Hacker News Reader

Hacker News

Comment by throwaway48476 | Hacker News Reader

Parent Full thread

throwaway48476

Fake frames, fake gains

View on HN

Are DLSS frames any more fake than the computed P or B frames?

Yes.

how so?

P and B frames are compressed versions of a reference image. Frames resulting from DLSS frame generation are predictions of what a reference image might look like even though one does not actually exist.

Compression and prediction are the same. Decompressing a lossy format is guessing how the original image might have looked like. The difference between fake frames and P and B frames is that the difference between prediction of fake frame and real frame is dependant on the user input.

... now I wonder ... Do DLSS models take mouse movements and keypresses into account?

But MPEG is lossy compression which means they are kind of a just a guess. That is why MPEG uses motion vectors.

"MPEG uses motion vectors to efficiently compress video data by identifying and describing the movement of objects between frames, allowing the encoder to predict pixel values in the current frame based on information from previous frames, significantly reducing the amount of data needed to represent the video sequence"

There's a real difference between a lossy approximation as done by video compression, and the "just a guess" done by DLSS frame generation. Video encoders have the real frame to use as a target; when trying to minimize the artifacts introduced by compressing with reference to other frames and using motion vectors, the encoder is capable of assessing its own accuracy. DLSS fundamentally has less information when generating new frames, and that's why it introduces much worse motion artifacts.

it would be VERY interesting to have actual quantitative data on how many possible I video frames map to a specific P or B frame vs how many possible raster frames map to a given predicted DLSS frame. The lower this ration the more "accurate" the prediction is.

The fps gains are directly because of the AI compute cores, I’d say that’s a net gain but not a the traditional sense preAI.

Kind of a half gain: smoothness improved, latency same or slightly worse.

By the way, I thought these AI things served to increase resolution, not frame rate. Why doesn't it work that way?

It's both. And it's going to continue until games are 100% and AI fever dream.

Right, I should have said "generate extra pixels in every frame, not interpolate whole frames". Doing the former also increases frame rate by reducing computation per pixel.

The human eye can't see more than 60 fps anyway

Can definitely see more than 60, but it varies how much more you can see. For me it seems like diminishing returns beyond 144Hz.

Though some CRT emulation techniques require more than that to scale realistic 'flickering' effects.

You are right, but diminishing returns technically start around 60.

i can tell up to about 144Hz but struggle to really notice going from 144 to 240Hz. Even if you don't consciously notice the higher refresh rate it could still help for really fast paced games like competitive FPS if you can actually generate that many frames per second by reducing input latency and if you can actually respond fast enough.

Same with me. At least on LCD. I'm still gonna get 480hz OLED display because I'm curious.

The human eyes are analog low pass filters, so beyond 60Hz is when things start to blur together, which is still desirable since that's what we see in real life. But there is a cutoff where even the blurring itself can no longer help increase fidelity. Also keep in mind that this benefit helps visuals even when the frame rate is beyond human response time.

This is factually incorrect and I don't know where people get this idea from.

Just moving my mouse around, I can tell the difference between 60 and 144 fps when I move my pointer from my main monitor (144 hz) to my second monitor (60 hz).

Watching text scroll is noticeably smoother and with less eye tracking motion blur at 144 hz versus 60.

An object moving across my screen at 144 fps will travel fewer pixels per frame than 60 fps. This gain in motion fluidity is noticeable.

I remember when it was "the human eye can't see more than cinematic 24fps" sour grapes by people who couldn't get 60fps

This is laughably false and easy to disprove. Blurbusters did an analysis of this many years ago and we won't get "retina" refresh rates until we're at 1000Hz.

Reply on news.ycombinator.com