davis.anthony
Veteran
Because Sony never promised 45% faster, they said up to 45% faster, there's a big difference between the two comments.
Digital Foundry Article Technical Discussion [2024]
- Thread starter Shifty Geezer
- Start date
- Status
Because Sony never promised 45% faster, they said up to 45% faster, there's a big difference between the two comments.
1ms mprt is equivalent to 1000 Hz sample and hold with no strobing or bfi. It’s basically 1ms where the image is visible and the rest of the refresh period you have a black screen. That’s basically as minimal blur as you can get.
What I’m saying is in that scenario the higher frame rate will give you smoothness that a lower frame rate will not. Neither will have motion blur.
In terms of display tech, I’d like to see advancement in frame generation. It’s a technology that exists and works reasonably well.
A new hypothetical display tech that doesn’t use sample and hold, and doesn’t exist, is no reason to stop advances in existing technology.
Moving more and more towards gpu-driven rendering (maybe work graphs) to remove the cpu as much as possible is an important step.
High frame rates are one of those things you have to experience to understand.
What I’m saying is in that scenario the higher frame rate will give you smoothness that a lower frame rate will not. Neither will have motion blur.
In terms of display tech, I’d like to see advancement in frame generation. It’s a technology that exists and works reasonably well.
A new hypothetical display tech that doesn’t use sample and hold, and doesn’t exist, is no reason to stop advances in existing technology.
Moving more and more towards gpu-driven rendering (maybe work graphs) to remove the cpu as much as possible is an important step.
High frame rates are one of those things you have to experience to understand.
davis.anthony
Veteran
CRT laughs at this comment.
Yes, lets destroy image quality even more than we are already.
It is when it's a dead end, you have seen reviews of an RTX 4090 yes?
Chasing high frame rates to match high refresh rates will stop advances in rendering, those high frame rates don't and won't come cheap in terms of GPU frame time.
I have, on an OLED PC monitor.
Last edited:
But this isnt 2014 anymore. HDR (and VRR) and strobing will be a problem. OLEDs cant even do 300 Nits fullscreen and strobing will massivly reduce brightness.
If we're just going to invent a new display tech that can solve these issues, why not just invent a new computing system that can render 2000 fps? ¯\_(ツ)_/¯
With all the improvements in technology they can't do a 4K CRT that would be much lighter than before?
No. CRTs are dependent on a vacuum tube for the electrons to fly through. You need a minimum amount of glass to hold that, which ensures they are huge and heavy. Best case, you could have some advanced magnetics to focus the beam over a smaller distance, but you'll still have a minimum that's still huge and heavy. And no-one is going to want to give up that flat screen display mostly used for TV+movies to accommodate gaming needs. You could try and convince the monitor companies there's a market for 1440p etc desktop CRTs. Price would probably be phenomenal given a niche audience and relatively high production costs, making it even more niche.
Well for starters the former is vastly more power efficient. Our eyes don’t need 2000 displayed fps in order to perceive smoothness and rendering 2000 fps even with frame gen is a huge waste of computing and electrical power.
Yes that’s the trick, need display tech that’s bright enough to light pixels for a very short time then go dark. I’m simply suggesting research money goes there instead of chasing nonsensical physical refresh rates.
Okay, how about inventing a technology that just beams the images into our minds without even needing a display.
Not really sure where you’re going with this. Is your point that in a world with finite resources we shouldn’t have opinions on where those resources are spent?
“Brighter pixels” isn’t some magical unreachable goal. “Beaming images” is sci-fi nonsense.
Charlietus
Regular
I don't understand how clocking the GPU 45% faster wasn't the better and more economically feasible choice than putting 67% more CU's.
I mean, I'm sure they crunched the numbers, but it still doesn't make sense to me.
GhostofWar
Regular
John could try lossless scaling if he wants to try that 480hz monitor with higher framegen options, maybe magpie if he doesn't want to spend money although i'm not sure how that compares to lossless.
I was on the internet the other day, and there was this video where this guy had to make a choice between two.. uh, well there were two choices. Anyway, it ends up that both of those choices at the same time ended up being the right choice. It was a very informative video and I learned a lot from it. Anyway, I think we should have higher refresh rates and new display technology at the same time.
@trinibwoy The companies that make display technology are lg, Samsung and au optronic. They have no roadmap for a new super display with no downsides. If they could make one in the next five, or even ten, years then where is the information? They’re still trying to make micro leds.
AMD and nvidia in the meantime can continue to improve their existing display technology.
But back to my hypothetical. A 1ms MPRT display should be as good or better than a CRT. If you had two CRTs and one was 120 Hz and the other was 480 Hz, which would produce the smoother image while panning? Do you really think your eyes would interpolate and they’d look the same?
AMD and nvidia in the meantime can continue to improve their existing display technology.
But back to my hypothetical. A 1ms MPRT display should be as good or better than a CRT. If you had two CRTs and one was 120 Hz and the other was 480 Hz, which would produce the smoother image while panning? Do you really think your eyes would interpolate and they’d look the same?
It should but the question is how do we get to 1000 fps. Rendering isn’t getting any easier and hardware advances are slowing down.
I'm not blaming Nvidia and AMD. However the more bandages they apply the less incentive Samsung and LG have to improve their tech. Nvidia is happy to sell you "fake frames" all day long.
Already addressed above. Higher refresh is obviously better all else equal but that's not what I'm talking about.
Last edited:
The complaints against current tech are comparing current tech to a non-existent alternative, with a completely unrealistic expectation that someone should just come up with an alternative that's better. If your going to want some better tech that doesn't exist and isn't on any roadmap, why limit yourself to a box on the wall and not just go with something completely new? My sardonic argument against the 'we need new tech' argument.
You don't need, and wouldn't want, 1000 fps game rendering for what Dictator describes. You render 30 fps or whatever, but draw a part of the screen at a time to emulate a CRT scanning beam, and make multiple passes across one game frame. A 120 Hz CRT would draw a 30 fps game 4 times per game update. This is a display solution.
Disagree. There's active work to improve pixel brightness. What do you mean it's not on any roadmap?
How is that a solution to image persistence blur? Rendering the same frame multiple times will produce the same blur.
I wasn't talking about brightness. I was replying to davis.anthony's position...
What is this new display tech that doesn't use sample-and-hold?
I don't honestly know. People talk about CRT scanning as a panacea to all displays ills, but I remember juddery frames on that too, along with flicker and blurring and stepping and sawtooth aliasing.
I'm just explaining that Dictator's position about CRTs being emulatable on very high refresh OLED/LCD doesn't need 1000 fps games. Whether that's actually a solution t the issue of smooth motion at lower framerates, I don't know.- Status
Similar threads
