4 ms·

This comes up in an ancient and often heated discussion a lot on game boards as to the "optimal" FPS to have in a game - many proponents saying anything under 1

This comes up in an ancient and often heated discussion a lot on game boards as to the "optimal" FPS to have in a game - many proponents saying anything under 100 FPS is "unplayable" for them, and another group saying basically anything over 60 is either not noticeably perceivable to the human eye, or barely noticeable. I'm starting to see the pop-science-gamer-journalism interpretation of this study as being "study proves some humans see at different FPS" and all the 100+ FPS gamers being like "AHA! Told you!" when really the results of the study support the ~60 fps conclusion that's been mainstream for a while (if I'm interpreting correctly). I think the science here is pretty clear - there is a variation in human visual temporal resolution, and this paper's a bit over my head but if I'm understanding correctly it isn't much (at least in this sample size and the gamer context - I'm aware they concluded the variation was large between individuals). TLDR if you're fussing that you're "only" getting 90 FPS vs 120 or whatever, and genuinely feel like it's affecting your performance - don't worry, seems very similar to the audiophile stuff to me. I think that there are probably big outliers in this range (thinking specifically professional baseball players), but perhaps the "perceived" difference is because if you have 130 FPS your average framerate is likely to always stay above your perception, and maybe at 70-90 it will occasionally dip below that threshold, causing you to perceive a difference.

Terr_ 3y ago
To complexify things a bit more, there's also a difference in testing/implications between: 1. Someone can detect a difference between rate X and Y. 2. Someone can detect that X is lower than Y. 3. Someone receives information better when increasing from X to Y. It's self-evident that there are values of X and Y where those are all true, but it's not certain that they will become false at the same conditions.
causi 3y ago
Gamers have testably higher in-game performance at higher framerates. https://www.youtube.com/watch?v=OX31kZbAXsA https://www.youtube.com/watch?v=OX31kZbAXsA
- kanbankaren 3y ago
  The US military has done research on refresh rates to find out the optimum refresh rate for pilots and their conclusion was the same. Beyond 60fps, there is not a benefit. There was another research which discussed about a latency in the brain of around 15ms. Even if the visual system could detect a change, not much could be done with that change due to the inherent latency.
  - 082349872349872 3y ago
    Note that hearing and touch have higher "frame" rates.
  - thfuran 3y ago
    >There was another research which discussed about a latency in the brain of around 15ms. Even if the visual system could detect a change, not much could be done with that change due to the inherent latency. That claim does not support that conclusion. 15 ms is roughly the frame time at 60 fps, meaning framerate potentially accounts for a significant and reducible portion of overall latency.
  - causi 3y ago
    A 45,000-pound airplane has a slower response time than a cursor on a videogame.
    - kanbankaren 3y ago
      The research was in relation to HUD for fighter aircraft from what I recall. Nothing to do with how the aircraft responds rather identifying targets on screen.
drjasonharrison 3y ago
It's a bit difficult to draw reliable conclusions from this critical flicker fusion paper to the entire visual field and to the detection of changes in position or appearance. My take away from the paper is that there is variation in the CFF across individuals for foveated visual targets that are much brighter than the surround. It remains to be seen the effects of visual targets that are dimmer, lower contrast, colour, and peripherally placed.
bee_rider 3y ago
I’m quite happy to have grown up in the n64 era, totally happy with 30fps. And you can get some sweet effects and high resolution on a cheap card at 30fps. OTOH, some games seem to add, like, a frame or two between input and response? Whatever it is, it feels really annoying. But I’ve played plenty of games that feel perfectly tight and responsive at 30fps so I think there must be something else wrong. I’m sure it could be fixed with a higher frame rate, but I’d rather not resort to that.
tithe 3y ago
This is fascinating. I often wonder about the different "clock speeds" of various functions in the brain and the mechanisms in play to keep these functions in sync. Tangentially, I wonder if the brain's "FPS" (sample rate?) in hearing and processing audio differs between individuals in a similar way.

Terr_ 3y ago

To complexify things a bit more, there's also a difference in testing/implications between: 1. Someone can detect a difference between rate X and Y. 2. Someone can detect that X is lower than Y. 3. Someone receives information better when increasing from X to Y. It's self-evident that there are values of X and Y where those are all true, but it's not certain that they will become false at the same conditions.

causi 3y ago

Gamers have testably higher in-game performance at higher framerates. https://www.youtube.com/watch?v=OX31kZbAXsA https://www.youtube.com/watch?v=OX31kZbAXsA

kanbankaren 3y ago
The US military has done research on refresh rates to find out the optimum refresh rate for pilots and their conclusion was the same. Beyond 60fps, there is not a benefit. There was another research which discussed about a latency in the brain of around 15ms. Even if the visual system could detect a change, not much could be done with that change due to the inherent latency.
- 082349872349872 3y ago
  Note that hearing and touch have higher "frame" rates.
- thfuran 3y ago
  >There was another research which discussed about a latency in the brain of around 15ms. Even if the visual system could detect a change, not much could be done with that change due to the inherent latency. That claim does not support that conclusion. 15 ms is roughly the frame time at 60 fps, meaning framerate potentially accounts for a significant and reducible portion of overall latency.
- causi 3y ago
  A 45,000-pound airplane has a slower response time than a cursor on a videogame.
  - kanbankaren 3y ago
    The research was in relation to HUD for fighter aircraft from what I recall. Nothing to do with how the aircraft responds rather identifying targets on screen.

drjasonharrison 3y ago

It's a bit difficult to draw reliable conclusions from this critical flicker fusion paper to the entire visual field and to the detection of changes in position or appearance. My take away from the paper is that there is variation in the CFF across individuals for foveated visual targets that are much brighter than the surround. It remains to be seen the effects of visual targets that are dimmer, lower contrast, colour, and peripherally placed.

bee_rider 3y ago

I’m quite happy to have grown up in the n64 era, totally happy with 30fps. And you can get some sweet effects and high resolution on a cheap card at 30fps. OTOH, some games seem to add, like, a frame or two between input and response? Whatever it is, it feels really annoying. But I’ve played plenty of games that feel perfectly tight and responsive at 30fps so I think there must be something else wrong. I’m sure it could be fixed with a higher frame rate, but I’d rather not resort to that.

tithe 3y ago

This is fascinating. I often wonder about the different "clock speeds" of various functions in the brain and the mechanisms in play to keep these functions in sync. Tangentially, I wonder if the brain's "FPS" (sample rate?) in hearing and processing audio differs between individuals in a similar way.