Description of the 9500I’s sensitivity & filtering system:
As we all know the 9500I has some unbelievable filtering of false alerts, and some people find Highway mode to be Farley quiet as well, even though Highway mode has full sensitivity.
In "H" mode there is no filtering.
In "C" mode there is strong filtering of X and weak K band signals.
In "A" mode at 15mph and less there is strong filtering of X & K band, and at 15mph to 55 there is a brief delay (less than a second) to see if there are any matching/pairs radar, like in/out door falsies. At approximately 55 and above there is no filtering/sampling at all.
The end result would be under 15mph strong filtering of X & K band, 16-55mph approximately 1/10 second delay if a paired signal is found, and 55 and above no filtering at all.
In a nut shell "A" mode works just like the "A" mode in the 8500x50 from 16mph trough 55mph.
The one big addition that i beleve Escort did to inhance the above is the horn on the 9500I is designed to help filter out those useless off axis falses “
Detection range” verses “Sensitivity”. The 9500I has very good sensitivity but on signals passed 35 degrees the sensitivity starts to decrease(
detection range is not affected and that’s also why “H” mode is still quiet). The horn actually takes into account for “Cosine error/effect” any radar more than 40 degrees off axis will show less than 80% of your speed so this is considered a low threat. For example if you are going 70 in a 55 zone and the LEO’s radar is 40 degrees off axis, his radar will show only 56mph. This dramatically reduces the falses from automatic doors without sacrificing real threat
detection range.
I figured the angle of the walls on the horn at 15 degrees would equate to the off axis decrease. Because 15 degrees times 2 (each side at 15 degrees) plus add the antenna width so the total equals 35 degrees and that’s where the off axis starts to decrease (all horns seem to be similar).
For example:
Signals from motion sensors are, in most cases off axis from the side of the road, not sitting in the middle of the road facing you or very strong. So by just shaving a little off axis you maintain detection range with creating less falses.
Most falses will alert your RD for a couple hundred feet when you are near them because they are usually over 40 degrees shooting across the road. Now if theirs a LEO off axis down the road you will pick up his signal much, much farther. Even if he’s 40 degrees off axis he’s so far down the road when you receive his signal he's essentially only 10 degrees off axis. Because of this, detection range would be maintaned for around-the-curve or over-the-hill encounters.
A discussion of the M4 horn: K band door openers...Opinions please 
Just some notes and observations.
Radar from a gun can acquire your speed in most conditions at approximately a maximum of 3000ft in moving mode, and a maximum 5500ft in stationary mode, how much more sensitivity do you really need? An extra 500, 1000ft, 3000ft? It only takes a few seconds to slow down. Besides the fact that the LEO needs to be able to make out who is speeding and 3000ft would be difficult to see.
If I pick up a real threat with 10 seconds to react on a less "sensitive" RD, and 13 seconds to react with a more "sensitive" RD, does it really matter when it only takes 4 seconds to slow down?
Sensitivity doesn’t equate to not getting a ticket, the ability to communicate a real threat and the driver’s reaction to that threat equates to not getting a ticket. Its ½ driver ½ the RD responsibility to prevent a ticket. The most sensitive RD in the world won’t prevent a ticket if it doesn’t communicate the real threat clearly and the driver doesn’t react correctly.
Speed lock:
sensitivity/filtering diffrence? 
