Surprise! Smart Bulbs Dim Differently

“Alexa, change the brightness of garden lights to 20%”

Voice commands like that are something I use all the time to operate the smart lighting at home. It’s one of the most useful features of smart lighting — with a single voice command or a touch on an app, you can adjust the brightness of multiple lights to a precise level you like on a 100-point scale, with a nice smooth transition, rather than a sudden jump.

However, if you take a closer look at how smart lights really dim, you’ll see that the 100-point scales used by the bulbs differ from one another. In other cases, you can dial the brightness down in a linear way — for example, you can crank down 10 percent of the total lumen output at the 10 percent setting and 60 percent of the whole lumen output at the 60 percent level. Set the bulb to 50% of its maximum brightness to obtain half of the maximum lumens. This seems sensible, doesn’t it?

Other lights, on the other hand, employ a logarithmic method, with the lumens rapidly dropping to lower-than-expected levels when the brightness is reduced. If you ask Alexa to dim a light like that to 50 percent, you could only receive 25 percent of the lumens you’d get if you left it on at full brightness. What’s going on?

Please bear with me while I explain the solution, which is obviously a little weird, but it’s intriguing and boils down to the difference between measured brightness and perceived brightness, as seen in the graph below. Final thoughts: The scientific evidence for how we see supports the logic of logarithmic dimming — but it isn’t a perfect solution.

This TP-Link Kasa LED dims the brightness in a straight line, while the Philips Hue White LED has a different dimming curve.

When I wanted to get a better look at the dimming curves of these bulbs, I put each one into a lighting lab’s integrating sphere and used the built-in spectrometer to measure the brightness at each of the 100 dimmable settings. With several hundred readings to go through, this was a time-consuming operation, but it was worthwhile since dimming an LED reduces the amount of energy it consumes and heat it emits. As a result, the bulb becomes a little brighter. The TP-Link graph on the left illustrates the impact somewhat; it takes a linear approach to dimming, but when I dimmed it down to take measurements, the values began to drift away from the dotted line representing the objective.

To the naked eye, there isn’t a discernible difference, but when it comes to drawing out precise dimming curves in Excel, it may be a real nuisance. Despite my best efforts to make the measurements as precise as possible for the sake of those graphs, I recommend allowing for a 10 percent margin of error on the exact lumen counts, especially in the centre of each dimming curve, to be conservative.

Follow us?for future updates, like when we launch our closed pilot test.

Twitter:?@SmartCityOSOps?| Instagram:?@smartcityos?| TikTok:?@smartcityos