Digital cameras with manual shutter lock?

Can anyone (here in Photography) recommend a good quality digital camera that has a manual shutter lock? Or even one where you can set the exposure time to minutes or longer? I'm putting together an Instructable that involves single-photon counting (few Hz rate), so I need a device that can accumulate hits for the order of an hour to get a recognizable image.

sort by: active | newest | oldest
1-10 of 13Next »
gmoon9 years ago
See killerjackalope's comment. The noise digital cameras produce is proportional to the exposure time. This is very much a break from the "film" analog. There are ways to get around it. I have two pro digital SLRs (D1X, D2X), and the newer camera takes a second "black" exposure of equal duration, and uses the noise in the second frame to cancel the noise in the first. This could be done with external software, however. For some weird reason, most DSLRs don't have a cable release anymore. An external release module is usually required...
kelseymh (author)  gmoon9 years ago
>> The noise digital cameras produce is proportional to the exposure time. This is very much a break from the "film" analog. Yes, of course (speaking as a particle physicist) :-) The thermal noise is a random Poisson process, where e-hole pairs are produced spontaneously. The longer you integrate, the more noise hits per pixel you end up with. >> takes a second "black" exposure of equal duration, and uses the noise in the second frame to cancel the noise in the first. That could be a problem for my application, where the actual pattern being built up looks very much like noise until you've integrated for many minutes. I suspect that doing an active subtraction would ruin the result. This is definitely turning into an interesting physics/engineering problem by itself!
gmoon kelseymh9 years ago
It's not only thermal noise--it's the noise inherent in the electronic circuits, too. Each has a "noise threshold," and trying to capture near those thresholds is problematic (as you know.)

It's a little like trying to capture the squeak of a mouse behind the walls, while the refrigerator is running...(and the mic / amplifer itself is noisy at hi-gain settings.)

I'm not certain why subtracting a "black frame" wouldn't work, though. In theory, any external source of luminance that was strong enough to trigger the CCD would leave a discernible image after subtracting the "black frame," regardless of it's pattern. I believe the technique was pioneered by astronomers for very low-light imaging.

NASA's been doing similar things with images since Mariner in the 60's--subtracting noise masks to remove image artifacts (analog processing in the beginning, perhaps?)

I certainly see why an instructable that requires an expensive camera is of only limited use....

As an aside, the "digital analog" for the film process would be an image sensor that's always on, and has been capturing the background noise (heat, electrical, cosmic rays, etc.) from the day of its manufacture until it's developed.

Film, while still susceptible, has a phenomenally low noise threshold. Of course, it's a one-shot deal. ;-)

Incidentally, I've worked in quite a few darkrooms, and they all had light leaks that were obvious after 30 or 40 minutes. Duct-tape might be in order...
kelseymh (author)  gmoon9 years ago
You wrote: >> I've worked in quite a few darkrooms, and they all had light leaks that were obvious after 30 or 40 minutes. Duct-tape might be in order... I was thinking gaffer's tape. However, it is becoming really clear to me that trying to seal a whole room (even a closet) is overkill, and probably doomed to failure :-/ I want to think about designing (or finding an I'ble) for a good light-box with a tight-gasketed insert for the laser, and a viewport for the user or the camera. However, that discussion should be moved to a separate topic (if I don't find an I'ble or off-site equivalent).
kelseymh (author)  gmoon9 years ago
You wrote:
>> It's not only thermal noise--it's the noise inherent in the electronic circuits, too. Each has a "noise threshold," and trying to capture near those thresholds is problematic (as you know.)

Sorry, my fault. All of that stuff is what I tend to call "thermal noise." Any kind of
circuitry that you're dealing with here is so far away from the quantum limit that its irrelevant. The electronic noise comes from fluctuations in the
electron/hole energies, scattering off phonons in the lattices, etc. All those things tend to scale the same way with temperature (kT).

For most electronics, the cross-section for cosmic ray interactions is tiny (there are a few hundred cosmics per square meter per second, and a typical surface mount cap is a couple of square millimeters :-)

I definitely need to think more about black-frame subtraction. I have no professional experience with modern astronomical or astrophysics detectors, but I can talk to my GLAST colleagues downstairs and learn something :-)
There's a large number of SLR's that have an indefinite shutter option, you just click it closed afterwards. Would film be an option to you? It's a very simple process to lock a film camera's shutter open, alternatively if it's in an area of size small enough you could use a large direct exposure to film and have no camera...
kelseymh (author)  killerjackalope9 years ago
Cool. I had no idea (having no experience with real photography, just taking pictures). And now you've given me the right keywords to search ("indefinite shutter"). Thanks very much for the help! Film is an option for the I'ble, but it's less desirable -- it would require the user to do developing, and they wouldn't get to see the quantum interference effects build up "live." In fact, if I were going to use film, I would use direct exposure (the experiment has to be done in a darkroom or sealed lightbox anyway).
Hmm... The other option I can think of would be a video with all the frames superimposed upon themselves, letting you watch it live or look at the total result, the process of making the image from the video would be like rotoscoping minus the tricky bits...
kelseymh (author)  killerjackalope9 years ago
Yeah, I was thinking about that, too...however, with a 1-few Hz photon rate, most of the frames are going to be dead empty (or thermal noise). Unless the video has really intelligent realtime compression, it's going to be a horrible waste of bandwidth :-) The more I think about this, the more I realize that the detector (camera) is a fairly complicated project all by itself, let alone the cool quantum mechanics application.
Well the advantage of the video is that it can be replayed and would be easy to do, the camera one holds it's own problems, the longer the shutter speed the more a CCD is likely to encounter noise, your best bet for viewing them may be some for of box attached directly to the lens otherwise it's likely to be noisy. The other issue is see is that forcing the camera to record an image for an hour could be too long to make it work - it'd average out to absolutely nothing, maybe I've experimented with similar photography processes so I'll test that with various things to see what can be done with it on those terms. the obvious way to remove thermal noise would be to keep the temperature of the camera and sensor not much higher than the bottom functioning levels. Another option may be to use two cameras choreographed to use 30second or longer exposures alternatively so you get a set of 120 clear frames in the whole hour which can easily be overlaid to show a complete pattern and print all of them and make a tangible flipbook for confused board members... Your other issue is the camera's power source, recording that data for so long will murder the battery so look for one with a power in bit, so you can power them from mains rather than batteries. Also one good hopeful point - larger sensors have been getting cheaper, I suspect a physically large sensor would negate some issues about alignments and such.
1-10 of 13Next »