Flicker-free tablets

Discussion in 'Hardware' started by Jamon, Oct 22, 2014.

  1. Robgoren

    Robgoren Pen Pal - Newbie

    Messages:
    30
    Likes Received:
    5
    Trophy Points:
    16
    You're misunderstanding me. Through PWM, a current remains at its absolute power output and is switched on and off at a particular frequency (the duty cycle) when the brightness is dropped below 100%. (You can eliminate PWM in a makeshift way by setting the brightness to 100%, downloading a dimming app, or fiddling with the graphic card properties to reduce the brightness and contrast.)

    Nevertheless, during the OFF period, the backlight is consuming NO power. Therefore, using a very low frequency or duty cycle, like the Surface Pro 3 below 50% brightness, means the backlight can sip power, and the battery life can be extended. Opting for a higher duty cycle would entail drawing more power because it's closer to constant current. The Surface 3 has no detectable PWM because the backlight is operating at a very high frequency, like (and I'm presuming here) all the tablets listed here as either PWM or "flicker free."

    Now obviously, an i7 does not necessitate the use of PWM. The point is, for CPUs with higher TDPs, where conserving energy is paramount, there are benefits to using PWM with a low duty cycle, which corresponds to lower power consumption.

    On the other hand, analog dimmers require additional circuits to supply a direct current to the backlight in proportion to the brightness of the screen. The main problem with running a direct current to the backlight is that LED diodes have a threshold voltage, below which they will not work at all, and also struggle to achieve the correct color temperature. This means that voltage-based dimming can require more power to surmount that threshold at lower brightness levels. Moreover, the circuits must deal with the heat generated by the LEDs, which increases the current flowing through them and greatly reduces their lifespan. With PWM, the current remains constant, preventing an overload. In analog dimming, the variable current supplied to the LEDs means the regulator supplying the current must also absorb the surplus power arising from the voltage across the LED. The regulator and semiconductors dissipate the additional power internally, but it ultimately results in wasted energy and heat -- heat which complicates thermal design.

    When implemented well, analog dimming has shown to be more efficient than PWM, but it's still complicated and expensive.
     
  2. Robgoren

    Robgoren Pen Pal - Newbie

    Messages:
    30
    Likes Received:
    5
    Trophy Points:
    16
  3. jtl

    jtl Pen Pal - Newbie

    Messages:
    2
    Likes Received:
    0
    Trophy Points:
    5
    That's a generic method for adjusting PWM frequency on laptop's with Intel graphics chip. Works for Windows, mac, linux.

    I recently spoke with someone at the display department of Apple and he says that recent MacBook's "flicker" at 100 KHz below 50% brightness. After that conversation I purchased a 2015 MacBook Pro with AMD Graphics. No problems!

    Test video using Canon T5i

     
  4. LucaD0

    LucaD0 Pen Pal - Newbie

    Messages:
    1
    Likes Received:
    1
    Trophy Points:
    6
    Can somebody confirm if there is pwm flicker on the Microsoft Surface Pro 4?

    "Update 12/01/2015: Thanks to a readers tip, who told us that he is suffering from problems caused by flickering / PWM, we looked at the Core i5 version again with the newest firmware and drivers. Now we could also measure flickering at 50 Hz below 50 percent brightness without a connected power adapter. Analyzing our readings, it does not seem to be classic PWM but some other form of flickering."
    (http://www.notebookcheck.net/Microsoft-Surface-Pro-4-Core-m3-Tablet-Review.153843.0.html)
    I'm really curious to know what "other form of flickering" is used.

    I personally was unable to detect any PWM with iPhone slow motion @ 240 fps on the Surface Pro 4 display.
    I was able to use the Surface Pro 3 for 1 year because there was no pwm flicker if the brightness was higher than 55%. I need to know if it's the same with the Surface Pro 4 or i will return it.

    I've made a video comparison of the monitor flicker (PWM) on the Surface Pro 3 vs Surface Pro 4 vs MacBook Pro, please watch it, at 4:33 you can se some slight random flicker, is that pwm?

     
    thatcomicsguy likes this.
  5. TallSean

    TallSean Pen Pal - Newbie

    Messages:
    2
    Likes Received:
    0
    Trophy Points:
    5
    Hello. I'm new to this forum and am VERY thankful to see this information online since PWM triggers my migraines. I bought two tablets for Christmas gifts (Galaxy Tab S2 and Galaxy Tab A) and was able to test them for flicker using the digital camera test. The results were very clear, and somewhat surprising. The cheaper Galaxy Tab A 8" had no flicker even at the lowest brightness setting. The more expensive Galaxy Tab S2 9.7" had flicker at all brightness settings below 80%. That tablet is being returned to the store today.
     
  6. Jamon

    Jamon Scribbler - Standard Member

    Messages:
    260
    Likes Received:
    64
    Trophy Points:
    41
    I doubt it's as simple as "PWM is bad". Many of the "flicker-free" displays use PWM, but at higher frequencies. The main problem was that for a time all the LED-backlit displays, in everything from desktop monitors to laptops, were using very low frequencies where at anything less than 100% brightness it was like a strobe light. I mean literally like a strobe light, where when I'd be working in a dark room staring into my dimmed display, as I looked around it was part of my visual experience to see trails, moving waves, and other kinds of things you see at a dance party with a strobe light. I preferred the pencil test, which I did using my finger, where I'd wave my finger or stylus in front of the screen, and televisions, monitors, laptops would often show a motion trail of sharp pencil traces. Whereas if you wave your finger in front of a sun-lit window, you see motion blur, and only sharp fingers at the start and end points, not a ton of them in between like someone is copy/pasting your finger in the sky.

    That strobe effect was nasty. I'd stare into it for hours every day, and it didn't kill me, but it was stressful and difficult sometimes to focus and see things on the screen through the distraction of all the flashing, which was most apparent whenever my eyes moved to look somewhere else. If you hold perfectly still and look only in one spot at a non-moving image on the screen you might not notice any strobing effects at all. Then you turn your head, and there's a trail of rectangles from the display tracing along the path of motion through time.

    But right now I'm using a desktop monitor that is marketed as flicker-free, and yet it still uses PWM. If you hold your digital camera up to it, you won't see anything on the viewfinder. You'll be tricked into believing because it's advertised as flicker-free and passes the camera test that it doesn't use PWM. But it does. It's just that the PWM frequency is in the upper kHz, so instead of a very obvious strobe effect, it blends together better and isn't really visible. Maybe you can sense it in some hairline way. In trying displays that supposedly are flicker-free, I've still felt like something was off. That I was not seeing a perfectly stable light, or something natural feeling like outdoors. There's still some weird motion to these devices, but I can't see it. So I don't know why I feel that way, and if I'm just crazy. That's why I'd like to try an E ink display to compare.

    But even looking at this display with PWM, it isn't really a problem. Maybe my brain is having to do more work to filter things because of the pulsing light, but it's much better than when I could see it obviously like a strobe light that was distracting in more direct ways.

    So be careful not to get too carried away with the purity pursuit, because there is no purity in the practical world, and if you get too extreme you'll be dying in a bubble. None of these displays are perfect, you just don't want something ridiculous like 200 Hz. It should be in the kHz.

    I think they need PWM for consistency and high-quality properties of things like color. I doubt there's any high-quality displays that don't use PWM at all.

    But because the camera test is breaking down now for a lot of people, here's the new version:
    1. Increase digital camera shutter speed to fastest
    2. Point it at a white screen
    3. Slowly move a finger or pencil back and forth in front of screen (in the air between the screen and camera lens)
    4. Compare this to the same in front of a sunlit window
    I think with this test, you will see the strobe trails in the camera preview or the recorded video in almost all flicker-free displays. But you won't see it with your eyes if you repeat the same test using your vision directly instead of through a camera. That's because the camera is shifting the temporal aspect beyond what your eye perceives, like using a microscope to see things you otherwise can't. The microscope shows you things that do exist, and if you see scary little mites and bacteria on your skin you might freak out. You might feel like something is wrong, and take a long shower scrubbing yourself clean. But the thing is, that stuff is always there, you just don't ordinarily see it with your normal vision. It took a microscope to let you in on the secret. Similarly, most of the screens you find in the marketplace today will flash at the microscopic level. The only real problem is when it's a practical distraction, not the fact that there's PWM.

    Unless of course we discover someday that everyone is going blind at age 55 because they grew up with these devices. But we're pioneers and using the most advanced technology we can, so even if that happened, such is life.

    In general, if you see obvious rolling waves and bands cycling across the screen in the digital camera viewfinder, then it's probably too low frequency. If you only see motion trails with the pencil in the digital camera viewfinder, it's probably high enough frequency to not cause major visual stress. If you see the motion trails with a moving pencil and your bare eyes, then it's too low frequency. Compare all these tests in front of a window, and when a display comes close to matching what motion looks like in front of a window then it's good enough. If you need something more biocompatible, try stuff like Sony Digital Paper, or just use paper.
     
    Marty likes this.
  7. jtl

    jtl Pen Pal - Newbie

    Messages:
    2
    Likes Received:
    0
    Trophy Points:
    5
    My Canon DSLR can do 1/4000 shutter speed video recording. Wonder what the frequency bottleneck of oscilloscope+photodiode testing is.

    I emailed NotebookCheck.net (who does osliscope PWM tests) and got this response.

    Hi,
    we do PWM tests on phones and tablets, and use a Osram BPW-34 photo diode (100ns + 100ns according to the specs sheet for rise + fall). However, we only use a 50 MHz 1GBps Rigol osci, so that could limit our detection rate.
    Best Klaus
    __________________________________
    Dipl.-Ing. Dr. Klaus Hinum
    CTO Notebookcheck Publishing GmbH
     
  8. Jamon

    Jamon Scribbler - Standard Member

    Messages:
    260
    Likes Received:
    64
    Trophy Points:
    41
    Does 1/4000 shutter speed mean you could capture a dark frame from square-wave PWM up to 4 kHz with short duty cycle?

    If you really want to benchmark displays with a camera you need to do the more advanced test, where you make a black screen with white vertical line, then pan the camera horizontally at a constant speed while taking a picture. Then counting the number of strobe trails, and multiplying that number by the bottom number in the shutter speed. Also just looking at the picture to see if the ghost trails are smooth and connected by a gradient of light, or harsh and separated by distinct black bands. The one with strong stripes is more of a strobe effect than a variation in brightness, and more abrasive on the senses.

    My main interest is just to avoid ordering a tablet that looks great on paper, then being blasted by flashes, which has happened multiple times in the past. I used a tablet and desktop monitor for years on 100% brightness only to avoid that effect, which was too bright. I have physical access to an electronics store now, so I can test some tablets before purchasing. Previously I'd just order based on what I saw online, and the reviewers never mentioned the screen flickering, even when in their YouTube videos it was obviously showing bands cycling across. You'd think they'd notice while editing their review, and say something about it, but it wasn't on most people's radar. But also they tend to crank up the brightness to max for video reviews, so many videos don't even show it.

    Now however after enough forum posts like this one, there's a bit more awareness, and sometimes reviewers test it, and hardware designers seem to be responding with less obnoxious low-frequency PWM being used in some products.

    I tried capturing some PWM effect from my desktop monitor, which supposedly uses 17 kHz PWM. I used the line test, with various shutter speeds up to 1/2000. I tried recording video up to 1/12800. I tried moving the camera as fast as I could, and slowly. But I guess the shutter speed isn't fast enough, and my muscles aren't fast enough to move the camera quickly enough to spread the time factor out spatially so the slower shutter speed could capture something. Once it gets into such tiny micro-slices of time, the basic tests break down. But at that point it means it's fast enough where it shouldn't be such a problem.

    Maybe you'll still get eye strain or other subtle symptoms if it uses PWM at all. But that's part of the price to pay. If you type all day, it doesn't matter which keyboard you use, your hands could still feel cramped. If you write on a paper notebook for hours, your hand could be sore after. That's why we should try to use our tools for doing something that matters, so when our eyes hurt, and fingers are sore, at least we have something to show for it.

    "My lifetime of Wacom use gave me hand cancer; but look what I created!"
     
    WillAdams likes this.
  9. xader

    xader Pen Pal - Newbie

    Messages:
    18
    Likes Received:
    0
    Trophy Points:
    5
    Hi,
    The Samsung Galaxy Tab Pro S is just announced and I just read a review here in youtube



    Please watch 1:20
    It seems the screen is flickering, but i am not quite sure whether it is problem of camera or really pwm.
    Can anyone confirm it ?

    Thanks
     
  10. Jamon

    Jamon Scribbler - Standard Member

    Messages:
    260
    Likes Received:
    64
    Trophy Points:
    41
    That's from the overhead lights. The way you can distinguish between waves from the backlight and environmental lights is by looking outside the screen. If the screen is the only place with waves, then it's from the backlight. But if you look beyond that and see waves on the entire view of the camera, then it's from the external lights. The waves in this video get most severe when the tablet screen is acting like a mirror and reflecting the external lights directly where you can see them in the reflection.

    However, all the new Android AMOLED displays from Samsung lately have had PWM at all times, even at maximum brightness. Hopefully they increased the frequency, because before it was only 240 Hz. I'm hopeful they increased the frequency, but have my doubts. We should know for sure soon.
     
Loading...

Share This Page