Speaking as an FCC-licensed radio engineer, I can confirm Darkker's point about the directionality of the antenna. At the extremes, an isotropic radiation source (effectively a point source radiating equally in a sphere around the point) sees its signal strength in volts per meter drop off as the inverse of the square of any change in distances from the source, while a MASER (microwave version of a laser) will theoretically keep the same energy level through space indefinitely (though, in practicality, there is a limit, as with LASER light, to how perfectly the beam can be columnated). Every other kind of antenna system falls somewhere between those extremes, with the cell phone being closer to the isotropic source in nature and the radar beam being closer to the MASER.
Back to the original question:
First, to check your chronograph, Dr. Ken Oehler has pointed out the base qualifying test for his 35P chronograph is to successfully read the velocity of a BB or an airgun pellet fired over it. The size of the shadow is small and the speed is slow, allowing maximum opportunity for noise-related errors to appear. If you can't read a BB you can't be sure how a .22 will go, for example.
Dr. Oehler has also mentioned in the past that if you fire
RIFLE MATCH .22 LR rimfire ammunition through a
RIFLE barrel, then you can expect the average velocity at 15 feet to be within 50 fps of the velocity number printed on the ammunition box or otherwise published by the manufacturer. The reason this works is .22 LR rifle ammunition uses powders that ignite easily and burn out completely somewhere between about 16 and 19 inches of bullet travel down the barrel (see Geoffrey Kolbe on this) and aftward is not making any more gas. As long as the barrel exceeds the powder burn length, you don't have the variable in exact percent of powder burned by the time the bullet exits that you do in most centerfire loads. Additionally, the expansion ratio of a .22 LR in a rifle is huge and that tends to neutralize the effect on bullet velocity that variations in throat condition and exact chamber dimensions have on final velocity. As a result, unlike centerfire guns, you tend to get about the same velocity in your rifle that a tight specification SAAMI P&V barrel (24" long) gets from the same ammunition. The bore friction is low enough that you don't see a lot of velocity loss from the bullet coasting down the rest of the tube past the burnout point.
Incidentally, chronograph clock speed, by itself, is not what determines the instrument's measuring resolution. One of my pet peeves about the Chrony has been their accuracy claim just considers their clock period. In fact, clock speed times the distance between the start and the stop screens is what determines the resolution limit, and then it will be accurate only if the combination of the detection window of the sky screens and the precision of the spacing between those screens is as narrow or narrower than the distance the bullet travels in one clock cycle. The Oehler, at 4 MHz and using four-foot spacing will have the same intrinsic resolution as a 16 MHz clock with one-foot spacing, assuming the light sensor window and spacing precision requirements are met. One of the features of the Oehler is the cylindrical lens system over its detectors keeps the detection window narrow. The precision of the screen spacing determined by indentations in EMT pipe is a limiting factor, but it is at least as good or better than integral photo transistors stood off by their leads on a circuit board are.
If it turns out the chronograph is OK, other possible culprits are:
Bullet unseating by the primer. 45 Auto not infrequently sees bullets unseated by primers before the powder burn gets well underway, and this leads to variation in barrel time and final velocity. It gets worse as the powder gets slower. Try working the load back up with a magnum primer.
Inconsistent slow powder ignition. Again, try using a magnum primer.
Inconsistent bullet pull. Brass mixed either as to headstamp or load history can vary in how hard it holds onto the bullet and thus causes start pressure variation. Try buying a sample of new bulk brass so it all comes from the same lot and has the same number of reloadings (none). I find Starline as consistent as it gets in .45 Auto. Also, they make a special
extra thick wall version for +P loads. You can buy direct from the factory for 500 pieces or more, but can get just 100 at a time from Midway (for a higher price, but worth it if you just want to try it) when they have it in stock, which
they do at the moment. That's plenty for a test to see if your brass is part of the problem.