I assume you're talking about the singal to the left. This is a three headed signal, and the top unit is obscured by the beam of the station canopy. From the engineer's or fireman's point of view, it isn't.

>These signals are carefully aimed to provide the clearest most direct view to the engineer.

That's only the theory, in practise the situation is far, far different. For all three colors to be equally visible from any distance, the light sockets have to be placed behind the [fresnel] lenses so the lamp filaments are at the focal points, the lamps themsevles have to be manufactured so the light center length and axial alignment tolerance is within a 1/64" (this is always true for the 9S11/SC 12 V signal lamps used), all the lamp housings have to be collimated (lined up so their beams are parallel), and the signal head has to be aimed in the proper direction. Even the slightest deviations in any of these tolerances can cause wide swings in the resulting light beams.

In older colorlight signals which were three lamp housings in one casting, like what the Reading electric lines have/had, the light socket locations could be adjusted so the beams were collimated at the factory, and this would be preseved upon installation. All that was necessary in the field was to use the built-in aiming device to align the signal.

With these newer ones, each lamp housing is a separate unit. This allows for "flexible costs" of manufacture and captial investment whenever you need more or less than three indications. If you look carefully, the upper right signal lens on the lower unit has been blanked, meaning there are only two housings in the bottom unit.

For each color to be equally visible, all three housings must assembled so the light beams are collimated (parallel), and this is almost never the case. Even though the factory may properly locate the lamps at the focal point of each lens in each housing, there is no provision for precision alignment of the housings relative to each other, no dowel pins to recreate any precision alignment of the housings which may or may not have been done at the factory. So when the heads are assembled in the field to the desired configuration (in this case triangular rather than vertical), where they are also subject to rough handling during installation, the beams don't necessarily wind up being parallel (there is only one aming device, in only one of the housings). These optics of the housings are so directional, it doesn't take much to jar them out of alignment with each other.

These things are highly directional, making their appearance less visible to someone on the platform. There is also a "hot spot" in the center of each lens which directs a beam off at an angle for close-up viewing, whoich would be aimed at the locomotive cab, not the platform (there is a small arrow cast into the lens to indicate the hot spot's direction).

>The top one looks darker to the photographer than the bottom one because he/she is nowhere near cab height. Trust me, I've seen thousands of signals from the cab and they look completely different than they do from the ground or even just a few degrees off from cab position.

I've seen thousands of these myself. I've also had the opportunity to compare them to an equal number of Hs, H-2s, and H-5s, and these new ones don't measure up, not by a long shot (misaligned searchlights aren't unheard of--although with (I've seen headlights do the same thing)them, all three colors are misaligned in exactly the same direction). With these new ones, one color may stand out like a sore thumb for a mile or two in the daytime while another color on the same signal [head] is barely visible from 1/4 mile away (sometimes even in the dark).

With fresnel lenses, there is also the problem of reflections arising internally because of their prismatic structure. When sunlight shines directly on one at just the right angle, it can reflect off one or more of these internal surfaces, making that lens appear to be lit (I've seen headlights do the same thing). If the desired color is "dim" because of the alignment issues, then these internal relfections can cause false indications. Seen this phenomenon myself many times. Since so may have reported a clear signal at Chatsworth, I suspect this has more to do with the accident than most want to admit.