Screenshots of the contraption used to test this phenomenon will be added at a later date.

Take a repeater output from the back of an observer, and set it to 3-4 ticks of delay. when the repeater powers a piston, the piston will stay open for longer than 1 tick, as if the repeater is acting as a pulse extender.

In the contraption I used to test the phenomenon, I wired the repeaters the observers were facing so that all trials of the experiment would activate simultaneously on the rising edge of the pulse. The outputs were varied from 0-5 ticks of delay, to cover most of the possible states I could think of. Using two repeaters extended the pulse the same amount as the four tick repeater, the pulses from the repeaters on 0, 1, and 2 ticks of delay were not extended at all. the repeater on 3 ticks extended the pulse but less than the 4 tick repeater.