Repeated images on sample and hold displays at low frame rates

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
fojodot
Posts: 2
Joined: 03 Jan 2022, 16:39

Repeated images on sample and hold displays at low frame rates

Post by fojodot » 03 Jan 2022, 19:56

This likely isn't anything new to anyone here, but the effect I'm seeing is a kind of double or quadruple image effect at 30 FPS and 15 FPS respectively in the UFO Test, independent of refresh rate and without using any kind of motion blur reduction settings like backlight strobing or black frame insertion. It's visible on a 60 Hz TN laptop screen, 10-15 year old 60 Hz LCD TV, and 120 Hz OLED TV.

I would like to have a better understanding as to why I'm seeing that. My understanding before was that double images are more of a characteristic of impulse-driven displays when flashing a frame twice, but I'm seeing a similar effect with single, persistent frames, albeit with much more blur/stutter. It's even there at 60 FPS, though much less obvious.

My thinking is that this is related to the stutter/blur threshold with sample and hold, and that these repeated images are a result of this stutter at lower frame rates, where the UFO is vibrating back and forth so fast that we're seeing the UFO (and stars) placed in different positions across the screen as our eyes track it, but not fast enough to completely blur together, but is that completely off the mark?

I couldn't find any real mention of this effect in articles and older posts I came upon. Maybe I'm misunderstanding what persistence blur actually is in the first place :lol:

Feel free to move this to a more suitable sub-forum, if mods wish.

User avatar
Chief Blur Buster
Site Admin
Posts: 9896
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Repeated images on sample and hold displays at low frame rates

Post by Chief Blur Buster » 03 Jan 2022, 20:16

fojodot wrote:
03 Jan 2022, 19:56
This likely isn't anything new to anyone here, but the effect I'm seeing is a kind of double or quadruple image effect at 30 FPS and 15 FPS respectively in the UFO Test, independent of refresh rate and without using any kind of motion blur reduction settings like backlight strobing or black frame insertion. It's visible on a 60 Hz TN laptop screen, 10-15 year old 60 Hz LCD TV, and 120 Hz OLED TV.

I would like to have a better understanding as to why I'm seeing that. My understanding before was that double images are more of a characteristic of impulse-driven displays when flashing a frame twice, but I'm seeing a similar effect with single, persistent frames, albeit with much more blur/stutter. It's even there at 60 FPS, though much less obvious.

My thinking is that this is related to the stutter/blur threshold with sample and hold, and that these repeated images are a result of this stutter at lower frame rates, where the UFO is vibrating back and forth so fast that we're seeing the UFO (and stars) placed in different positions across the screen as our eyes track it, but not fast enough to completely blur together, but is that completely off the mark?

I couldn't find any real mention of this effect in articles and older posts I came upon. Maybe I'm misunderstanding what persistence blur actually is in the first place :lol:

Feel free to move this to a more suitable sub-forum, if mods wish.
This is not unheard of. Not everyone sees this -- but I have seen this (very faintly, some sees it more clearly). It is typically a very faint double-image optical illusion caused by a human-brain-related stimuli that is different than from strobing.

A sample-and-hold double image effect is probably triggered by the extremes of the stutter distance (amplitude of the object jump). Our brains seem to more easily register the stutter-jumps (the slideshow-to-motion threshold is a flagged threshold here) than the motion blur in between the frame-steppings. Which can mean our brains sometimes register the extremities of the stutter (furthest leftmost and furthest rightmost), as separate images, during certain low frame rates.

Many can force themselves to see a double-image effect at certain frame rates, simply by staring at the smoothest moving UFO followed immediately by the UFOs below them.

Your smooth eye pursuit, versus the stepping of the frames, can create an (essentially) optical illusion of double image effect (even if fainter than on an impulsed display)



There is a threshold where stutters blend to motion blur, which has to do with the stutter-vibrations become too fast to see (blends to motion blur) like a fast-vibrating string. Sometimes when we stare at a fast-vibrating string, we see two strings -- the outer edges of a string vibration.

You can watch the stutter-to-blur continuum of a sample-and-hold display by watching the variable speed version of TestUFO Eye Tracking (watch the second UFO through the entire framerate range). Sometimes this animation, too, kind of amplifies a sample-and-hold double image effect at certain frame rate thresholds for some people's. Although research is fairly scant on this, I attribute it to the stimuli of the extremities of the stutter-jump registering as two separate objects.

A secondary explanation is that pixel transitions can be relatively slow, and there can be ghosting (which can create a brief flash of a corona or a ghosting, typically for one refresh cycle, even if the frame rate is far lower than refresh cycle). As the 2nd object fades in, the 1st object fades out, so temporarily, there are two simultaneous UFOs. This may or may not amplify the stimuli.

I have noticed over the years, some people perceive a low sample-and-hold frame rate exhibit a double-image-effect -- and even on fast OLED displays (even more visible than on slow LCDs) simply because it's the outer extremities of the stutter jump distance.

There is, however, a different TestUFO with an intentional double image effect: blatantly intentional TestUFO double-image effect simulated by double-strobing a software-based black frame insertion.

Obviously, more research may be needed for a non-impulsed double image effect, to see if this pertains to the lower sensitivity threshold or the higher sensitivity threshold (the slideshow-to-motion threshold, versus the stutter-detection threshold), but there is some blending between some of the thresholds to create combined motion illusions. Some of these motion illusions are not as reliably consistently seen by a human population.

(crossposting from Area 51)
Chief Blur Buster wrote:Many people misunderstand the different sensitivity thresholds, such as "Humans can't see above 75Hz" -- but that is only a flicker threshold. The purpose of this post is to show that there are extremely different orders of magnitude that refresh rate upgrades do address.

Even in a non-gaming context, one thing many people forget is that there’s many thresholds of detectable frequencies.

These are approximate thresholds (varies by human), rounded off to nearest order of magnitude for reader simplicity of how display imperfection scale.

Threshold where slideshows become motion: 10
This is a really low threshold such as 10 frames per second. Several research papers indicate 7 to 13 frames per second, such as this one. This doesn't mean stutter disappears (yet), it just means it now feel like motion rather than a slideshow playback.
Example order of magnitude: 10

Threshold where things stop flickering: 100
A common threshold is 85 Hz (for CRTs). Also known as the “flicker fusion threshold”. Variables such as duty cycle (pulse width) and whether there’s fade (e.g. phosphor fade) can shift this threshold. This also happens to be the rough threshold where stutter completely disappears on a perfect sample-and-hold display.
Example order of magnitude: 100

Thresholds where things stop motion blurring: 1000
Flicker free displays (sample and hold) means there is always a guaranteed minimum display motion blur, even for instant 0ms GtG displays, due to eye tracking blur (animation demo). The higher the resolution and the larger FOV the display, the easier it is to see display motion blur as a difference in sharpness between static imagery and moving imagery, blurry motion despite blur free frames (e.g. rendered frames or fast-shutter frames).
Example order of magnitude: 1000

Threshold for detectable stroboscopic effects: 10,000
Where mouse pointer becomes a continuous motion instead of gapped. This is where higher display Hz helps (reduce distance between gaps) and higher mouse Hz (reduce variance in the gaps). Mouse Hz needs to be massively oversample the display Hz to avoid mouse jitter (aliasing effects). If you move a mouse pointer 4000 pixels per second, you need 4000Hz to turn the mouse pointer into a smooth blur (without adding unwanted GPU blur effect).
Example order of magnitude 10,000

An example:
Image
(From lighting industry paper but has also been shown to be true for stroboscopics on large displays, including VR displays intended to mimic the real world)

More information can be found in Research Section of Blur Busters.
Many of these thresholds are VERY well researched, with hundreds of scientific papers, but the double-image-effect on sample-and-hold (with no PWM, strobing, or impulsing effects) is a niche subsegment-of-humans display motion artifact that is fairly research-scant.

Some of the double image on sample and hold display is a LCD GtG artifact (and is definitely captured on pursuit camera), while some of the double image on a sample and hold display is a human-brain-triggered optical illusion (not captured by pursuit camera). I personally see the double image effect on the 36fps UFO on a fast-GtG 144Hz NanoIPS panel, and it is apparently faintly captured by pursuit camera so I treat it as a kind of a LCD GtG artifact. So there appears to be a fuzz of multiple prevailing factors.

Double Image Effects Capturable by Pursuit Camera
Here's a pursuit camera of a double-image effect caused by LCD ghosting:
Image
This often happens when LCD GtG pixel response is longer than one refresh cycle. -- Look the faint yellow double-image trailing behind the yellow UFO dome. That's a LCD GtG-derived double image effect, as pixel response is metaphorically like kicking soccer balls multiple times ever closer to the finish line (imagine goalie line = final pixel color). The 2nd refresh cycle fades it partially (kicks the pixel closer to final color but not completely), then the 3rd refresh cycle fades it more fully (final kick into goalie line). So you have a LCD GtG stairstepping effect caused by the voltage of the multiple refresh cycle passes pulsing the same pixels once per refersh cycle, kicking the LCD pixel colors ever-closer to final pixel color. The stairstepping usually doesn't exist on ultrafast-GtG displays such as OLEDs or direct discrete-LED displays (no LCD layer). Yet, this is not the only cause of the sample-and-hold double image effect.

Double Image Effects NOT Capturable by Pursuit Camera
However, the double image effect (not capturable on pursuit camera) also occurs on an OLED too, and I attribute this to an optical illusion effect created by the human brain, caused by the brain more suddenly registering the extremities of the stutter amplitude, rather than the universally capturable persistence-blur in between.

So there is apparent existence of more than one factor that generates a human-perceived double image effect on motion on a sample-and-hold LCD, at least among a small portion of humans.

BTW, where did you hear about Blur Busters Forums? You came to the right place for these kinds of nuanced display-behaviours questions. :D
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

       To support Blur Busters:
       • Official List of Best Gaming Monitors
       • List of G-SYNC Monitors
       • List of FreeSync Monitors
       • List of Ultrawide Monitors

User avatar
jorimt
Posts: 1739
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Repeated images on sample and hold displays at low frame rates

Post by jorimt » 03 Jan 2022, 20:42

I was under the assumption that with low framerates on a sample and hold display without strobing during panning motions (ruling out any GtG or overdrive artifacts), it's mainly because since each frame scan is effectively wiping over the other, the lower the average framerate, the longer it takes the next frame to be scanned fully over the previous, thus the longer the previous frame persists under the next, with the offset between manifesting as a visible "gap" (or afterimage) of sorts when eye tracking.

I.E. an evident doubled image, with the "gap" being the wider and more noticeable the higher the average frametime is (30 FPS = ~33.3ms persistence gap, ~15 FPS = 66.7ms gap, etc).

Such a phenomena would be somewhat obscured by the higher GtG blur of older LCD displays, but would become more apparent as GtG on modern LCDs decreases (which it has been), leaving MPRT more evident instead, especially at lower average framerates.

This is a reason many people complain of the "judder" in 24Hz content on a OLED, even with 5:5 pulldown, since with its virtual 0ms "GtG," it can't obscure the persistence with any form of blur at all, making the high MPRT of 24 FPS content all the more apparent than on sample and hold displays with slower pixel response times (like a VA-type LCD TV, for instance).

Anyway, I clearly perceive the oscillation of the TestUFO at 30 FPS and 15 FPS as well. If you fix your gaze just above the 15 FPS pattern and use your periphery, you should notice that it stops looking like oscillation and starts looking more like a linear slideshow instead.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG279QM, LG 48CX OS: Windows 11 Pro MB: ASUS Z690 Extreme CPU: i9-12900k GPU: EVGA RTX 3080 FTW3 UG RAM: 32GB G.SKILL Trident Z5 @5600MHz

thatoneguy
Posts: 109
Joined: 06 Aug 2015, 17:16

Re: Repeated images on sample and hold displays at low frame rates

Post by thatoneguy » 04 Jan 2022, 05:20

Chief Blur Buster wrote:
03 Jan 2022, 20:16


Interesting that only the 3rd row(14fps for me) feels like a double image to me when I'm eye-tracking.
The 2nd row(26-28fps) feels more like strobe-talk where you can see a trail but it's not exactly full-on double image.

EDIT: Wait, nevermind. It wasn't loading properly for a second there.
I see the double image in the 2nd row now.

User avatar
Chief Blur Buster
Site Admin
Posts: 9896
Joined: 05 Dec 2013, 15:44
Location: Toronto / Hamilton, Ontario, Canada
Contact:

Re: Repeated images on sample and hold displays at low frame rates

Post by Chief Blur Buster » 04 Jan 2022, 12:04

jorimt wrote:
03 Jan 2022, 20:42
I was under the assumption that with low framerates on a sample and hold display without strobing during panning motions (ruling out any GtG or overdrive artifacts), it's mainly because since each frame scan is effectively wiping over the other, the lower the average framerate, the longer it takes the next frame to be scanned fully over the previous, thus the longer the previous frame persists under the next, with the offset between manifesting as a visible "gap" (or afterimage) of sorts when eye tracking.
It was already determined that global refresh brhavior has the same double image effect.

Low frame rates at 1/60sec scan vs 1/360sec scan does not appear to affect this (in enough people tested, and myself), so finite speed of scanout appears not the cause.

It’s more the human stimuli of the sudden frame change, registering both previous frame (disappearing) and new frame (appearing), counting both into a double image when this stimuli is not masked by blurrings (GtG and/or MPRT combined).

Also, existing optical illusion science suggests this may play a role too, given sufficient study.

Most of the rest of your post is accurate, however.

It would be cool to see peer reviewed research on this, albiet there’s a backlog of higher priority papers many researchers are doing…
Head of Blur Busters - BlurBusters.com | TestUFO.com | Follow @BlurBusters on Twitter

       To support Blur Busters:
       • Official List of Best Gaming Monitors
       • List of G-SYNC Monitors
       • List of FreeSync Monitors
       • List of Ultrawide Monitors

User avatar
jorimt
Posts: 1739
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Repeated images on sample and hold displays at low frame rates

Post by jorimt » 04 Jan 2022, 12:28

Chief Blur Buster wrote:
04 Jan 2022, 12:04
It was already determined that global refresh brhavior has the same double image effect.
Ah, good to know. I've never witnessed global refresh, so that rules finite scanout time out as being a contributing factor or cause.

Then it would seem to be triggered almost entirely by how the eye tracks, and whatever illusory effects that creates optically. This seems especially confirmed when considering this part of my previous comment:
If you fix your gaze just above the 15 FPS pattern and use your periphery, you should notice that it stops looking like oscillation and starts looking more like a linear slideshow instead.
That said, regarding:
Chief Blur Buster wrote:
04 Jan 2022, 12:04
Low frame rates at 1/60sec scan vs 1/360sec scan does not appear to affect this (in enough people tested, and myself), so finite speed of scanout appears not the cause.
I was more generally referring to this "illusion" on sample and hold displays (without strobing) being caused by the increase of time between each frame due directly to the lower average framerate (AKA higher MPRT, as in unique frames will persist on-screen longer, regardless of global or traditional scanout), and not as much due to the given scanout time of the monitor (though I can see how it could be interpreted that way by my original wording); it stands to reason 15 FPS, for instance, would appear the same in this respect whether you're viewing it on a 60Hz or 360Hz display, assuming both have comparable GtG performance.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG279QM, LG 48CX OS: Windows 11 Pro MB: ASUS Z690 Extreme CPU: i9-12900k GPU: EVGA RTX 3080 FTW3 UG RAM: 32GB G.SKILL Trident Z5 @5600MHz

fojodot
Posts: 2
Joined: 03 Jan 2022, 16:39

Re: Repeated images on sample and hold displays at low frame rates

Post by fojodot » 04 Jan 2022, 18:19

Thanks for the informative responses, much appreciated.

It does seem to show up as strong or even stronger on OLED to my eyes, despite the very fast GtG response. That's what made me lean towards the more harsh stutter being a trigger for it. But, yes, it shows up very clearly to my eyes whether it's OLED or LCD (though with a layer of sample and hold fog over it).
Chief Blur Buster wrote:
03 Jan 2022, 20:16
BTW, where did you hear about Blur Busters Forums? You came to the right place for these kinds of nuanced display-behaviours questions. :D
I think my exposure to this site has continually grown over time, so I'm not sure I'd be able to point to a single place and say that's where I heard of you from :) I suppose having an interest in display technology and general perusing of the web is what led me here.

User avatar
jorimt
Posts: 1739
Joined: 04 Nov 2016, 10:44
Location: USA

Re: Repeated images on sample and hold displays at low frame rates

Post by jorimt » 04 Jan 2022, 21:44

fojodot wrote:
04 Jan 2022, 18:19
It does seem to show up as strong or even stronger on OLED to my eyes, despite the very fast GtG response. That's what made me lean towards the more harsh stutter being a trigger for it.
It would be because of the more instantaneous pixel response of the OLED that this phenomena is more evident, not despite it.
fojodot wrote:
04 Jan 2022, 18:19
But, yes, it shows up very clearly to my eyes whether it's OLED or LCD (though with a layer of sample and hold fog over it).
The persistence caused by sample and hold ultimately allows the effect, with low framerate content making it more evident. Good strobing on sample and hold eliminates/mitigates the persistence, but introduces it's own characteristics at that point.

All told, however, this can be defeated on sample and hold (without strobing) simply by brute forcing through it with higher refresh + framerate ratios; at 240 FPS+ in this thread's test pattern (on my 240Hz display), for example, the effect is virtually imperceptible (at least at that scrolling speed).

The bottom-line is, the lower the average frame rate and the higher persistence display (especially those with instantaneous pixel response), the more the illusion of smooth, continuous movement is broken, exposing display technology for what it is; a series of static images shown in sequence really, really fast.

So in that context, I guess you could say the effect is not so much creating an illusion, as much as it is "breaking" the intended one.
(jorimt: /jor-uhm-tee/)
Author: Blur Busters "G-SYNC 101" Series

Displays: ASUS PG279QM, LG 48CX OS: Windows 11 Pro MB: ASUS Z690 Extreme CPU: i9-12900k GPU: EVGA RTX 3080 FTW3 UG RAM: 32GB G.SKILL Trident Z5 @5600MHz

Post Reply