MQTT brightness with garage closed

When the garage is closed the brightness reported is always 29 or 30%. If I cover the reflector (opening the garage without actually opening it) the brightness immediately drops to ~3%. When I actually open the garage door it displays >95%

Is there a reason it doesn’t report the correct value when the door is closed? Am I doing something wrong in the setup?

Is there a lamp on your opener? It is possible that the light from the lamp is reflected into Garadget’s photo sensor. What is the brightness reading when the lamp is off?

There is a lamp on the opener, however this happens when the lamp is off or on.
Recently I changed the MQTT topic to match the device id and the brightness now changes sometimes but it is still not matching the actual brightness of the garage.
Here is an image of the brightness level in a 24h period without the garage door opening at all. The garage light may have been on a few times but there is no spikes in the brightness above ~35% (the lights are bright).
I have also checked the brightness with MQTT Explorer and the results mirror the Home Assistant image.
brightness

Here is another 24 hour period without opening the door
image

Here is what happened when I opened and then closed the door. You can see when I opened the door, it updated to be the correct brightness before updating to bright as the door fully opened. Then when I closed the door it got lower until it fully closed when it changed to ~30% again
image

I’m trying to understand if the issue is with the measurement or reporting. Does the ambient light from the mobile app match the values reported over MQTT?

Yes the app measurement matches the MQTT measurement. I guess adding MQTT has added to the confusion here.

From your initial message it sounds like the photo sensor is picking some reflection from the retro-reflective tag on the door. For this to happen the source has to be aligned in front/behind/next to the sensor, that’s why I asked about the lamp on the opener. If there is no other sources of light, the only thing I can think of is some residual light from the laser.

To test this theory, instead of covering the reflector, please try covering the laser in the way that the light from it doesn’t reach the reflective tag or leaks into the photo sensor.

BTW: This issue is device/situation specific: I don’t observe it in other units.