I tried to measure frontlight power on an Oasis 2 at different brightness settings by measuring current from a charger to a fully charged Oasis 2. This did not work, partly because my USB multimeter only has a resolution of 10ma, but I suspect better resolution would not have helped because even maximum brightness did not give good results. Measuring accumulated mah over long intervals might work.
But it turns out that the linux device driver for the Voyage power supply exposes some useful information in /sys/class/power_supply/max77696-battery/, including current and voltage, now and averaged, which can be accessed on a jailbroken device. In addition, there is a directory /sys/class/backlight/max77696-bl/. (I suspect amazon inherited a device driver assuming a phone or PDA.) I made a table of /sys/class/backlight/max77696-bl/actual_brightness value versus Voyage brightness setting, which is plotted in the attached brightness.png for logarithmic and linear scales ( human brightness sensitivity is approximately logarithmic
https://en.wikipedia.org/wiki/Weber-Fechner_law ). 10**(0.2*x) is also plotted for comparison. (I'm using 0.2 just because it is such a close match for the first half of the plot.) There is every reason to believe that the increase in power used for each step is larger than the increase for the previous step. The questions are how much, and does it matter? If you see no problem with how long your battery charge lasts, or don't care how brightness affects battery life, then it does not matter to you. Others might care or just might be curious.
So far, I am in the curious category, but for the Oasis 2, I might also want to take power use into account. Anyway, I wrote RUNME.sh scripts to make battery discharge current measurements at each brightness level. The results are plotted in the attached led_ma* PNG files. I've also included the scripts and measurement data in LEDma.zip attached in the next post .
More later.