Modern computer complexity has heavy impact on keyboard latency

[HEXUS]

Only the newest gaming PCs can approach fast response of an Apple 2e or Commodore Pet.

Read more.

[yeeeeman]

This is because it is not important really...

[aniilv]

Luu measured the input latency between a keypress and the display of a character in a terminal using multiple runs, using a high speed camera (up to 1000fps = 1ms resolution).

what?!

I think any 14 year old CS:GO player could make a test with more common sense than this rubbish

[tag75]

i loved my acorn electron when I was little and that was before game pads and mouse's ish .

[adidan]

Has anybody in the history of anything actually found this to be a problem?

Actually, don't answer that, I really don't care.

[ZephyrSurfer]

Luu measured the input latency between a keypress and the display of a character in a terminal using multiple runs, using a high speed camera (up to 1000fps = 1ms resolution).

what?!

I think any 14 year old CS:GO player could make a test with more common sense than this rubbish

This comment section today is filled with such ignorance. How would you measure latency across all these decades old systems to modern day computers using CS:GO, or other games reliably. That doesn't make sense.

The approach given is fine although not perfectly accurate. With multiple runs and rounding to the nearest 10ms for end to end latency it is repeatedly reliable though.

[Rubarb]

This is something I've suspected for a while now, and yes it's not a big deal but try loading a lot of processes then firing up a game you will notice it much more... mainly in the way it effects key stokes and even mouse stutter and lag.
I wouldn't say this has anything to do with the keyboard or mouse interface but more to do with the lazy way software is written now, drivers 10 time larger than say windows 3.1 operating system etc.
While most of you say or think this is irrelevant what do you think will happen in the future?

[adidan]

While most of you say or think this is irrelevant what do you think will happen in the future?

By the time it may ever be an issue I doubt we'll even be using keyboards.

[DDY]

I think those who are questioning the relevance or importance of this have missed the point.

This is was an observation that was tested using a reasonable methodology, explained with plausible mechanisms and someone wrote about it - it's science.

I enjoyed the article; https://danluu.com/input-lag/

I liked that it was written in plain static HTML too, and not some ad-ridden blog template spread across 10 pages titled "Computers were better in the 70's and here's why"

[Saracen]

This comment section today is filled with such ignorance. How would you measure latency across all these decades old systems to modern day computers using CS:GO, or other games reliably. That doesn't make sense.

The approach given is fine although not perfectly accurate. With multiple runs and rounding to the nearest 10ms for end to end latency it is repeatedly reliable though.

If you re-read what aniilv said, it was that a 14YO CS:GO player could design a better test, not that he (or she) would necessarily use CS:GO to do it.

i found the article quite interesting, on an academic level, though I agree with yeeeeman about its importance.


He's another couple of flies in the ointment, though.

Somewhere in-between finger hitting keyboard and character appearing on-screen, there's a vast difference between what's going on in the 'black hox' in-between. For a start, a single simple CPU like the 6502 works very differently to moden multi-core, multi-threading CPU's with all sorts of hardware-level context-switching, buffering, cacheing on so forth going on. And without that, much of what we take for granted these days simply ain't possible.

Secondly, inside that black box you have software, too, and the vast, vast bulk of today's software works in an entirrly different paradigm. There's no onject-orient event handling going on in Apple IIe software, for instance, where the PC has to potentially react to a vast array of different events from different sources. Instead, you have linear programming wuth a start, an end, and probably a keystroke handling subroutine looking for menu options or data entry.

Oh, and I still have a fully operational Apple IIe, with a variety of extras and peripherals, bought in the late '70s. I wonder who else does, too?

I'd say one reason these old machines feel faster is our expectations and perceptions.

[adidan]

I think those who are questioning the relevance or importance of this have missed the point.]

Not seeing the point is very different to missing the point.

I'm always open to seeing a point if someone cares to make one.

[iamlorro]

Interesting article, thanks for posting hexus.

Reminds me of a top gear episode a while back where they compared modern cars with the same models from a few years /decade before.
I think it's fair to say that most of us assume that newer= better.

... Until scepticism /realism kicks in!

[philehidiot]

Luu measured the input latency between a keypress and the display of a character in a terminal using multiple runs, using a high speed camera (up to 1000fps = 1ms resolution).

what?!

I think any 14 year old CS:GO player could make a test with more common sense than this rubbish

This comment section today is filled with such ignorance. How would you measure latency across all these decades old systems to modern day computers using CS:GO, or other games reliably. That doesn't make sense.

The approach given is fine although not perfectly accurate. With multiple runs and rounding to the nearest 10ms for end to end latency it is repeatedly reliable though.

I think he's saying that the person deriving a test is a 14 year old CS:GO player and not that you should use CS:GO...

Hooowever, I disagree entirely. What has been complained about is the perceived slowness. I use a PC at work where the time from keyboard to character is around half a second. Some of the letters just don't appear when touch typing as the damn thing can't keep up. This PC was rubbishrubbishrubbishrubbish when it was bought around 8 years ago and is now dying on its arse.

So, if the problem is perceived responsiveness, it makes perfect sense to see if this is real at all with a pilot study looking at the entire process as seen by the person reporting the problem. This process starts with keypress and ends with character appearing within a typical usage scenario. In order to remove the overheads of software to the maximum degree possible, a command prompt seems reasonable. A 1000FPS camera is more than sufficient in terms of temporal resolution and whilst it doesn't help narrow down where the source of the lag is, this is essentially a pilot study looking to see IF there's a problem or if it's just perception. It doesn't need controls like the monitor being the same to stnadardise input lag or whatever as an issue as the complaint is modern hardware Vs older hardware. Adding in controls like this will be useful when trying to narrow down the culprit but there's no need where all you're doing is exploring whether there's an issue or not.

The main source of error would be standardising the keypress especially on different hardware but this is very easy to fix by either just shorting the switch or using some other mechanism to trigger a timer when the key bottoms out (yes, not the same as actuation point but the error introduced as long as the actuation force / speed is sufficient will be minimal and standardised). This would be easily done with a microphone and looking for the click on the audio timeline and using this to begin the timing.

That's my opinion anyway, I'd be interested to hear why you think this is such a poorly designed study? High speed cameras at that frame rate are very accurate. I measure the individual wiggles within someone's heartbeat and we only use around 40FPS most of the time. Really high temporal resolution for us would be around 100FPS. 1000FPS is insanely high. That's like 3 mins of footage from around 0.2 seconds off the top of my head.

[Iota]

Somewhere in-between finger hitting keyboard and character appearing on-screen, there's a vast difference between what's going on in the 'black hox' in-between. For a start, a single simple CPU like the 6502 woris very differently to moden multi-core, multi-threading CPU's with all sorts of hardware-level context-switching, buffering, cacheing on so forth going on. And without that, much of what we take for granted these days simply ain't possible.

It's a shame he can't run the same test using the same methodology with different hardware and the same OS. That would probably give a much better indication of hardware differences.

The OS is the real variable here, even with multiple runs.

[Millennium]

Latency in modern computing has long been a problem. There have been many instances where I've thought my computer is not as subjectively fast as my old BBC Model B or Spectrum 48k Rubbery. It's not something new.

Hopefully manufacturer and users focus on the latency problem and eliminate it, at an OS and a hardware level. I'm sure it can be done, even not taking into account any technological evolution.

[Saracen]

Interesting article, thanks for posting hexus.

Reminds me of a top gear episode a while back where they compared modern cars with the same models from a few years /decade before.
I think it's fair to say that most of us assume that newer= better.

... Until scepticism /realism kicks in!

Hmmm. Good subject, but for a different forum. I'd say that it depends on the relative points of reference between older and newer, and the criteria. For instance, replacjng suspension bushes on my old 70s car cost about £1 per corner for a little rubber bush and took me hslf an hour per side, on my own driveway. That same job on the wife's VW took a professionak mechanic with fully equipped workshop most of the day, cost some £700 and the parts alone were about £300, and a ruddy great sluminium contraption, not an ounce of hard rubber.

In many respects, newer cars are better - faster, more reliable (usually) and more comfortable BUT .... the price is that ability to do jobs at home without specialist eauipment is heavily reduced, and servicing far, far FAR more expensive.

In some ways, that's a good analogy for PCs too, and it seems to me industry is doing everything it can to take the P out of PCs pun intended) with take-control OS's like Win10 on the one hand, and un-upgradeable "devices" like tablets and smartphones on the other.

The PC, as such, is rapidly becoming a niche product for enthusiasts, and I question how much longer that niche will exist? For how much longer will we be able to buy components, build a machine to our spec and choose our OS? Or will we end up with a choice between offerings from a few a mega-corps (Apple, MS, Google, whatever)?