentence even makes sense. I’m not sure what the definition of 0ms input lag even is. And I say this as someone who has deeply profiled true end-to-end input/output systems. 0ms relative to what exactly?
If you spend 20 minutes getting used to something you could maybe tell the difference if 3ms of input lag was artificially injected. Maybe. But that’s an exceedingly small value and if I had a toggle switch that turned it on/off I’d bet against you being able to tell when it’s on.
It does depend on the context. In a game like Guitar Hero 3ms is going to be exceedingly difficult to distinguish imho.
Humans are incredibly sensitive to audio rhythms/lag, and we’re talking about a rhythm game. Visual processing isn’t likely to notice, but this is probably between hearing/feeling the keystroke and hearing the reaction. 3ms is entirely believable; detection is based on a frequency domain transformation of the overlaid waveforms, not tuning as such.
At the same time, sure, we easily compensate.
Yeah I don’t buy that. It’s a long pipeline and there’s almost definitely more than 3 milliseconds of jitter. I’d wager that success/failure sounds are quantized to audio buffers that are a good bit longer than 3ms.
This is a deep and interesting enough topic that you could publish a pretty good and valuable paper on what the JND is for poor, average, elite, and “professional” players is.
If the line is 3ms for pro I will gladly bet the over!
Have you ever had the experience of having a malfunctioning phone system play back your own voice to you on a slight delay? I have. It's maddening to the point where I find it difficult to continue the phone conversation, and I'd bet that's on the order of 3-5 ms audio delay.
https://arxiv.org/abs/1202.6106
". In general, human speech is jammed by giving back to the speakers their own utterances at a delay of a few hundred milliseconds. "
Trust me. It's way less than a few hundred ms.
That’s completely different from being able to identify 3ms of input lag.
Oh, I’m expecting there’s 3ms of jitter. On a good day.
The total lag is almost certainly an order of magnitude above.
You can set up a double bound really easily for this. Play on speaker, and have someone move the speaker a couple of metres. See if you really can tell whether it has been moved.
Can you tell direction of a sound just by turning your head? Congrats, you can detect the difference of less than 1ms.
After reading some of the science on auditory perception it seems you are wrong. If you perceive a signal in the left ear and then get a louder signal on the right ear a few milliseconds later you will perceive this as one sound coming from the right. The precedence effect breaks down.
https://en.m.wikipedia.org/wiki/Precedence_effect
With simultaneously present signals...
Edit: I can also tune a guitar by plucking two strings at once, but I can't by plucking them one at a time.
This really has nothing to do with noticing the absence or presence of a 3ms latency difference in a situation where your baseline is already dozens of ms of latency in the best case.
Adapting isn't the issue though? It just feels much nicer to play on my 144Hz monitor than it does on my 60Hz monitor. This is first-world problems territory at most. It's not just gaming either. Doing anything on my 144Hz feels so much nicer.
as long as the lag is consistent, then I can concur :)
nothing is worse though than a rhythm game that can't keep proper rhythm itself.
Sorry, but:
3ms is how long it takes sound to travel a meter.
Do you notice the difference in audio/video sync between sitting in the front or the back of the cinema?
rhythm games are literally scoring your timings, a cinema offers no such equivalent of a scored feedback loop.
No, I wouldn't notice 3ms on the lips of an actor. Drummers are known to be able to perceive 1-3ms differences in timing; again, a tactile action.
If a human is involved in the action their sense of rhythm and prediction takes over to allow for manipulation within those timescales. It's not uncommon, either.
I do absolutely not doubt a 3ms difference in timing being perceptible with two rythmic signals. I absolutely do doubt a 3ms latency though.
You are claiming that someone can tell the difference between playing the game oneeter or two metres from the speakers.
Put another way, on a large stage the sound from the bands monitors takes ~10ms to reach the drummer.
I don't have data, and neither does anyone else here apparently, but in the absence of evidence it seems exceedingly likely that any perceived differences due to 3ms latency (difference ?) are placebo. We know that placebo is real and large so unless someone has an experiment that tries to account for this (e.g. properly blinded) my prior is that you can't tell when playing a rythm game whether the speaker is 1m or 2m from you.
Sound travels distance between ears in less than 1ms max. Are you able to detect direction of the sound with your ears?
There's also a difference in volume and freq response due to the head being in the way and also the shape of the ear that helps you determine the direction.
I don't think we could tell direction based on the distance and less than 1ms difference alone.
I have both signals coming in at the same time and can superimpose them. Completely different situation.
Like tuning strings by hearing when you can pluck both strings at the same time, vs doing it using absolute pitch.
That's about phase, not frequency.
It sounds insane, but I agree with him, it's absolutely perceptible.