4 ms·

I have no idea how reliable it is, but quoting wikipedia [1]: "Testing has found that overall "input lag" (from controller input to display response) times of

by Xixi 7y ago

I have no idea how reliable it is, but quoting wikipedia [1]: "Testing has found that overall "input lag" (from controller input to display response) times of approximately 200 ms are distracting to the user. It also appears that (excluding the monitor/television display lag) 133 ms is an average response time and the most sensitive games (fighting games, first person shooters and rhythm games) achieve response times of 67 ms (excluding display lag)." These numbers are from the PS3/XBox 360 area, which certainly sold massively well. Since display lag is not included in these 67ms, this is what these streaming services would have to aim for. It doesn't sound completely insane given the 30ms round trip latencies of home connections. On a more subjective note, I think it's going to be really hard hunting down high end gamer PCs. On the other hand lots of games don't run particularly smoothly on consoles (especially "normal" PS4 and XBox One, rather than PS4 Pro or XBox One S) that these services are directly aiming at. I can perfectly imagine a stadia version of The Witcher 3 running circles around the PS4 version that lags so much... But I don't think it would hold against the PC version. But I don't game much at all anymore, so my knowledge is mostly outdated. The last games I played were 80 days and A Dark Room on my iPhone (both pretty good games). So I guess I'm now more in the target demographics of Apple Arcade. [1] https://en.wikipedia.org/wiki/Input_lag https://en.wikipedia.org/wiki/Input_lag

ehsankia 7y ago
You can probably do even better than 20-30ms round trip if you're one of the large Cloud companies with data centers everywhere. Amazon so far doesn't seem interested, but Google and Microsoft definitely have the infrastructure to pull it off. Sure you won't get high end PC gamers, that's not really the target demographic here. It's the millions of console and mobile gamers.
TheOperator 7y ago
Actual video gamers right now go out of their way to get displays with very low input lags. 1080p144hz TN low input lag displays are used quite a bit in esports to shave a few msec. You literally win more often if you react faster. People aim to get TVs with sub 16ms input lag. However that's not most of the market. The appeal of cloud gaming is for single player non-competitive games. These are the most pirated games and the games where input lag matters the least.
borkt 7y ago
I can easily notice the difference between 20 and 40 ms in FPS games. I have gotten out of FPS's due to keeping up with the hardware requirements, but a streaming service like this could make it possible. That being said, if they cant stay below 50ms I probably wouldn't waste my time.
davemp 7y ago
I followed up on the wiki article. The citation was a just eurogamer article that had little to with human perception of latency. It did have this nugget: “Criterion said it was aiming for 50ms latency [for burnout paradise]” which agrees with what I’ve heard making handheld electronics. The goal was always <50ms latency. Personally I noticed a significant improvement switching to a 165hz monitor in basically every first person game. The only difference here is going from 16ms to 6ms frame granularity (10ms off the worst case). I’m not sure how much is visual smoothness or latency, but I’d gladly take refresh rate/fps over UHD. I’m definitely in the enthusiast group though. I didn’t even realize that console players dealt with such high latency. It will be very interesting to see how the market plays out.
- gmueckl 7y ago
  Hm... the topic is quite complex. Human perception is actually much slower than many people want to believe. Parts of the human vision system react no faster than around 1/25th of a second (other parts are faster - its complicated!). Hearing is similarly divided. AFAIK, auditory and visuak stimuli are processed together and merged over a rather large time window (to the point where one system can induce illusions in the other). But we are able to perceive time delays within a complex stimulus with very high precision. I am not totally convinced that a high refresh rate monitor allows human vision to respond faster. I would rather suspect that most game loops are VSync-locked and a higher refresh rate leads to better input sampling and processing. Maybe I should set up a simple test where I show a non-interactive sequence at various framerates and ask users to identify the highest framerate. And then repeat with interactive camera controls.
  - hnuser123456 7y ago
    >I am not totally convinced that a high refresh rate monitor allows human vision to respond faster. Have you ever watched a high-level FPS player? They can aim and shoot in the span of about ~1/5 of a second, occasionally even faster. That's 200ms. At 25hz, that's 40ms. A big part of it is going to be increased input processing, but a lot of it is muscle memory at that skill level. At 165hz, that's 6ms. With two equally skilled players, one will see the other up to a whole 34 ms sooner, in a process that takes about 200ms total. That's going to show up as a highly statistically significant advantage. Of course, the increased sense of connectedness to the game at higher rates is another big advantage, but these effects combine. Pit one skilled shooter against another at equal refresh rates, but with one dealing with 34ms of additional input latency, and they're not going to play as well as they should. Competitive players explicitly disable vsync, because the higher the game framerate, the faster it processes input. Also, the sooner you see updated information on the screen mid-refresh, but that doesn't help as much as a wholly higher refresh rate. All of this only matters if cloud gamers were playing against locally-rendered players in the same match though.
romwell 7y ago
>I have no idea how reliable it is This is not reliable at all. You can really feel 100ms latency. That's why in the world of pro music, barely anything above 10ms keypress-to-sound latency is deemed acceptable. Quake III will make all these streamed games feel slow.
- DontGiveTwoFlux 7y ago
  Well sound uses different neural pathways than audio, so it's not a totally fair comparison.
  - kibibu 7y ago
    Is this a typo? Anyway, audio feedback is a critical part of games. Pressing the "fire" button and waiting 100ms for the sound of the gun would be insanely noticeable
    - josephg 7y ago
      The inverse is also true. Sound triggers and local animations are often used to mask latency in games like DOTA which need a server round trip before your commands have any affect on the game world. It works remarkably well to make games feel responsive even when they’re not. However, with these streaming systems I’m not sure developers will be able to use tricks like that. It seems like these tech stacks are running all code remotely. I feel pretty skeptical about whether they can make the game feel good enough this way.
  - tlear 7y ago
    100ms is pretty bad in shooter. Around 60-70 is where I cant distinguish the difference. But I am very far from a great player. I can see this working very well for games like Total War where hardware specs are super high but response time is not critical
  - vincnetas 7y ago
    you mean video? "pathways than audio"
voldacar 7y ago
In arena shooters like quake and similar games, input lag of 10ms (1-2 frames) is noticeable. 200ms would be an absolute joke.

by Xixi 7y ago

ehsankia 7y ago

You can probably do even better than 20-30ms round trip if you're one of the large Cloud companies with data centers everywhere. Amazon so far doesn't seem interested, but Google and Microsoft definitely have the infrastructure to pull it off. Sure you won't get high end PC gamers, that's not really the target demographic here. It's the millions of console and mobile gamers.

TheOperator 7y ago

Actual video gamers right now go out of their way to get displays with very low input lags. 1080p144hz TN low input lag displays are used quite a bit in esports to shave a few msec. You literally win more often if you react faster. People aim to get TVs with sub 16ms input lag. However that's not most of the market. The appeal of cloud gaming is for single player non-competitive games. These are the most pirated games and the games where input lag matters the least.

borkt 7y ago

I can easily notice the difference between 20 and 40 ms in FPS games. I have gotten out of FPS's due to keeping up with the hardware requirements, but a streaming service like this could make it possible. That being said, if they cant stay below 50ms I probably wouldn't waste my time.

davemp 7y ago

I followed up on the wiki article. The citation was a just eurogamer article that had little to with human perception of latency. It did have this nugget: “Criterion said it was aiming for 50ms latency [for burnout paradise]” which agrees with what I’ve heard making handheld electronics. The goal was always <50ms latency. Personally I noticed a significant improvement switching to a 165hz monitor in basically every first person game. The only difference here is going from 16ms to 6ms frame granularity (10ms off the worst case). I’m not sure how much is visual smoothness or latency, but I’d gladly take refresh rate/fps over UHD. I’m definitely in the enthusiast group though. I didn’t even realize that console players dealt with such high latency. It will be very interesting to see how the market plays out.

gmueckl 7y ago
Hm... the topic is quite complex. Human perception is actually much slower than many people want to believe. Parts of the human vision system react no faster than around 1/25th of a second (other parts are faster - its complicated!). Hearing is similarly divided. AFAIK, auditory and visuak stimuli are processed together and merged over a rather large time window (to the point where one system can induce illusions in the other). But we are able to perceive time delays within a complex stimulus with very high precision. I am not totally convinced that a high refresh rate monitor allows human vision to respond faster. I would rather suspect that most game loops are VSync-locked and a higher refresh rate leads to better input sampling and processing. Maybe I should set up a simple test where I show a non-interactive sequence at various framerates and ask users to identify the highest framerate. And then repeat with interactive camera controls.
- hnuser123456 7y ago
  >I am not totally convinced that a high refresh rate monitor allows human vision to respond faster. Have you ever watched a high-level FPS player? They can aim and shoot in the span of about ~1/5 of a second, occasionally even faster. That's 200ms. At 25hz, that's 40ms. A big part of it is going to be increased input processing, but a lot of it is muscle memory at that skill level. At 165hz, that's 6ms. With two equally skilled players, one will see the other up to a whole 34 ms sooner, in a process that takes about 200ms total. That's going to show up as a highly statistically significant advantage. Of course, the increased sense of connectedness to the game at higher rates is another big advantage, but these effects combine. Pit one skilled shooter against another at equal refresh rates, but with one dealing with 34ms of additional input latency, and they're not going to play as well as they should. Competitive players explicitly disable vsync, because the higher the game framerate, the faster it processes input. Also, the sooner you see updated information on the screen mid-refresh, but that doesn't help as much as a wholly higher refresh rate. All of this only matters if cloud gamers were playing against locally-rendered players in the same match though.

romwell 7y ago

>I have no idea how reliable it is This is not reliable at all. You can really feel 100ms latency. That's why in the world of pro music, barely anything above 10ms keypress-to-sound latency is deemed acceptable. Quake III will make all these streamed games feel slow.

DontGiveTwoFlux 7y ago
Well sound uses different neural pathways than audio, so it's not a totally fair comparison.
- kibibu 7y ago
  Is this a typo? Anyway, audio feedback is a critical part of games. Pressing the "fire" button and waiting 100ms for the sound of the gun would be insanely noticeable
  - josephg 7y ago
    The inverse is also true. Sound triggers and local animations are often used to mask latency in games like DOTA which need a server round trip before your commands have any affect on the game world. It works remarkably well to make games feel responsive even when they’re not. However, with these streaming systems I’m not sure developers will be able to use tricks like that. It seems like these tech stacks are running all code remotely. I feel pretty skeptical about whether they can make the game feel good enough this way.
- tlear 7y ago
  100ms is pretty bad in shooter. Around 60-70 is where I cant distinguish the difference. But I am very far from a great player. I can see this working very well for games like Total War where hardware specs are super high but response time is not critical
- vincnetas 7y ago
  you mean video? "pathways than audio"

voldacar 7y ago

In arena shooters like quake and similar games, input lag of 10ms (1-2 frames) is noticeable. 200ms would be an absolute joke.