I do describe what I am doing somewhere in the journals and the advanced search should be able to find where I first start talking about it. The theory on both ends, the light just before dawn and just after sunset, is that those lights are imitating what happens in nature as the sun hits the horizon. During that time, just for a few minutes, dramatic deep and far red shows up on the horizon.
In the morning, if the plants are standing tall enough to see that light at the horizon, something in their biology tells them that they are then tall enough, and have no need to stretch to see the sun that is quickly approaching. Without seeing this light they figure that something is in the way and that they need to get taller in order to be able to see the horizon. My lights hitting them from above doesn't seem to matter... they see the morning red light and consequently, stop stretching.
In the evening the red lights appear again on the natural horizon, and if the plants can see about 10 minutes of this as the light is slipping away, they know that nighttime is approaching, and they quickly get into the nighttime mode. Without this trigger light, the plants will stay in the daytime mode for much longer, some theorize that it takes 2 hours or so for plants not convinced by the red light that the daylight is gone for the rest of the day, and to then get going with nighttime activities.
So I have my lights set to come on for 10 minutes immediately before the lights come on at 6am, and for 10 minutes after the lights go off at 7pm, 13 hours later. The evening trigger light might be giving them as much as 2 hours extra nighttime mode, since they happily transition after seeing that powerful evening red light that we humans are just realizing the importance of.