Can you post the videos you took?
I'll try checking a few other local apple stores to see if the showroom M2's flicker or not.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Now that we have no PWM on MBA M2 what to do with FRC/Dithering?
- Thread starter jackblack123
- Start date
- Sort by reaction score
I'll try checking a few other local apple stores to see if the showroom M2's flicker or not.
The video I captured is a 50 second, 240 fps, 500MB slow-mo. I don't know if I can upload it here, and I don't have a youtube (or anything else) account.
I also don't know how to capture normal speed footage with the slow-mo cam mode (240fps), because otherwise the video is much too long. I'll try to think of a way...
Last edited:
If you have a Dropbox account or other cloud storage, you could put it there and have it generate a public link to share.
Last edited:
Here you go. Low lighting conditions, MBA M2 on the left, MBP M1pro 16" on the right.
Shot with iPhone 13 mini slow-motion cam, 240fps.
With the naked eye, I cannot see any flicker on either screen, but using the MBP at night gives me sore eyes and head-aches, whereas using the MBA is bliss.
In normal daylight, the MBP is still more eye-fatiguing than the MBA, but the difference is less noticeable than with low light.
This doesn't seem like 240 fps, unless your hands are moving much faster then normal human speed?
The transition at 0:09 from one screen to the next looks like it's happening at regular speed, not at 8x slow-mo.
Last edited:
As I said, the vid is captured with the 240fps slow-mo cam. It was originally almost 2 min long and over 500MB. Editing it with the iPhone Camera app, I could choose normal playback speed instead of slow-mo, so it became 28 seconds and exported to 40MB. In other words the same recording, but easier to view, store, upload.
Anyway I seem to miss your point. Do you not see the flicker on the MBP?
If you go to any Apple store and view any MBP 14" or 16" using your iPhone's slow-mo mode, you will find exactly the same flicker on all of them (below 6 brightness bars). Doing the same with several MBAs M2, I was unable to see flicker on any of them. You just need to view it through your iPhone, there is no need to record it. And if you record it, it doesn't matter at which speed you play it back, you can still see the same thing, only much slower in slow-mo.
Last edited:
So do you realize that after the export process your video is somehow playing at normal speed, as if it wasn't captured in slow-mo mode at all?
Like I said "The transition at 0:09 from one screen to the next looks like it's happening at regular speed, not at 8x slow-mo."
Can you recapture the video and put it on a dropbox/google cloud/something without the post-processing, so that the slow-mo footage is preserved?
Because otherwise it has no informational value.
I fail to see your point again. Yes, my whole video is at normal speed and it doesn't matter. The slow-mo cam can see the flicker and show it to our eyes, which it does. Can you see it on the MBP screen or not? Why do you want a slowed-down version of the same thing you can clearly see now?
Because it might also appear on the MBA M2 screen as well? Which is the whole point of this thread:
"Now that we have no PWM on MBA M2 what to do with FRC/Dithering?"
Are you somehow confused as to what this thread is about?
If you are, let me spell it out explicitly:
We don't need video proof of how the display performs at 30 fps. It's simply not necessary information as there are hundreds of videos online showing 30 fps footage.
Plus shutter speed timings on cameras can make any screen flicker, depending on the combination, so even for the MBP it's pointless anecdata.
Slow-mo footage is less likely to have shutter timings intefering and cause visual artifacts, hence why people reading this are not interested in the 30 fps footage posted.
Last edited:
I don't think the flicker seen on Mac LCD screens is due to PWM. That can be an issue on OLEDs, since it's typically in the 100's of Hz, which can be noticeable to some. But LCD PWM frequencies are far higher. For instance Notebookcheck lists the PWM frequency for the 14" & 16" M1 MacBook Pro LCD screens as 14,880 Hz, which shouldn't be detectable.
My guess is thus that it's instead due to FRC. From Wikipedia (https://en.wikipedia.org/wiki/Frame_rate_control#:~:text=FRC is a form of,(half frame rate) flicker ):
"FRC is a form of temporal dithering which cycles between different color shades with each new frame to simulate an intermediate shade. This can create a potentially noticeable 30 Hz (half frame rate) flicker. FRC tends to be most noticeable in darker tones, while dithering appears to make the individual pixels of the LCD visible. TFT panels available in 2020 often used FRC to display 30-bit deep color or HDR10 with 24-bit color panels."
Note: By "30 bit" and "24 bit" color, they mean what we are referring to as 10 bits and 8 bits, respectively, since they are multiplying the bit depth by 3 colors/pixel.
My guess is thus that it's instead due to FRC. From Wikipedia (https://en.wikipedia.org/wiki/Frame_rate_control#:~:text=FRC is a form of,(half frame rate) flicker ):
"FRC is a form of temporal dithering which cycles between different color shades with each new frame to simulate an intermediate shade. This can create a potentially noticeable 30 Hz (half frame rate) flicker. FRC tends to be most noticeable in darker tones, while dithering appears to make the individual pixels of the LCD visible. TFT panels available in 2020 often used FRC to display 30-bit deep color or HDR10 with 24-bit color panels."
Note: By "30 bit" and "24 bit" color, they mean what we are referring to as 10 bits and 8 bits, respectively, since they are multiplying the bit depth by 3 colors/pixel.
14880Hz / 240(fps) = 62. Does anybody from the 90s remember how much more visible the 60Hz CRTs flicker would be to the naked eye than the 80Hz or 100Hz CRTs?
And does the FRC theory imply that if we set the display to some lesser color profile (or maybe even to Low Power Mode) the flicker will disappear? Cause that does not seem to be the case in practice with the MBP screen.
Last edited:
Your division gives you the number of modulations per frame at 240 fps, and you're comparing it to the number of modulations per second. I don't see how that makes sense—it seems to be apples and oranges. More broadly, it's generally acknowledged that PWM frequencies in the 10 kHZ+ range are imperceptible to everyone. Have you found a study that says otherwise?
I guess you and I have different ideas as to what constitutes better representation of a flickering screen.
The video I posted is meant to show the experience you can have when looking at the MBP display through the 240 fps iPhone camera. That's it. Never mind recording anything at any speed of FPS, and never mind re-encoding a recording. If you can see those ripples waving across the display, that's what I see through the 240 fps camera. I cannot see them with a 30fps camera, I cannot see them with a 60fps camera.
This visual experience I speak of is captured in the video you saw, at normal playback speed. My aim was never to provide the best possible technically encoded proof for the flicker, it was to show you what I see with my eyes.
As for "shutter timings can make any screen flicker" I dispute that.
You cannot make a non-flickering screen (always on, 0 Hz flicker if you will allow the figure of speech) look like it's flickering, unless you are also making it seem like all the other always-on light sources around that screen (including the ambient daylight) are flickering at the same frequency.
This is clearly not the case here. Via the 240 fps camera I can see some light bulbs in my house flicker while others do not. And the Sun outside does not flicker on this camera either, just for the sake of being needlessly thorough. And none of the ambient light sources affect how the MBP screen flicker is perceived. That is always just as you see it in my "30fps" video.
And yet none of those 30 fps videos show the flicker waves that appear when using the 240 fps camera. And my "30 fps" video does? A bit ironic, no?
I guess I find your logic as flawed as you do mine, even if we set off to take on the same technical problem.
You know, I've been an active Electronics and Computer Engineer for almost 20 years, and while I'm sure there are many better Engineers than me, I still used to believe that I have a pretty decent understanding of technical things in this field. You've been making me question this assumption, which is kinda nice, albeit in a weird way... But at the end of the day, the more I read your posts, the less I think there's something valuable to address in your requests.
Last edited:
Your using a pseudonym, unless there's proof, like linking to a real world identity, other readers will not believe your claim.
Especially when as theorist9 pointed out, your doing math that doesn't make sense.
Plus the vast vast majority of 'active Electronics and Computer Engineers' aren't in any way involved with LCD display R&D, production, etc. So it's exceedingly unlikely, in any case, that you would know any better then several folks here who've done the research.
For other forum members to take this seriously, you need to show solid evidence. Otherwise, feel free to leave, threads on macrumors are entirely optional to participate in.
I don't feel the need to prove anything to you or anybody else, but thank you for proving my point. And for your kind invitation. I do feel free to leave, just as I feel free to stick around and just ignore your snark and continue the conversation with people who want to discuss this technical matter in a friendly and constructive manner. Be well.
What point?
The math doesn't make sense, the video doesn't make sense, the claims don't make sense.
Plus, ignoring theorist9, who as far as I can tell is a genuine and long-standing forum member, makes this seem like a pretty low-effort deflection.
So to be frank, I find it unlikely there is an 'active Electronics and Computer Engineers for almost 20 years' behind this account.
My reasoning is this: A 14.88Khz PWM frequency means the display brightness LEDs are switching ON and OFF 14880 times per second. Because of the nature of this modulation, those ON/OFF pulses will be longer or shorter (variable duty cycle). So depending on the brightness level selected by the user, the screen will be ON more or less time in a second, and it will be OFF the remainder of the period. The raport can deviate vastly from 50/50, and please read the last paragraph if you care about the effects of this.
Now, the 240 camera takes one picture and then shuts off for a short while 240 times per second. Each time it blinks it loses ~62/2=31 display flickers (assuming a 0.5 duty cycle; it could be more or less, I haven't studied cameras much).
But for the sake of simplicity let's asume it's 31 flickers. If that batch of flickers is not perfectly aligned (time-wise) to the camera shutter), then it's possible to (sometimes) catch the display mid-blink. Then on the next frame, it will catch it slightly sooner or later than mid-blink. Rinse and repeat, and so you will be able to visually notice artifacts across the display (such as the ripples in my video, and the videos of others). These artifacts of course are much more likely to happen at low display brightness, when the PWM duty cycle is way off 0.5. For the curious, please search 'macbook pro pwm' on youtube and simply watch the first few results.
But anyway the nastiest use case for the human eye is when we are selecting low brightness in the dark. That means short spikes of full screen brightness alternating with long (relatively speaking) periods of zero brightness. This prompts the iris/pupil to dilate (in darkness), only to have it promptly and repeatedly violated by a very bright surge of light. So while it's trying to do that thousands of times per second, even if the brain is consciously unaware of this effect, it's only natural for the eyes to feel a burning sensation (and many other adverse effects - some users report the effect of seeing "oily" and jiggling text) after a short time of using the screen, especially in a dark room. The only thing I find baffling about this whole story is how many people swear by the MBP displays and call them the best ever. I guess some may always use them at high brightness (or even low brightness but in an otherwise well lit room, which is not conducive to pupil dilation). Otherwise, whether aware of it or not, the eyes will suffer. If you need to read a study about it, go ahead. You will find them with either conclusion, but follow the money and see who is funding the one you want to believe. Or my personal advice: you could just kindly ask your own body if it's good or harmful. As a consequence of experience, I'm all for scientific studies and get a kick out of many, but have learned to take them with a grain of salt and let the truth percolate its way over longer periods of time and in any case abstain from making a god out of "science".
Do you realize the Macrumors forum software shows to every reader that your followup response to theorist9 was 18 minutes after I posted my comment?
It looks odd, especially after saying "I don't feel the need to prove anything to you or anybody else, but thank you for proving my point."
If you want to prove something, just say it clearly. If you want to not engage, then leave. These sorts of games really don't help the discussion.
Last edited:
ah never mind. Found the setting. First you have enable the daemon. but mine was already at millions. Does that mean my MBP doesn't need ResX?
Last edited:
This was a really interesting thread to read through. I'm not sure exactly why, but my eyes really don't like long periods with either my M1 Air or my work 16 Pro. Red/dry eyes, eye strain, the works.
Edit: I've been hit and miss with screens in the past, too. An old Intel Air (2011) used to annoy my eyes, but my iPads and LED iPhones haven't, and my trusty Sony monitor from 2004, that's still going strong, does not in the slightest.
Edit: I've been hit and miss with screens in the past, too. An old Intel Air (2011) used to annoy my eyes, but my iPads and LED iPhones haven't, and my trusty Sony monitor from 2004, that's still going strong, does not in the slightest.
Last edited:
Possible breakthrough!
This tool is supposed to disable GPU dithering on Apple Silicon Macs:
ledstrain.org
This tool is supposed to disable GPU dithering on Apple Silicon Macs:
I disabled dithering on Apple silicon + Introducing Stillcolor macOS M1/M2/M3
tl;dr disable temporal dithering on your M1/M2/M3 with Stillcolor. I got 16” M3 Max MBP about a month ago and it’s been absolute hell. By far the worst scre...
ledstrain.org
This is consistent with what I wrote above, in post #60--that the eyestrain seen in Mac LCD screens is due to FRC rather than PWM.Possible breakthrough!
This tool is supposed to disable GPU dithering on Apple Silicon Macs:
I disabled dithering on Apple silicon + Introducing Stillcolor macOS M1/M2/M3
tl;dr disable temporal dithering on your M1/M2/M3 with Stillcolor. I got 16” M3 Max MBP about a month ago and it’s been absolute hell. By far the worst scre...
FRC is a form of temporal dithering MacOS uses to increase the effective color depth of its 8-bit panels to 10 bits. The only native 10-bit panel used by Apple, and that thus doesn't employ FRC, is that on the Pro Display XDR.