Zakim Bridge, Boston in UV at night

[Andy Perrin]

I took this photo last night, truly pushing my equipment to the limits. The Zakim bridge was the crowning achievement of Boston's Big Dig a few years ago. It's typically lit with blue lights, as it was last night, although sometimes the colors change. On a whim I decided to see what could be seen in the UV, since I had my camera and tripod with me from another long walk yesterday, and North Station is adjacent to the bridge.

 

The resulting image is very noisy. The settings were ISO16000, F/3.5 on the Novoflex Noflexar, 330WB80 filter rear-mounted, and a 30 second exposure. I have to figure out how to get longer exposures out of that camera!

 

 

With a lower ISO of 6400 and F/8, and a swooping angle:

[Cadmium]

Andy, Cool! Very interesting!

Side note: You know these new brighter bluer headlights they have now? I have often wondered if those have a lot more UV that the older lights.

[Andy Perrin]

Not sure about the comparison to older lights (although probably, with the xenon in the new ones, right?) but in the bottom picture you can see what a difference there is between the headlights and the (red filtered) rear lights on the two sides of the highway.

[OlDoinyo]

Some variations in apparent color may be due to variations in vehicle speed. Another complication is that xenon lights are being displaced by LED lights which emit almost no UV. The xenon HID arc lamps would probably be the most UV-bright, but many halogen incandescents put out a surprising amount as well.

 

Taillights are a different conundrum. Most taillight plastic looks utterly opaque in daytime UV photos. I have had trouble believing that a recordable amount of UV could leak through such a thick red medium, even when it apparently cannot pass through a common yellow filter. Perhaps I am wrong on this; I, too, have photographed taillights in UV exposures. What alternative explanation could there be? At night we are pushing the margins of UV, and out-of-band leakage is harder to contain. Red/infrared leakage is one thing to consider, but I am not sure the color I see here is consistent with what I would expect. Blue leakage is another possibility; also be cognizant that license-plate lights are white. Are we seeing taillights at all in some cases? Perhaps some experiments are in order here. I would expect that taillights that use red LEDs would be UV-dark; that has been my experience with other LED sources, such as the electronic billboards in Times Square. The sign-illuminating lights above the roadway in your picture are also dark--that may indicate LEDs in use there.

[Andy Perrin]

It could be out of band leakage, I guess, since there is nothing to compete with it (as there is in daylight) and the exposure time is really long. The white balance was done early in the day off asphalt, and the UV colors were the traditional ones back in the daytime, so I don't think it's a WB issue. It's also possible that I'm getting indirect headlights reflecting off the ground/car in front? Seems pretty bright for that, though, even with the long exposure.

[Andy Perrin]

Here is another photo. There IS some reflected light, as you can see from the fact that the road sign is visible from reflection only.

[OlDoinyo]

At least one of the traces in the last frame shows a turn signal prior to a lane change, so that is definitely either a taillight or a marker light, not a headlight reflection.

 

The dim illumination on the sign looks too uniform to be the result of base lamps. I think it is throwing back some residual light from headlights.

[Cadmium]

You have me wondering.

So the cars on the right are moving away from us, and the cars on the left toward us? So this makes me wonder why the lights are basically all the same color, other than a few brighter whiter streaks on the left side.

At least two of the pics have some 'blinker action', and curiously all the blinkers look slightly brighter, even maybe whiter than the other bluer lights.

The first pic has a whiter/brighter streak on the left that is interrupted, maybe when passing by another car?

The overhead lights and street lights are all whiter, making sense, because of the wider spectrum of such lights as metal halide, etc..

There is some basic illumination from those halides on the pavement and surfaces.

But why the approaching and departing lights are seemingly the same color, other than a few brighter approaching lights, I am a little puzzled by.

This has me really wanting to try this now! I may go take a little drive and look at lights for a while...

[Alex H]

I have to figure out how to get longer exposures out of that camera!

 

You can stack exposures in the same way as astrophotographers do. And if you add dark frame, you can get rid of some of the noise too.

[Andrea B.]

Go to bulb mode for exposures longer than 30 seconds. Using a remote (either manual or electronic) helps. Some remotes permit you to set the exposure length so that you don't have to use your watch.

 

You didn't say what your white balance setting was. It is possible you are recording violet/blue visible light. I don't have that filter for testing for Visible light leakage. I don't think it has any. But then I'm not exactly sure whether light in the range 380-400nm records records as violet or violet blue and whether that filter may have a long right-hand tail to hit that range. Probably so given the "80".

 

Anyway, cool experiment!

[Andy Perrin]

The filter doesn't SEEM to have any visible leakage under normal use, although I have not done a careful experiment to estimate how much at these bizarre settings. No filter is perfect, right? Here is the manufacturer's spectrum (red curve):

 

I don't know quite how strong the visible blocking is, but certainly it seems to go to 0 on the normal scale well before 400. Up till now, the filter has behaved normally and given the standard UV colors. Here are dandelions:

--

 

I don't have a remote, so I guess I need one to use the "bulb" setting?

[Andrea B.]

Agreed that there's probably no vis leak. Twas just a thought.

 

To use Bulb without a locking remote, you just have to hold the shutter down manually for the length of the exposure. It can get tedious and may induce some camera shake when your finger/hand gets tired as you try to hold longer than 30 secs. "-) I'm pretty bad at this myself. "-)

[Andy Perrin]

Haha, I just bought a remote for the grand total of $7.95. The only trouble is that the IR port for the remote seems to be on the FRONT of the darn camera! Great for selfies, not so great for triggering the camera when you are standing behind it. I'm hoping I can make a reflector or something that will let me trigger it from in back.

[Andrea B.]

oh la!! I have one of those too.

 

BTW, I've recently been using a Vello Free Wave remote on the new Sony a7R conversion. Receiver fits in the hot shoe with a cable to the port. Or you can tack the receiver onto camera elsewhere with velcro dots. The small trigger for remote use has single, multiple, bulb and timed settings. Setup runs on 4 AAA batteries so I'm using rechargeables. Good value for the money. Several other options exist of course.