Bue,
I use two seperate chronies, that are mounted in line (one behind the other). If the arrow is flying straight through the chronies, the displayed speeds don't differ more than one fps, with the further away chrony showing the slower speed.
If the arrow doesn't fly straight, because of deflection, more than normal paradox etc. the two speeds are quite a bit different.
Here's why:
If the arrow is not passing the first sensor point on, the sensor is triggered by the shaft passing over the sensor in an angle. Since the arrow is straightening out on his flight, chances are, that the second sensor is triggered by the point or a section of the shaft, that is further up front. The speed reading will be faster, than the real speed of the arrow.
By the time, the arrow is passing the second chrony, which is further down the flight path, the arrow's flight is already more straightened out and the reading will be more acurate. So if the two readings are quite a bit different, the arrow didn't pass both chronies in a straight line.
The biggest problems occur, when I'm testing non-center shot bows, or bows, that are cut a bit before center. By using a mechanical release, there's almost no paradox. That means the arrow doesn't clear the bow like shooting with fingers. The arrow hits the sight window, will be deflected and will not pass the chronies in a straight line.
Hope this explains, what I'm doing.
Blacky