It all depends on the zigbee execution of it all. Zll has the oppertunity to group message the bulbs so that they go on at the same time. Hue bridge supports that via the api. But HA has to send that too for it to work. And since a HA light group can be of a number of different bulbs from different sources with different protocols. Well I think we get the point.
Most zigbee integrations can create a zigbee group. So if you don’t have to, use a zigbee group instead of a HA light group.
2 Likes
I have innr spots GU10, I grouped them in deconz, no visible delay on my end.
I do not use any cloud integrations to control them except from nabucasa gateway.
Same goes for any other of my bulbs, lightstrips, sockets and motion sensors. Using a mix of Hue, Innr, Aqara, Ikea and Osram
For the sake of the discussion, with both native Milight bulbs and remotes, a number of 50/100/hell, an infinite number of bulbs (as long as they are in 5-8 m radius of the remote) can be turned on/off, changed color, brightness, saturation instantly (less than 1/60 of a second or what an observable delay would be considered as appropriate).
However, current Limitless integration in HA (which uses a Milight gateway connected over wifi) gets broken with every HA release.
With an alternate HA integration over MQTT (https://github.com/sidoh/esp8266_milight_hub) it can get to similar responsiveness to a physical remote, however only about 95-98% of the bulbs would get the payload. Add another similar gateway in the mix and the success rate would go up to 99.99% (regardless of the number of bulbs to be controlled).
Some issues with Milight:
- communication is not encrypted (neighbors’ kids can fiddle with your lights as long as they’re less than 10 m away of the bulbs);
- sometimes bulbs are hopping channels and they need a command repeat to change as they don’t react immediately (not an issue with a physical native remote, though).
I won’t keep belaboring this thread because it really was just something me laugh, not an attempt to crap all over ezlo, but ezlo did finally publish their test environment, and it’s a farce.
No smart bulbs involved, and it’s a random mix of smart switches behind the scenes, but not the same set of switches on both sides. Most notably, the smartthings side has at least 3 dimmers in the mix while the ezlo side has 0, and the SmartThings side seems to have newer equipment in at least a couple cases (which means less likelihood of a local integration).
1 Like
Now if they were Tuya devices I would be impressed right now.
They’re cowards, try ezro vs Tasmota 