[lm2s]

Would love to know what kind of noise and sound level was used for this test. Was it a continuous monotonous noise (such as white/pink/brown noise) at a fixed volume? Or was it some random noise that would "pop-up" randomly?

Does anyone know?

I would imagine that a noise that would randomly "pop-up" would be worse. But would be curious if that's not the case.

[cheeseomlit]

That's also what I was wondering- I sleep with a fan on for the white noise, otherwise I sleep very poorly if at all. I wonder if that sort of 'covers up' the spikes in noise you'd be hearing if the room were otherwise silent

[brandonb]

This was ambient noise in the room, as measured by an Apple Watch. So “random,” real world noises that pop up, as opposed to a controlled level of white noise.

[maerF0x0]

+1 also thought about this.

Something like the variability of noise (eg, maybe figure out the 25th percentile dbs across the night, and then count the spikes above that? or maybe count the number of times the slope goes above a certain value indicating sharp rises in volume that would disturb someone?)

I also would love to simply see the data based on the average of "N loudest moment(s)" during the sleep. eg: treat the dbs score for that night as the average of the N loudest moments over the night, and plot a series of graphs that show various values of N. (or make it 3d, but i've found many folks are not capable of reading those kinds of graphs)