Hacker News Reader

Comment by TaylorAlexander | Hacker News Reader

Agreed. It’s possible the improvement is due solely to a reduction in total light output (which I assume is a side effect of reducing blues) but something about it really seems to help. I still wish my monitor could go dimmer.

View on HN

It is an anecdote but I also much prefer redshifting all my displays. At this point, it is really uncomfortable to look at a blue screen on someone's phone. If the science isn't finding an effect, they are either not measuring the right thing or they are not measuring it with a large enough sample size.

I agree. https://github.com/jonls/redshift has been a gamechanger to me. Hint for those with desktops: redshift may fail to detect your location so you can use -l lat:long to set it.

Is that only because it’s bright and contrasts strongly with the darkness? A dim one without a color filter looks fine to me.

Well I always turn my monitors to the lowest brightness at night but when the program turns down the blue tones I assume it reduces the light output further. That may be the main effect that helps (brightness reduction) or it could be that reducing blue specifically helps. I’m not sure.

> If the science isn't finding an effect, they are either not measuring the right thing or they are not measuring it with a large enough sample size.

This is a very dangerous idea to have. Placebo is a very well known effect.

True, but poorly designed studies and non-replicated results are also very common. I think if a majority of people find some effect to be anecdotally true I’m their own lives (“using flux makes my eyes hurt less and I fall asleep more easily”) then you should need to meet a high standard of very rigorous science to prove them wrong.

Placebo is controlled for in these kinds of experiments by using a comparison group exposed to a carefully designed placebo.

To be very clear, you can test for the abscence of an effect. It is a little bit more challenging than finding a "true" effect, but it is feasible.

It is a dangerous, and common, misunderstanding to not distinguish between "no significant effect" and "statistically significant zero effect".

Reply on news.ycombinator.com