the framerate line: get on a side

still every so often think about having played hyper light drifter pre 60fps patch and how what i really played was myself

8 Likes

also i don’t believe rudie

1 Like

I made a game that ran at 24fps once.

https://postsecret.com/

5 Likes

People who dislike playing video games competitively mocking folks who care about things like framerate and input lag and resolution really crack me up.

Like, if I’m in a noise band it’s probably cool if my guitar doesn’t stay in tune but in any other context it sure makes things more unpleasant and difficult.

12 Likes

well, yes, this hobby has a very complicated relationship to obsessives

I think that’s unavoidable

1 Like

not being able to hear dialogue in tv and film because it’s assumed everyone has some ridiculous 7.1 setup fucks me right off

9 Likes

I’m looking for a metaphor to get this across - jump out of skin, nails on chalkboard - but out of tune instrument also hits the idea of input. If the machine is not providing consistent responses to my inputs it is viscerally upsetting.

It’s not about overall fidelity - Nintendo stuff on the Wii U and Switch is great with the exception of BotW.

Holy shit. When I have subtitles on and a subtitle appears and there’s no audible voice coming from my 2-channel stereo TV I get fucking livid.

2 Likes

That should be an incorrect setup problem; everything should downmix to whatever your actual speaker setup is, including if it’s mono.

I’m happy with how dynamic audio mix controls have spread in the last decade – they’re usually labeled as ‘TV’, ‘Headphone’, ‘Speaker System’, or the sensual, ‘Midnight Mode’ in some. Takes advantage of the audio processing a computer can do (on recorded stuff, you have to lean on your TV or player to do ‘dynamic mixing’, which will offer these days, if you’d like to compress the mix).

1 Like

I think that’s a reasonable metaphor for the difference between the rabid FPS hounds and the standards of mastery and technical work that goes into presentation; there’s a body of knowledge that exists and is needed to produce this to a more precise level, and then there’s the half-true information that filters into people with more energy and anger than sense.

And the way people can teach themselves aesthetic values based on a misunderstanding

1 Like

Preferred arc of format that leans into aesthetic or varying taste, obviously we have subjective degrees of fuzz and clarity. Emulation enhancements or fine tuning of shaders are like an entire meta projection of this.

My eyes just roll backwards out from their sockets into a bowling bag when you see someone imply it’s all made up just for a sense of prestige bullshit. Elitist pressuring is bad, but if you’re in any scenario sifting through the qualities of x, everyone has their base standard of measuring and definition involved, ideally flexible.

and yeah high framerate has been a core foundation of a lot of The Best Shit like q3a, sega arcade games (excellent example thanks @doolittle) etc.

but at the same time i mean, i fuckin’ love demon’s souls and it’s a really boring thing to build a personality around

1 Like

60fps Youtube videos freak me the hell out like they’re reverse stop-motion animation

Thank you and goodnight

5 Likes

I think its born from two things:

  1. We are in an era of HD and generally pretty potent hardware, even for consoles. Compared to previous generations, current consoles are way more powerful than ever. Such that, we have the luxury of caring about framerate. Back in the PSX and N64 days, we didn’t really have the resources for games to always be “good” on framerate. Its just the way it was and we were used to it. In example, I was recently surprised to find out that Goldeneye 64 spends a ton of time under 20fps. It wasn’t until the days of 360 ports of games Vs. PS3 ports of games, that framerates really started to become mainstream. Because for the first time ever, there were "better’ versions of games available. With Goldeneye, there was never a faster version available. So no one ever talked about it.

So that’s what I think, for point 1.

  1. On the tech savvy side or actual technical editorials: A few years ago there was sort of a renaissance and marked change in how framerate is generally talked about. You see, AMD had a real big problem at the driver level. Its was a fundamental problem with their driver stack, which affected virtually their entire suite of graphics cards. (its a problem which they fixed or at least improved a hell of a lot, BTW).

Even though framerate monitors such as FRAPS or MSI Afterburner were reporting “good” framerates: games still felt bad. They didn’t always look or feel smooth. Sometimes it was so bad and so apparent, it would be known as hitching or micro-stutters. Which was an actual thing. Not just made up by some over analyzing nerd. This is pointing to a much different problem than this scene is too demanding, so the framerate is lower.

Certain benchmarks or other programs which visually plotted framerate in a visible graph, would show these frequent dips. Even though the average framerate reported was “good”.

So what was going on? It turns out, average framerate isn’t the only thing we should be concerned with. And these visual plots are an indication of this.

Frame times (also referred to as frame pacing). How many milliseconds does it take for each frame to be completed and sent to your framebuffer? AMD’s problem got the entire industry educated on frametimes. Now its often talked about generally. And some review sites actually have a standard way of testing it for their reviews. Turns out somebody at Nvidia had developed a special tool for it called FCAT. But even FRAPS will spit out frametimes in a report. I think FCAT is regarded as the most accurate way, but FRAPS is still pretty good. But nobody really “knew” about it.

If we evenly split 60 frames per second, you have about 16.7 milliseconds per frame. This also relates to input lag. Your inputs fit into the frametimes.

In reality, no average framerate is gonna be a perfect split of 16.7 milliseconds. Because the demand of what needs to be rendered is constantly changing. Some frames are just really easy to deliver. So some reviewers have come up with the “99 percentile” where they show the average time of delivery for 99% of the frames. The smaller this number is, the better.

If you look at this page for a specific test in Far Cry 5

Both the non ti 2080 and the Radeon VII deliver an average of about 60fps in this Far Cry 5 test. However, if we look at the 99th percentile, a much different picture emerges:

The 2080 has 99% of its frames coming in at about 21ms. If all frames were 21ms, you’d be looking at about 48 frames per second. So in this case, the remaining 1% are quick enough that the average framerate is still 60fps.

For the Radeon VII, 99% of the frames are coming in at about 32.8ms. If all frames came in at that speed, we would be looking at 30 frames per second. Again, the remaining 1% are fast enough that the average is 59 frames per second. In this case, the Radeon’s 1% has to be relatively a lot faster, to equal the same average framerate.

So…what? Lets talk about input lag again. For 99% of a second the quickest the Radeon VII can update, is 32.8ms. Whereas the 2080 is much faster, at 21ms. The 2080 is going to feel faster. In this particular case, for 99% of a second, the Radeon VII feels like a 30fps game. Even though a framerate monitor says 59fps. That’s really bad. Comparatively, the 2080 feels like 48fps, for 99% of a second. That’s much better.

This data also means that the 1% frames for the 2080 aren’t going to feel as different, relative to the other 99%. Whereas with the Radeon, the 1% are going to feel relatively much quicker. So the 2080 is also going to feel smoother or more consistent.

**Keep in mind, this is one example. I’m not in this moment trying to say that Nvidia is better than AMD. However, a person may come to a preference for a brand, by looking at data across several games at several resolutions, etc.

This knowledge of frametimes has crept into general public knowledge. So we occasionally see talk about it. A good example would be the Souls games by From Software. Their console ports are notorious for frametime/frame pacing issues. With noticeable stuttering and hitching. Even though an average framerate report would say “locked at 30fps”. People would feel it anyway. But now they have commons terms to talk about it, which actually refer to real technical stuff. Not just “man this game really stutters”.

If you wanna deep dive on frame times, read more here:

5 Likes

Hello everyone I love stable frame rates but I am too bored by the idea of arguing about it with people so you’ll just have to put up with my opinion ok bye

9 Likes

HDR on projectors is tough because there’s no real local light control so it’s just tone-mapping the image with metadata. Hence, darker, probably washed out colors too.

1 Like

How do I setup the sound on my old plasma’s built in speakers so music and effects aren’t three times as loud as dialogue

3 Likes

yeah to make HDR movies look good without being washed out or hard to look at I have to both turn the setting up from “standard” to “bright” and turn off the eco mode/dynamic black or else the bright setting crushes them.

it works, but compared to my newfound oceans of film grain it’s not as immediately effortless. “you need to turn off the eco mode” is probably a fair situational thing tbh it just grates

Best bet is probably a “night mode” or other lower dynamic range setting on the input device, not the TV if it’s old.

2 Likes