Blurinator 9000 - Blur simulation software

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.
Haste
Posts: 326
Joined: 22 Dec 2013, 09:03

Re: Blurinator 9000 - Blur simulation software

Post by Haste » 12 Jul 2021, 17:12

Would it be possible to have an option to simulate stroboscopic stepping?
Monitor: Gigabyte M27Q X

MCLV
Posts: 43
Joined: 04 Mar 2021, 15:04

Re: Blurinator 9000 - Blur simulation software

Post by MCLV » 13 Jul 2021, 16:18

Discorz wrote:
12 Jul 2021, 07:26
MCLV wrote:
12 Jul 2021, 06:38
Guess I can't improve it further then.

Do you have any future plans with this software, updates? If yes, can we expect improvements like wider/slower response times range and a possibility of mixing ghosting with inverse ghosting (overshoot)...?
Not at the moment. Frankly, even doing this GUI took more effort that I originally wanted to invest. And it also confirmed my past experience that once a GUI is made, one usually gets in a "Give them a finger and they’ll take the whole hand" situation with its users :lol:

Slowest response time corresponds to GtG with roughly the same value as frame time (i.e. 16.7 ms for 60 Hz) which I thought to be sufficient. Furthermore, Blurinator generates blur trail based on actual and two additional frames. Having a slower GtG would result in a need to extend the number of frames which are included in blur calculation. Otherwise, the blur trail would be truncated. This would also generate larger sideways shift of the blur image. So I decided to keep it simple and just not allow slower response times.

There is a lot of things that can be done but many of them are much simpler to do by changing the code directly rather than by creating a GUI which gives this flexibility to the user. So I encourage you to look at the code and play with it.

User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: Blurinator 9000 - Blur simulation software

Post by Discorz » 14 Jul 2021, 06:11

MCLV wrote:
13 Jul 2021, 16:18
That is true tho.

Unfortunately I never dived into coding world. It looks like hieroglyphics science to me.

Now I only hope Chief finally does something similar as he always wanted. This is a solid base to add to :)
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Blurinator 9000 - Blur simulation software

Post by Chief Blur Buster » 10 Dec 2021, 18:00

Haste wrote:
12 Jul 2021, 17:12
Would it be possible to have an option to simulate stroboscopic stepping?
MCLV wrote:
13 Jul 2021, 16:18
There is a lot of things that can be done but many of them are much simpler to do by changing the code directly rather than by creating a GUI which gives this flexibility to the user. So I encourage you to look at the code and play with it.
Discorz wrote:
14 Jul 2021, 06:11
Unfortunately I never dived into coding world. It looks like hieroglyphics science to me.

Now I only hope Chief finally does something similar as he always wanted. This is a solid base to add to :)
I'd like to add other stuff like:

- A PWM emulator (to simulate PWM dimming)
- A phosphor decay emulator (to simulate KSF/CRT)
- A temporal dithering emulator (to simulate plasma/DLP)
- A temporal color emulator (to simulate DLP color wheel)

These are actually very simple shader math functions, so rewriting/porting Blurinator 9000 to use shader math may actually allow it to emulate more displays.

In theory, it could also do the simulation in real time (generate 240 blurinatored frames at 240 frames per second), and become the algorithm for a CRT electron-beam emulator (e.g. RetroArch BFIv3)

In fact, I may be happy to create a 4-figure or maybe 5-figure (USD) bounty for this, or hire someone part-time to create this codebase (stipulation: commercialzation-friendly open source licenses such as Apache or MIT), to create something that can later be added to RTSS or Special K or a windows indirect display driver. (Emulate 60Hz Windows CRT rolling-scan on a 360Hz monitor!). Apache or MIT, take your pick.

Since I have long tended to be the Robin Hood of low-cost high Hz, I'm generally a big fan of opensourcing this kind of stuff, even if I hire someone to expand this to a C/C++ codebase that can easily be ported into other projects. Too many people tend to be greedy on this stuff, and I have abandoned an earlier BFI-driver project because someone else was greedy and crossed certain uncrossable red lines. So I need someone else to create a clean-room implementation, completely from scratch, and completely open source. Fortunately I probably don't need to post a bounty soon, since multiple people has already started on different parts of such a project.

For example, Special K now has software-based BFI built in, and I know others are working on multiple Windows indirect drivers. But I'd love to have a high speed real-time Blurinator (capable of Blurinatoring 240 framebuffers of 1/240sec simulated exposures per second for realtime display on a 240Hz monitor in a Windows indirect display driver), especially a Blurinator enhanced to add other sub-refresh temporals than LCD GtG. Such a realtime Blurinator would correctly simulate LCD GtG of a slower LCD on a faster-LCD-GtG display, as an example. Now imagine enhanced so you could simulate a CRT tube, plasma display or DLP projector on a future 1000Hz LCD!

This would have many spinoff applications:
- Simulated pursuit camera images of ANY display (LCD, CRT, OLED, plasma, DLP, etc) simply by plugging the appropriate GPU shader
- Easier to simulate displays before building them (to discover if there's any unexpected artifacts)
- Easier to realtime simulate displays on non-retro displays (e.g. simulate a CRT on a high-Hz LCD)

A universal display emulator library (in a highly portable language) would be able to do all the above! Whether for offline analysis, slow-motion analysis, or real-time display.

GPUs are so hyper-fast today, that an RTX 3080 can be made to simulated real-time temporal dithering for a 4K DLP with 1920 1-bit refresh cycles per second (with some algorithmic optimizations). This was formerly the territory of FPGAs but shadertoy tests shows we now have enough computing power to emulate all retro displays in realtime now! So that means if we had a 1920Hz display, it'd be accurately simulating the artifacts of that DLP projector, for example (including contouring, pixel noise, and rainbows). And a VGA CRT tube can be fairly accurately (to human vision error margins) temporally simulated on a 1000Hz+ LCD. Etc.

That said, until someone goes far enough along, I'm willing to incentivize speeding up such a project. Developers, inquire within, PM *and* email mark [at] blurbusters.com with something that authenticates your identity (e.g. same email that you use for LinkedIn) and your experience.

Ideally I could do this myself, but there are so many Blur Busters projects that Chief Blur Busters can concurrently focus on.
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!

User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: Blurinator 9000 - Blur simulation software

Post by Discorz » 11 Dec 2021, 16:18

Blurinator 9000 is by far my favorite post on BBForums. Very good educational tool. I would absolutely love to see something very similar but more advanced, I feel like we need it. If anyone makes it let me know. I will be glad to test it out.
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Blurinator 9000 - Blur simulation software

Post by Chief Blur Buster » 01 Oct 2022, 16:27

Discorz wrote:
11 Dec 2021, 16:18
Blurinator 9000 is by far my favorite post on BBForums. Very good educational tool. I would absolutely love to see something very similar but more advanced, I feel like we need it. If anyone makes it let me know. I will be glad to test it out.
Long-term, I want to see someone release a CRT simulator, because I'm working on such algorithms.

I added a new post to the RetroArch CRT electron beam simulator feature request -- that mentions Blurinator.

Basically creating a Blurinator style algorithm but for simulating a CRT tube instead of an LCD.
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!

User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: Blurinator 9000 - Blur simulation software

Post by Discorz » 03 Oct 2022, 02:33

Chief Blur Buster wrote:
01 Oct 2022, 16:27
Long-term, I want to see someone release a CRT simulator, because I'm working on such algorithms.

I added a new post to the RetroArch CRT electron beam simulator feature request -- that mentions Blurinator.

Basically creating a Blurinator style algorithm but for simulating a CRT tube instead of an LCD.
This is good news! We're getting there.

Though why not simulate all panels? Ideally the program would need to properly simulate display physics so all technologies could be replicated. A program that allows panel layer stacking and modifying their order. But I can imagine how hard would that be.
Compare UFOs | Do you use Blur Reduction? | Smooth Frog | Latency Split Test
Alienware AW2521H, Gigabyte M32Q, Asus VG279QM, Alienware AW2518HF, AOC C24G1, AOC G2790PX, Setup

User avatar
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Blurinator 9000 - Blur simulation software

Post by Chief Blur Buster » 04 Oct 2022, 22:12

Discorz wrote:
03 Oct 2022, 02:33
Chief Blur Buster wrote:
01 Oct 2022, 16:27
Long-term, I want to see someone release a CRT simulator, because I'm working on such algorithms.

I added a new post to the RetroArch CRT electron beam simulator feature request -- that mentions Blurinator.

Basically creating a Blurinator style algorithm but for simulating a CRT tube instead of an LCD.
This is good news! We're getting there.

Though why not simulate all panels? Ideally the program would need to properly simulate display physics so all technologies could be replicated. A program that allows panel layer stacking and modifying their order. But I can imagine how hard would that be.
Yes, the concept of Blurinator should be ported into a universal display simulator.

We can already simulate DLP color wheels on a 240Hz+ display at www.testufo.com/rainboweffect

I imagine at 500Hz, we can simulate DLP and plasma temporal dithering in slow-motion. Although the LCD GtG would distort/muddy the results, the GtG's between white and black are pretty fast, and would produce educational (even if dim / low quality) results.

Factors to simulate:

Scanout velocity (simple for CRT, LCD, OLED)
Multipass refresh algorithms (DLP, plasma)
Sample and hold curves
...Simulate LCD
...Simulate OLED (OLED GtG is fast but can color-distort OLED impulsing if GtG is different for each color)
Impulsing curves (rise and fall)
...Simulate CRT
...Simulate OLED rolling scan
...Including impulsing overlays on top of sample and hold curves, to allow for backlight phosphor behavior on LCDs
...More accurate if channel split (R/G/B) including KSF simulation www.blurbusters.com/red-phosphor

I believe this can be converged into simple GPU-shader logic.
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!

Post Reply