Originally Posted by TheEndHK
If you ask for true low light test then probably it is just testing long exposure time combined with slow shutter
Not necessarily. Larger pixels means you can potentially capture a better shot in low light with same shutter speed. At least that is the whole theory of the "Ultrapixel" camera. I personally don't see the "low light" test in the article as low light at all. Too much ambient light from the neon lights.
In any case, although the M8 shot is overexposed in the lighted window area, it captures a lot more detail in the darker areas, that shows up as completely dark on the other 2 devices. So its arguable either way which one is "better". In fact, the Butterfly 2 isn't capturing much at all in terms of "low light" its mostly just capturing the well-lit areas illuminated by huge neon lights. Hardly "general low light".