Always on display (sleep mode) behaviour and rules on newer AMOLED watches

I'm wondering about Always on display (sleep mode) behaviour and rules on newer AMOLED watches (Epix Gen 2, Venu 2, and Marq Gen 2 etc) compared to the older Venu and Venu Square devices. Is the display technology or behaviour or rules different? Someone showed me a photo of their always on display on a Marq Gen 2 for an analog watchface, and it looks almost the same brightness as high power mode, with adjacent pixels on. (not using a checkerboard mask). I said that I suspect that it violates Garmin AMOLED burn-in rules and it would likely turn off after a while. But they say that the display doesn't turn off.

Do the rules of no more than 10% of pixels on and no pixel on for more than 1 minute still apply for these watches? Or can the user override the burn-in protection (display shutdown) functionality somehow using settings? Is the display technology different such that the colors are more vivid in sleep mode?

I don't have one of these watches to investigate this myself.

Top Replies

myaro over 2 years ago +1 suggested

I've gathered there are two separate issues at play, which are often thought of as the same thing.

The first is whether you've "lit" up more than 10% of the pixels when in "sleep" mode.…
myaro over 2 years ago in reply to myaro +1

Here are some images to demonstrate what I was describing. Not the difference between high power / low power (asleep) mode. Also notice the different indicators on the simulator for "Screen protection…
FlowState over 2 years ago in reply to myaro +1

myaro said:
To prevent burn-in, I just have a small, variable offset for where I draw the center of the watch face on the screen. (Leaving enough spacing to the edge so that nothing gets clipped).

Interestingly…

All Replies

0 Shent-Aurkhan (SHN) over 2 years ago

hi, pretty sur they send you a screenshot of native watchface,

for CIQ, we don't have choice, we have to respect the 10% (and I think for newest you enum not the 3min rule anymore).

As for many thing, native wf work diferently than CIQ ones.

0 trudelta over 2 years ago in reply to Shent-Aurkhan (SHN)

Interesting. This is the first I have heard of it. Can we get someone from Garmin to confirm.

Can you explain ->

I think for newest you enum not the 3min rule anymore).

0 Shent-Aurkhan (SHN) over 2 years ago in reply to trudelta

first time you've heard about that? I 've seen it here (forum) many times.

on "old" amoled devices, 1 pixel can't be turned on more than 3 min, it is not the same case for the devices you enum

developer.garmin.com/.../

0 myaro over 2 years ago

I've gathered there are two separate issues at play, which are often thought of as the same thing.

The first is whether you've "lit" up more than 10% of the pixels when in "sleep" mode. If so, your watch face will just go black. This is separate from burn-in. This is a power / battery limit.

The second is the "burn in" concept. That's something you can test in the simulator, but I haven't found any consequences (enforced by Garmin) when that suggests burn in is possible.

For sleep mode, I just minimize the number of pixels needed.

To prevent burn-in, I just have a small, variable offset for where I draw the center of the watch face on the screen. (Leaving enough spacing to the edge so that nothing gets clipped).

0 myaro over 2 years ago in reply to myaro

Here are some images to demonstrate what I was describing. Not the difference between high power / low power (asleep) mode. Also notice the different indicators on the simulator for "Screen protection" and "Burn in state".

If you click start, it will simulate 24 hours and then report if it detected burn-in.

One other strategy for burn-in seems to be to simply have the analog hands of the watch spin the entire circle and therefore toggle each pixel off at least once per hour.

0 FlowState over 2 years ago in reply to myaro

myaro said:
To prevent burn-in, I just have a small, variable offset for where I draw the center of the watch face on the screen. (Leaving enough spacing to the edge so that nothing gets clipped).

Interestingly, it looks like Garmin pixel shifts datafields for native activities, too. (A user filed a bug report when his data field seemed misaligned on an AMOLED device, and after taking a few screenshots, it turned out the boundaries/contents of data fields would shift over time.)

0 trudelta over 2 years ago in reply to myaro

Interesting. If you have any graphic element 2 pixels or more wide or tall (like your icons and hands), then a 1 pixel shift in 2 directions is not enough to avoid breaking the 3 minute rule.

0 trudelta over 2 years ago in reply to myaro

I think your hands and icons in the picture would violate the 3 minute rule. A 1 pixel shift up/down/left/right is not enough to prevent many pixels from being on more than 3 minutes.

0 Alex over 1 year ago in reply to Shent-Aurkhan (SHN)

Hi!
How did you find that is different?
Seems like : " On System 4 and above devices, Connect IQ watch faces can have a launchable configuration flow built into the watch face"
But I didn't find anything regarding rendering difference on OLED displays

0 jim_m_58 over 1 year ago in reply to Alex

Look in the section of the link about AMOLED as well as best practices.

Also see this and the section about AMOLED watch faces.

https://forums.garmin.com/developer/connect-iq/b/news-announcements/posts/connect-iq-3-1-now-available