The unbearable fussiness of the smart home

At my first job, I worked for a company that built (among other things) X-10 powerline-control interfaces that would let your PC do the same thing: switch on lights/fans, etc.

As far as I could tell, the only people buying that stuff were hobbyists with a lot of disposable income. Part of my duties was customer support and I can't recall a single instance of anyone doing it for any reason that didn't boil down to "it's cool."

Not making judgements, but I really don't see what the big deal is with getting up to turn on a light, and I live in a pretty big house.

I measured a few in a dark room with the bulb in a desk lamp aimed straight down at a lux meter on the desk. The bulb was 41 cm from the meter.

For each bulb I measured at 1%, 2%, 3%, 4%, 5%, 10%, 15%, 20%, 25%, 50%, 75% and 100% brightness.

The bulbs I had were a TP-Link white bulb (TPL), a Merkury Innovations white bulb (MIW), a Hue white bulb (HW), a Merkury Innovations color bulb set to white (MIC), and a Hue color bulb which I tested set to cold white (HCW), daylight white (HDW), white (HPW), soft white (HSW), and warm white (HWM).

         TPL   MIW    HW   MIC   HCW  HDW  HPW   HSW  HWW
  1%,     14,  135,   64,  135,   10,   7,  10,    5,   3
  2%,     14,  149,   64,  162,   10,   7,  10,    6,   3
  3%,     26,  162,   65,  170,   11,   8,  11,    6,   4
  4%,     43,  180,   65,  182,   11,   8,  11,    7,   4
  5%,     58,  190,   66,  200,   13,   9,  13,    7,   5
  10%,   128,  246,   77,  255,   20,  15,  20,   12,   8
  15%,   196,  303,   91,  322,   34,  25,  34,   21,  13
  20%,   261,  364,  112,  388,   54,  40,  54,   34,  21
  25%,   325,  428,  143,  451,   74,  67,  74,   46,  29
  50%,   630,  725,  400,  776,  282, 218, 282,  184, 118
  75%,   909, 1008,  820, 1099,  650, 483, 650,  410, 263
  100%, 1160, 1283, 1382, 1408, 1146, 850, 1146, 725, 463

For the smart bulbs the TP-Link and the MI bulbs where the most linear. The MI bulbs fit a linear function with R^2 of 0.9997 or better, but that function would be 130 at 0 so they aren't very good if you want really low light (like a nightlight for when you get up to in the middle of the night to go to the bathroom).

The TP-Link is almost as linear (R^2 = 0.9977) and its function would be 7 at 0, so it is great for nightlight use.

The Hue bulbs are quite non-linear. The color ones get closer to 0 than the TP-Link so make great nightlights. The white one doesn't get close to 0 so is disappointing as a low light bulb.

The curve for the Hue lights fits a quadratic polynomial well. For example the Hue color set to plain white matches 1168.3 b^2 -31.431 b + 10.89 (where b is the brightness from 0 to 1) with R^2 = 0.9999. Same for that bulb on the other white settings I tested. For example set to warm white it matched 464.82 b^2 - 3.2729 b + 3.1916 with R^2 = 1.

This raises a question. I believe human perception of brightness is not linear. Is it actually better from a perception point of view to have a dimming curve that is linear like the TP-Link bulb, or a quadratic curve like the Hue bulb?

When considering smart bulbs, check to see what power-on settings they have. The Hue bulbs can be set to power on off, power on to whatever state they were in when they lost power, or power on to any specific color and brightness.

The MI bulbs power on at full brightness. This can be annoying if you have one in the room in which you sleep and you get a short power loss overnight that normally would not wake you.

If you don't mind that, and not being able to get anywhere near as dim is the others, they can be a good deal. They are 3 for $15.88 at Walmart for white. A Hue bulb would call $15.88 a down payment. (MI color is 2 for $15.88 or 4 for $19.88).

I don't remember what the TP-Link does on power on.