What about something like this: Two small amber LED lights are attached to a strap/soft clamp. ...They wouldn't replace something like a PBSF, but they'd make a nice addition to a front/rear blinky.
You could have a 200-300 degree light bar mounted from the seat post and you could clip mulitple lights to it. Amber LEDs appear to reduce output VERY fast with heat. They aren't that bright to start with. So we may be stuck with red-orange (used for casrs, too) as the red also has less output and is not as visible.
OK. Just back from my favorite school access drive. Used the trunk of the car to stage the PBSF's so we can know what humans rather than video cameras see. The cell phone camera caught more glint off the car trunk and 4-way flashers than PBSF so lacked what was needed to document and show here what we saw.
1. Time 7:45-8:00 PM EDST near the western edge of the time zone. Sunset 8:35 PM. Will check near 1:00 PM solar time 2:00 PM EDST but this is best done on the weekend when it isn't raining and it will be. So next weekend at earliest. The brighter sun but at a higher angle I expect to cancel or maybe even be more of a wash-out.
2. Can you see the SBSF in a daylight low angle light situation:
Another vote for the PlanetBike SuperFlash.
I remember, once a fellow CPFer caught up to me in his race bike and told me he had seen my SuperFlash blinking from 3-5 blocks away on a sunny winter morning.
Kathryn and I confrimed that we could see them at over 400 feet though two side by side blend as one. This is 4-7 seconds at highway closing speeds. We did not see how far away you can still see them as pin points of light are meaningless to a driver, so we saw no point literally and figuratively.
BUT!!! Seeing the light AND not ignoring it or not understanding that it might be or is a bicycle is NOT the same thing. A bike tail light must not only be visible but proclaim BIKE! Or is isn't a lot of help.
The size of the lights seen at 400 feet may be a vehicle quite far away 15-30 seconds or more. Instead of a fairly close bike. A driver is going to lose precious seconds particularly if the bike was not in clear sight but happened upon over a hill or around a corner. The driver need to understand very quickly. Wimpy flashes of light won't do that unless a driver is expecting bicycles or is used to seeing such lights.
So, one KEY aspect of this process is the driver expecting a bicycle.
Another key is the recognizing of such a light AS a bicycle.
There isn't much time for Whazat? at highway closing speeds.
Your situation will differ from mine.
1. I ride a lot of roads where the closing speed on hills into the wind with speeders will be in excess of 50 mph (80 kph), 88 feet per second upper end, most 66 feet per second or less. I need recognition by the unsuspecting driver at 400+ feet or shortly thereafter.
2. There are few cyclists here. The few who do ride in the day, don't use lights or even particularly bright clothing. So there isn't a driver ed program in bike lighting in the day.
3. Even fewer bikes on county and state routes.
4. Somewhat derived from the above: even attentive drivers are not expecting to see a bike here. They see the light, they will think it farther away than it is, if they reconize it as a bike they will also likely think it is farther away than it is.
5. The PBSF is VERY directional. a cone of maybe 30 degrees so it is useless for a car entering the road when you have passed and the angle is greater than that. (A left or right hook into your rear.)
6. The helmet SBSF was all but useless even on full. When placed near the flashing lights it helped the light size issue. IMHO a PBSF in daylight on a helmet is a waste of battery power and weight. You simply cannot control where you need to look riding so the light is not going to be anywhere near a good angle most if not all the time.
7. The side view of the light in the day, under this light condition, is 50 feet and no more. They may as well be off as far as any side visibility is concerned beyond 50 feet. They are pitifully weak from the side in daylight.
8. Two side by side, set up a disconnected out of perfect sync call-and-response flashing that is pretty annoying and attention getting .
9. The 1/4 watt LED's are visible as separate flashes at 120 feet, or 1.5 seconds highway speed, 3 seconds in town, slow hill climb situation. Pretty ineffective in the vertical orientation.
10. When the SPSFs are switch end to switch end, the 1/4 watts are more apparent and the whole flashing effect is like a police car without the blue lights, only more erratic and if you can believe it, even more irritating. The effective light size is also much bigger.
Recommendations:
1. We need a larger light that makes us look closer for areas where cyclists compete for attention, or where their presence is a novelty and drivers need more time to figure it out.
2. Two larger PBSF's with 1 watt main and 1/2 watt small LED's running two AA's might be just the ticket.
2. If you have two SBSF's figure out how to mount them sideways end to end with at least one replaceable so you can turn them on.
3. Aim them STRAIGHT back. At a distance vehicle size is irrelevant. You hope they've figured out you are a bike by 2.5 to 4 seconds out, and the lights aren't important to identify you anymore.
4. Find a wider brighter helmet light. Or a pivot mount. A Catseye L1100 may have the lack of directionality needed. But this has not been evaluated by me. OR find a way to mount the PBSF so it can freely swing to the vertical regardless of how you tilt your head from chin down to neck back or a major part of that movement.
Anyone have A Catseye L1100 and want to report how well it works from angles in slanting sun?
5. SBSF's aimed forward (as much as your leg interference permits or 45 degrees whichever comes first) should help cross traffic, but a good bright floody headlight will likely do as well or better. A spot on the helmet to look at them seems to work well in the day for me, so far. I think I stopped three motorists from starting a launch in the last three weeks.