Elementary Explanation: Trying To Understand Pixel Refresh Rate (Was: "New Futuristic Temporal Resolution Hypothesis")

Advanced display talk, display hackers, advanced game programmers, scientists, display researchers, display manufacturers, vision researchers & Advanced Display Articles on Blur Busters. The masters on Blur Busters.
Post Reply
thatoneguy
Posts: 195
Joined: 06 Aug 2015, 17:16

Elementary Explanation: Trying To Understand Pixel Refresh Rate (Was: "New Futuristic Temporal Resolution Hypothesis")

Post by thatoneguy » 29 Feb 2024, 06:57

Disclaimer: This hypothesis is just a little fun thought experiment

An interesting concept came to my mind recently: What if the refresh rate matched the framerate? Would that eliminate all motion blur including Sample and Hold, blur from frame repetition etc. ?
Now I know what you’re thinking: ’’But bro that is Blur Busters 101 :roll:’’

No, this is a bit of a different concept
To explain it a bit more thoroughly let me explain it in a bit more detail:
  • Outer Motion Rate: This is self-explanatory as in the refresh rate denoting how many times the display refreshes per second
But what about Inner Motion Rate? Is it just the video frame rate?
NO!

Inner Motion Rate in this concept is divided into 2 categories:

Inner Motion Rate:
  • Line-level rate: In this concept it denotes the horizontal lines of the resolution of the video multiplied by the framerate of the video(e.g 3840 times 24fps)
  • Pixel-level rate: In this concept it denotes every pixel of the video multiplied by the framerate of the video(e.g 3840x2160 times 24fps)


A demonstration:

Line-level rate calculation: 3840(lines) * 24(fps) = Outer Motion Rate is 92160Hz
Pixel-level rate calculation: 3840x2160(pixels) * 24(fps) = Outer Motion Rate is 199065600hz or about 199.0656Mhz

Why would such an absurdly high refresh rate be neccessary?
You see, in each frame of a 4K video you have you have about 3840 lines horizontally an 2160 vertically, which means that in a 4K video in each frame you have over 8 million pixels.

So ultimately my hypothesis is that if the Refresh Rate matches (Pixel/Line-rate * Video Framerate) then motion blur could be completely eliminated.

Could a display ever be fast enough to achieve an absurd refresh rate like 199.09656Mhz?
I actually did some calculations and it turns out that you it’s about 5 nanoseconds so technically a MicroLED display has a response time that is fast enough to do it.

Will we ever have the neccessary bandwidth to ever try it?
Now this is a toughie, and I can’t say for sure. Even with compression techniques like Display Stream Compression it will take a long time before we can ever do it.

This hypothesis could of course be tested out with much lower resolutons and frame rates( for example 320x240 * 10fps= Pixel Rate of 768000hz or 160x120 * 10fps = Half of that).
Line level rates could be tested in the lab with low resolutions/framerates right now( e.g 320 lines x 10fps video = Only requires 3200hz), that is assuming they are using Emissive LED screens with 0 GtG. This is based on Chief saying in reddit that they already have 20Khz refresh rate displays on the lab.
But Pixel level refresh rates lke 768000hz seem to be beyond the scope of what we can do right now.

And that is all !
I may be wrong, I may be right. But it’s fun to hypothesise.

User avatar
Chief Blur Buster
Site Admin
Posts: 11726
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Elementary Explanation: Trying To Understand Pixel Refresh Rate (Was: "New Futuristic Temporal Resolution Hypothesis")

Post by Chief Blur Buster » 07 Mar 2024, 20:55

No, the display physics does not work like that.
That's not a temporal resolution metric, that's like an Internet connection speed.

What you are describing is bandwidth/speed (like the megabits per second on Internet). It has no effect on motion blur.

We already have pixel clocks built into signals;
pixelclock.png
pixelclock.png (32.94 KiB) Viewed 1208 times
Pixel clocks (a number already built into ToastyX and NVIDIA Custom Resoultion) has no effect on motion clarity, refresh rates, and frame rates.

It's simply a function of how fast a GPU can generate pixels, and transmit pixel from point A (GPU) to point B (display)

Like a polygon rate, a fillrate, etc. It's of no benefit to frame rates, just that a higher pixel rate means you can have a higher resolution at a higher refresh rate concurrently.

It's just a speed, like the speed of an Internet connection.

The faster the pixel clock, the more refresh rate you can get per pixel.

Motion clarity is dictated by pixel visibility time, and that has to apply to ALL pixels. If you refreshed one pixel 1 billion Hz, you don't have bandwidth left over to refresh the OTHER pixels 1 billion times anymore.

As a serialization of 2D picture data over 1D link between signal source and display - it generally can only refresh 1 pixel at a time, left-to-right, top-to-bottom like a calendar, see high speed video at www.blurbusters.com/scanout

Think of this; you're playing a compromises game like trying to split up a million dollars amongst a million people. Do you give $999K to one person (metaphorically massive refresh rate for one pixel), and less than a penny to everyone else (metaphorically less than 0.001Hz for other pixels)? Or do you give everyone a dollar equally (same refresh rate for all pixels of a screen)?

You can only gain motion clarity if ALL pixels refreshed EQUALLY as fast. But you would need unobtainium pixel clock in order to get enough per-pixel refresh rate for ALL pixels concurrently.

Think of a DSL/Cable/fiber Internet connection. Each pixel requires 24 bits to transmit over the cable. Without overheads (synchronization intervals/porches), 1920 pixels times 1080 pixels times 24 bits times 60 Hz:
calculator.png
calculator.png (4.85 KiB) Viewed 1193 times
This is because one picture is 1920x1080, and each pixel in the picture consumes 24 bits of data.
And you have to do it 60 times a second. And that's the amount of data that has to go over a cable, just to get 60 refresh cycles per second.

Now imagine, repeat sending 60 pictures per second over an Internet connection called "HDMI Cable" or "DisplayPort Cable".

Nearly three billion bits per second! (Actually it's over 3 billion with some of the overheads, like the synchronization intervals).
That's much faster than most fiber Internet connections, since it's completely uncompressed.

That means your video signal speed is about 3 gigabits per second, much faster than your Internet connection. It's metaphorically displaying 60 photographs images per second, one photograph after the other, rapidly, like a flipbook! That's what displays do, they rapidly display a series of stationary images. It takes a LOT of bandwidth.

Remember, the cable is transmitting pixels one at a time, left-to-right, top-to-bottom, like the letters of a book.

Think of this; a screen is metaphorically a page, and pages/second is the display refresh rate you're limited to --

You have to finish a WHOLE refresh cycle (all pixels on a screen) before you can declare it a true unit count refresh rate.

Now ask yourself; what's the minimum possible number of bits to do ONE refresh cycle at 1920x1080? That's a lot of bits and bytes (because each pixel is 24 bit, bit one refresh cycle at 1920x1080x24 = almost 50 million bits. Now imagine trying to do that 1000 times a second (1000Hz). That becomes 50 billion bits! Imagine a 50 gigabit Internet connection between your computer and display -- that's what very fast DisplayPort cables are -- very fast short-distance data links.

Which means your refresh rate & frame rate is limited, because you don't have infinite bandwidth between a computer and a screen (even inside an embedded screen too, like tablet/laptop/phone). Obviously your screen performance also matters.

But number of pixels/sec is a bandwidth speed metric, like 1920x1080 pixels times 60 refresh cycles per second (times overheads) to create the pixel rate.

So to reduce motion blur, you need to shorten pixel visibility time.
- More refresh rate for all pixels (which is hard, requires lots more bandwidth if you have more pixels)
- Or add flicker which some people hate (like strobe backlight, phosphor, etc)


Motion blur at the fastest pixel response time (GtG=0) is throttled by:
- Pulsetime on strobed
- Frametime on flickerless


Also, here is an educational animation about motion blur = pixel visibility time
If you have a 240Hz display, turn off strobing temporarily and observe TestUFO 240Hz Variable-Persistence Black Frame Insertion. Observe with that educational animation:
(A) One-quarter frame rate is four times the motion blur;
(B) Observe that motion blur corresponds to pixel visibility time for all pixels (frame visibility time).
(C) Observe that minimum motionblur is throttled by minimum refreshtime if you're not using hardware-based strobing (which can flash briefer than a refresh cycle).
But doing either, STILL requires bandwidth to display all pixels of a while refresh cycle.

Unfortunately, it's throttled by framerate too:
"Refresh rate" is like the outer refresh rate you were talking about.
"Frame rate" is like the inner refresh rate you were talking about.
So, if you're getting only 60fps at 240Hz, it looks like 60fps 60Hz (sample and hold), like this other animation at www.testufo.com

There are no shortcuts, just like the "one million dollars split across one million people" metaphor. You can't give everybody a million dollars if you need to split a million dollars among a million people. That would require a trillion dollars (one million times one million). A million divided by a million equals only $1 per person if you're trying to split a million dollars amongst a million people.

TL;DR: You have to refresh ALL the pixels to get extra motion resolution, and that takes bandwidth per pixel. More pixels = more bandwidth needed.

Remember:
"Refresh rate" = "Picture rate" = "Number of pictures per second a screen can display"
Remember, PIXEL = a pormanteau of PICture ELement, so PIXEL is like one element of a picture.
And you have to refresh the WHOLE screen, which is a lot of data. Which means you can't refresh too many times a second.

This is really just mere rudimentary elementary school/junior high school math stuff.
For more advanced reading, see www.blurbusters.com/area51 about display motion blur physics.
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

Image
Forum Rules wrote:  1. Rule #1: Be Nice. This is published forum rule #1. Even To Newbies & People You Disagree With!
  2. Please report rule violations If you see a post that violates forum rules, then report the post.
  3. ALWAYS respect indie testers here. See how indies are bootstrapping Blur Busters research!

thatoneguy
Posts: 195
Joined: 06 Aug 2015, 17:16

Re: Elementary Explanation: Trying To Understand Pixel Refresh Rate (Was: "New Futuristic Temporal Resolution Hypothesis

Post by thatoneguy » 19 May 2024, 21:48

Chief Blur Buster wrote:
07 Mar 2024, 20:55
*snip*
I had forgotten I made this thread.
Mark, I am aware of all that you wrote. What I was suggesting is highly(extremely) hypothetical(hence the original title of this thread).

I may not have explained my idea very well. I'll try to shorten it.
What I was suggesting was basically ''Hertz per Pixel''.
How traditional displays work: 1 frame refreshed in 1 hertz.
My (hypothetical)idea: 1 pixel refreshed in 1 hertz.

This method, as you stated, basically requires for the refresh rate to match(or perhaps double or on the extreme side of things...squared(for 4K@24fps that would be ~199 million times ~199 million)...OUCH) the pixel clock/bandwidth of the signal itself.

And now that I think about it, it sounds a bit like how the Nyquist-Shannon sampling theorem is applied to audio(e.g 44.1khz sampling rate for 22.05khz frequency response).

The refresh rate required is obviously as you put it ''unobtanium'', hence why I put ''futuristic'' in the title. Because it really is a futuristic idea.

I'm not saying I'm right either, hence why I called it a hypothesis.
I'm obviously no expert like you, but I do believe there is some logic behind my thought process and I wasn't just throwing shit at a wall to see if it sticks.

Not gonna lie, I don't like that you changed the title of my thread but hey it's your forum. Oh well!
Feel free to lock the thread if you like.

Post Reply