By the way, the camera banding test is not a very good indication of flicker. My phone camera (Google Pixel 3) will pick up even 2% flicker and make it look like we’re at a disco. Other flagship phones appear to have the same “issue.”
Here are some numbers from the original footage of my video.
First I exported one frame every second:
ffmpeg -i 00000.MTS -vf fps=1,scale=640:480 frames/%05d.png
Then I took the average brightness of each exported frame:
for x in frames/*.png; do convert $x -colorspace gray -format "%[fx:100*mean]\n" info: >> log.csv; done
This is the result:
There are some fluctuations. Not sure if they correlate with the mains voltage, too lazy to OCR the text from the multimeter. 
Oh, this doesn’t tell us if there’s a flaw in my setup. But random noise looks different.
edit: I manually compared the brightness and the mains voltage and both correlate with each other more or less. I think this is very similar to the result from maukka.
Thanks for that test. It’s very helpful.
If I’m understanding the graph, the camera is picking up minor brightness fluctuations of around 1% (from a trough of ~52.7% to a peak of ~53.7%). You also notice some slight visual changes in the output just by observing. Is that correct?
I don’t think the absolute values of the scale mean anything (although it is in percent, but percent of what?). The changes are very hard to notice and I think it depends on the situation and the observer if it is noticed at all.
IMHO the point is that the Badtime Bedtime Bulb has a behavior that is not found in other bulbs.
OK, thanks for the clarification. And it may well be a “badtime” bulb in this instance!
Oops. :person_facepalming:
So today i received a bulb from Yeutterg, there were 4 people at home, we all tried to see visible flicker in different scenario (direct vs indirect, dark room vs a little bit of light) all of us failed to see any fluctuation in intensity or flicker, i’ll do some further testing tomorrow.
Is there any kind of typical household load i can put on my electrical system that is susceptible to induce voltage fluctuation ?
I also must say that the light produced is superb for a bedside table or such in my opinion, very soothing and high CRI compared to my 3000k osram led bulbs.
Big AC, fridge, anything with a big honkin’ motor that draws a lot of current on startup (compressor motor kicks in, etc.).
Circular saws and other power-tools are good for that, too.
And vacuum cleaners. They will pull some juice!
For the last days I have used the Bedtime Bulb in my bedroom as the only light and have observed visible output fluctuations several times.
Thanks for the pointer to https://www.ies.org/fires/melanopic-green-the-other-side-of-blue/
This is why we use amber LEDs as household lighting in the evening. The amber floodlights available appear to mostly use Luxeon Rebel emitters:
And at critical locations around the house, we have this sort of thing for don’t-step-on-the-pets perambulation at night:
https://www.mrbeams.com/amber-stand-anywhere-light
Thank you for confirming the Bedtime Bulb has the fluctuations that maukka reported.
His review of the Remez SunLike E14 LED bulb (3000K, CRI95) reports:
I would buy the Remez, not the Bedtime
What sort of fluctuations ? Is it a rare short dip and back to stable or is it lasting alternating changes ?
The only detectable fluctuation we manage to see is a short dip in intensity when our AC compressor turned on when a lot of other appliances where on also
Yes, it appears as a dip whenever there is a jump in the mains voltage. A sudden jump of 1 V is enough to be noticeable. In both directions. The brightness an the mains voltage seem to be proportional.
I suspect when this happened in our little extreme electrical load experiment that the voltage change might have been much greater than 1v since i’ve been told even the plasma TV went dark for moment when i was looking at the bedroom bulb.
Thanks for the feedback so far. We have two people who have noticed any kind of brightness change during normal operation. Six people have not noticed any change. We are still continuing the investigation.
statistically
50% of people dont notice flicker
25% dont care if there is flicker
25% wont buy a product that has been tested by reputable people, who know what they are talking about, and have both calibrated instruments and trained eyes
out of every 8 people, only 2 care enough to find a light that does not flicker
I appreciate that you care a lot about the flicker issue. Trust me, I do as well (both visible and invisible flicker). That’s why we’re running this investigation and have delayed the launch of the product until we have more info.
However, that conclusion is not correct. I specifically asked all the testers to look for the issue. No instrumentation is required to do a check for the visible flicker we are talking about—you either notice it or you don’t. Seeing if it correlates to voltage changes is nice to know and helpful for when we do a redesign, but not necessary to verify the behavior.
Please take this in the kindest way that I would appreciate more patience as we carry out this test. You wouldn’t get this kind of attention and openness from pretty much any other manufacturer.
For the purpose of adding context, the local final transformer of my electricity provider is 20 meters from my house and my household is the only permanently living in the little hamlet, i wish i had the equipment to log my AC voltage fluctuations, i suspect it’s very clean.
Thanks a lot!