First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Many sites including LinusTechTips, RTINGS, TomsHardware, and others use the free Blur Busters pursuit camera invention. Now also available as a rail-less smartphone wave, too!
User avatar
Discorz
VIP Member
Posts: 999
Joined: 06 Sep 2019, 02:39
Location: Europe, Croatia
Contact:

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Discorz » 20 Nov 2021, 16:36

I think your camera is not good enough for eye distance display capturing. Fullscreen capturing is used for strobe crosstalk.

This is your second pursuit but with tuned exposure and sharpened a bit. That's how would I do it.

XG2431 240Hz Ultra Fast 960pps - PS edited
XG2431 240Hz Ultra Fast 960pps.jpg
XG2431 240Hz Ultra Fast 960pps.jpg (347.03 KiB) Viewed 10744 times
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

teo
Posts: 33
Joined: 15 Apr 2021, 19:28

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by teo » 21 Nov 2021, 12:45

that edit looks great to me! if anything the it looks slightly better in person, but I think that's mostly down to the colors being better/eyes being better than 4k. I'll mark that as 'the' 240hz example shot for now.

I've reorganized that drive directory a bit, here's a link to the top level. I appreciate you both providing guidance and oversight. I figure I'll try to get at least a 240hz 'normal' and then figure out the whole QFT thing for 60 and 120hz strobing before dumping those images and my subjective thoughts back in the main xg2431 thread.

I did one pass of strobed/pure xp pursuits, and boy does that make it easier to see the sync lines. how do y'all think this one looks?
xg2431 240 extreme 960
xg2431 240 extreme 960
best_ (1).JPEG (1.32 MiB) Viewed 10737 times

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

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Discorz » 21 Nov 2021, 14:32

This is amazing! Now we finally have a bit better view. Thank you for doing these.
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: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Chief Blur Buster » 22 Nov 2021, 21:45

Good strobed pursuit photo! Strobed pursuits is easier, yes.

Which refresh rate? It appears you used factory tuning at a high Hz.
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: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Discorz » 23 Nov 2021, 06:46

Can VTs be pushed further at 240Hz? If yes would it make any difference compared to factory 240 VTs?
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: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Chief Blur Buster » 23 Nov 2021, 15:48

Discorz wrote:
23 Nov 2021, 06:46
Can VTs be pushed further at 240Hz? If yes would it make any difference compared to factory 240 VTs?
Successful large VTs at max Hz is very rare on any panel.
This is because large VT (QFT) simply pushes low-Hz to max-Hz scanout velocity (e.g. 120Hz refresh cycles scanned out in 1/240sec).

The fastest a refresh cycle can occur on a XG2431 is approximately 4ms scanout. The XG2431 scaler/tcon is capable of refreshing 270000 pixel rows per second (270KHz horizontal scan rate). So it can refresh one refresh cycle in 1080/270000ths of a second -- that's 0.004 second -- 4 milliseconds (a little under 1/240sec). This is based on a 1080-pixel row signal in a 1125-pixel-row signal transport (1080 visible, 45 blanking). QFT modes with Large Vertical Totals simply allows a lower-Hz to have the scanrate of max Hz.

Not all pixels on an LCD refresh at the same time, it takes 4ms from the first pixel to last pixel to refresh. That's much faster than a 60Hz LCD, as filmed in high speed video at www.blurbusters.com/scanout

The fastest you can scanout the Innolux 240Hz panel is about 1080/270000ths of a second, regardless of Hz and QFT mode you create. So that's your essential theoretical maximum. So you can have a 4ms scanout at any Hz with the best possible QFT mode. For example, 4ms scanout at 60Hz means your LCD has lots of time to finish LCD GtG pixel response between refresh cycles. But at 240Hz, you literally only have less than 0.2 milliseconds of headroom between refresh cycles, unlike about ~13 milliseconds of headroom between refresh cycles at 60Hz, and ~8.5 milliseconds of headroom between refresh cycles at 120Hz.

Thus, that's why strobe crosstalk decreases more, the lower Hz you can go. You've got more headroom to create a larger blanking interval (pause between refresh cycles).

The world's first perfect zero-crosstalk strobed LCDs were only finally achieved recently; not a single faint duplicate image. With a full tuning, the XG2431 manages to have less crosstalk than NVIDIA LightBoost (a gold standard in low-crosstalk strobing), when tuned to the maximum with QFT and Strobe Utility, combined with a 120Hz-or-less mode. And do it without the color degradation of LightBoost. Even a CRT-like color gamut is preserved (with the exception of poorer blacks of LCDs versus CRT). Zero crosstalk at screen center, and ultra faint crosstalk at top/bottom that disappears the lower Hz you go.

Stock 60 Hz can't go zero crosstalk this because it's not a QFT mode, it slow-scans in (1080/1125)ths of 1/60sec (pretty much a bit above 16ms to scanout). But using QFT 60Hz (Vertical Total 4500) goes perfect zero crosstalk top/center/bottom after a good Strobe Utility tuning session at www.blurbusters.com/xg2431

Incidentially, it is also how Quest 2 VR LCD and a Valve Index VR LCD does it; it does an accelerated low-latency scanout at low-Hz (like a large vertical total). It is likely a ~240Hz-ish LCD being underclocked to 72Hz, but now has a 90Hz and 120Hz mode via firmware update (with improved strobe tuning that made higher Hz possible).

Modern 1ms LCDs (IPS and TN) with approximately a 3:1 refresh rate headroom (e.g. 240Hz LCD underdriven at 75Hz) can go perfect zero crosstalk top/center/bottom better than LightBoost, with proper tuning.

Image

With QFT modes, crosstalk can fall far below 1% and become perfect zero crosstalkless.

The XG2431 achieves less than 1% crosstalk when you use Strobe Utility + QFT Mode + Refresh Rate Headroom + Framerate=Hz. You have to combine all the possible tricks documented at www.blurbusters.com/xg2431 in order to pull off the perfect crosstalkless strobed LCD mode that you cannot achieve on any other gaming monitor.

Generaly this occurs in the 60Hz-80Hz region, but 100Hz-120Hz can get under 1% crosstalk for screen centre if you warm up for an hour first and tune expertly on a VT2250 QFT mode (120Hz) or VT2700 QFT mode (100Hz), there will still be faint crosstalk (typically under ~10%) for top/bottom that completely disappears at ~80Hz and less.

Choose your refresh rate headroom based on how quality-priority strobing you want to do.

In my opinion, if your game can do above 60fps, then using 120Hz is the sweet spot, but 100Hz can make it easier to do framerate=Hz in more games.
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!

teo
Posts: 33
Joined: 15 Apr 2021, 19:28

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by teo » 23 Nov 2021, 21:48

Chief Blur Buster wrote:
22 Nov 2021, 21:45
Good strobed pursuit photo! Strobed pursuits is easier, yes.

Which refresh rate? It appears you used factory tuning at a high Hz.
the one I posted in this thread is 240hz at 'extreme' and I also did one for 'normal.' my assumption (which I think is supported by your explanation) is that the factory 240hz modes are as best as you can get for a given brightness level. I could probably live with a brightness slightly in between the normal and extreme presets, but I already feel like I'm splitting hairs to spot the difference between those two modes. the difference between 240hz strobed vs unstrobed is pretty stark for me in ufo tests, though I'm not sure I appreciate the difference in game trying to A/B them. I'm staying on strobed for a few days to see if I can notice a change after adjusting to it.

I did give 60hz and 120hz a go after figuring out that the 'stock' 240hz mode is in the extension block. I'm able to double the VT to 2250 for 120hz. indeed at center screen (I'd say the middle 33%-ish) there's no detectable cross talk, on the lower third it's still difficult to spot, and the upper third is a bit more noticeable. I'm not sure what happened to the left-of-center column, some sort of photo artifact? shutter speed was 1/120 for this stationary photo.
120 strobe full screen (ROTATED).JPG
120 strobe full screen (ROTATED).JPG (5.01 MiB) Viewed 10533 times
[Click for Zoomed Image]

for 60hz, I wasn't able to achieve 4500 VT. I could get an image at 4225 VT, but the screen would flicker to black every few minutes. this didn't happen at a VT of 4200, so I tuned there and it did result in a cross-talk free image from top to bottom. using those steps, though, results in a refresh rate of 64hz. I admittedly didn't investigate much, but is there a way to get this to 60hz while maintaining that 4200 VT? or is the discrepancy of a 64hz refresh rate vs 60 fps content not going to be noticeable? I fired up sonic mania as a test case for real world content and...I'm not really sure I notice a difference even when I look for it. I'm also assuming the bottom row is a photo artifact of some kind. shutter speed was 1/60
60 strobed full screen (ROTATED).JPG
60 strobed full screen (ROTATED).JPG (5.15 MiB) Viewed 10533 times
[Click for Zoomed Image]

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

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Chief Blur Buster » 23 Nov 2021, 23:44

Fantastic 120Hz strobed crosstalk photo. Yes, there's a shutter artifact. A camera sensor slowly scans, so part of sensor captured one strobe, and another part of sensor captured a different strobe -- this can vary from camera to camera if you're not doing pursuit camera with strobing. Pursuit camera avoids this type of artifact.

Can you post a screenshot of your Strobe Utility settings for your VT4225 QFT mode?
I presume you followed instructions at www.blurbusters.com/xg2431 when tuning strobe.

GammaLyrae managed VT4525 with this Strobe Utility setting:
viewtopic.php?t=8113&start=360#p70213

Try re-read the Quick Frame Transport HOWTO to make sure you're properly configuring your QFT mode. You need to modify a max-Hz mode with unmodified Pixel Clock numbers;
teo wrote:
23 Nov 2021, 21:48
could probably live with a brightness slightly in between the normal and extreme presets, but I already feel like I'm splitting hairs to spot the difference between those two modes.
You need TestUFO Panning Map Test at 2000 pixels/sec or 3000 pixels/sec to begin to notice sub-millisecond MPRT differences.
Try this: TestUFO Panning Map, 2400 pixels/sec. The labels will remain blurry at 1ms MPRT, but becomes clear at 0.5ms MPRT.

Use Custom PureXP with Strobe Utility.
You can use Strobe Utility setting 25 to create a brightness exactly between Normal and Extreme. It is persistence (MPRT100%) in percentages of a refresh cycle.

Light = 40% refresh cycle persistence
Normal = 30% refresh cycle persistence
Extreme = 20% refresh cycle persistence
Ultra = 10% refresh cycle persistence
Custom = completely adjustable in 1% increments from 1% to 40% with Blur Busters Strobe Utility for ViewSonic monitors

When creating VT4500, it becomes easier if you first create a working 240Hz VT1125 mode first. Then derive 120Hz by doubling VT, (2250) and derive 60Hz by doubling VT again (4500) as per instructions at QFT HOWTO. A working maximum-VT 240Hz mode makes it easier to create maximum-QFT modes on the XG2431 with exactly the same "Pixel Clock" number and "Horizontal Scan Rate" number as the original working 240Hz mode you're basing your low-Hz QFT modes off. The litmus test is making sure the rate of pixels transmitted over the video cable is exactly as it was with the 240Hz mode, e.g. Pixel Clock of 600Mhz = 600,000,000 pixels per second transmitted over the DisplayPort cable. Keeping Pixel Clock exactly identical as the original 240Hz mode you're deriving from, helps ensure that the QFT mode still works reasonably reliably on the XG2431. It's the art of serializing a 2D image over a 1D cable, transmitting one pixel at a time (via a packetized pixel row transmission), and keeping the signal as perfectly compatible as 240Hz, is the easiest way to create a QFT mode that is compatible without glitches.

If your original 240Hz mode is not 1125 then remember to multiply the original number by exactly 4 to get exactly 1/4th refresh rate. So if your factory 240Hz mode is Vertical Total 1098, then you want exactly 4392, with the radio button in the correct locations in ToastyX to keep all ToastyX numbers perfectly unmodified except for Vertical Total (to generate the correct refresh rate). If you want VT4500, you may want to create your 240Hz VT1125 mode first.

TIP: Also for 60Hz QFT (signal VT above 4000), you want to use a Strobe Phase of between ~85 to ~98 (it's a wraparound effect) to eliminate the crosstalk at the very bottom row of your 60Hz for perfect full height zero crosstalk operations. Strobe Phase 100 is exactly equal to Strobe Phase 0, in a true wraparound fashion.
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
Chief Blur Buster
Site Admin
Posts: 11647
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by Chief Blur Buster » 24 Nov 2021, 00:00

teo wrote:
23 Nov 2021, 21:48
I'm not really sure I notice a difference even when I look for it
  1. Try temporarily using VSYNC ON if you want fluidity-priority strobing (though overkill framerates during VSYNC OFF compensates a lot)
  2. Many games will automatically switch refresh rates, which can interfere with PureXP benefits. Make sure your game is using the correct refresh rate when you launch the game. Check your monitor menus and refresh rate to make sure you're still running at the refresh rate that you strobe-tuned. Otherwise benefits are lost.
  3. Motion blur reduction benefits occur more if you track moving objects while panning, instead of stationary gaze. PureXP will usually benefit crosshairsless games more than games with crosshairs (because blur reduction doesn't have as much effect, albiet there's a sort of an aim stabilizer effect).
    HOWTO: Using ULMB Beautifully or Competitively.
  4. Strobing amplifies mouse microstutters, the lack of blur makes the tiniest stutters much more visible.
    Are you running your mouse in high-DPI low-sens operations, preferably at 2000Hz?
Sonic Mania in an emulator does not use the mouse, so skip this. However, if you're playing a game that uses mouse panning or mouse turning, you may want to also optimize your mouse to be more strobing optimized.

Recommended Mouse Settings For Strobed Operations

Only a few mice supports this; preferably a new >2000Hz+ mouse with a 3389 or 3399 sensor.
For FPS strobing, recommended ultrafluid mouse settings for strobing is:
- 3200dpi
- 1/8th sensitivity as you would normally at 400dpi, or 1/4th sensitivity as you would normally at 800dpi.
- 2000Hz or 4000Hz poll rate

CS:GO will not work well with this (bad mouse math at 1600dpi and 3200dpi), but Valorant and Overwatch works well with it.

Bluntly, the mouse jitters of yesterday's 1000Hz pollrate 400dpi and 800dpi kills benefits of strobing in many ways, so it kinda sabotages many advantages of strobing. The "strobe jitter" effect. But having the right mouse & right mouse settings fixes the strobe jitters.

This limits your mouse purchase choice to mouse models such as Razer Viper 8KHz.
Alternatively, to bypass mouse microstutters (that kills benefits of strobing such as PureXP), are you testing fastscrollers with VSYNC ON and gamepad or keyboard?

Image

Image
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!

woland
Posts: 27
Joined: 06 Aug 2020, 16:38

Re: First-Timer: XG2431 Pursuit Camera Photos By Hand-Wave Smartphone

Post by woland » 24 Nov 2021, 01:49

Hi Chief, you've mentioned in numerous posts now that "CS:GO will not work well with this (bad mouse math at 1600dpi and 3200dpi)" - what exactly happens in csgo at 1600-3200 dpi?

Post Reply