I'd like to use a digital camera to replace the Polaroid camera on our electron microscope, but need the capability of at least 64 seconds of exposure time. Many cameras allow 8 seconds, and some Olympus models 16 seconds. Does anyone know of a medium- to high-resolution camera capable of longer exposures? How about manual exposure?

-- Winfield Hill (whill@mediaone.net), June 18, 2000

Answers

The Nikon CP-990 has a "BULB" setting in its manual mode which permits essentially infinitely long exposure times. You should be aware however that it, like all high resolution, prosumer cameras will exhibit significant noise during extended exposure times. It is reported that this noise can be minimized by chilling the camera.

-- Bill Cook (billc@wcook.com), June 18, 2000.

Now before people start putting their cameras in the freezer before taking them out into the damp night to catch a time exposure or two...
Please be aware that the effect of chilling will be minimal while the risk of repeated use of this technique will probably result in damage due to internal condensation.
The noise is caused by thermal reaction of the light sensitive elements - and if they could remain very cool there would indeed be less noise. Some astronomical cameras utilize cooling chambers to keep the exposure plane cool - but these were specially built for the purpose.
There are better techniques for mortals by taking two or more exposures of static opjects (I'll assume they're static since we are talking about long time exposures). There are software packages (including Photoshop) that have the tools necessary to average these images. The theory is that the noise will be different from pix to pix, whilst the image remains the same. Averaging tends to cancel much of the noise and can noticably improve the picture. This, of course, only works when you take multiple pictures of a static subject.

Des

-- Dan Desjardins (dan.desjardins@avstarnews.com), June 18, 2000.

Des. Isn't most of the thermal noise due to self-heating of the CCD during operation, and you need some way of continuously extracting heat from the chip to make any difference? Does this mean the longer the exposure, the greater the noise? I'm not sure about this. Perhaps the self-heating only occurs when the image is being "shifted" out of the chip.

-- Pete Andrews (p.l.andrews@bham.ac.uk), June 19, 2000.

Honestly I don't know. The noise does increase with longer exposures and the actual dynamics of the noise is beyond me. It would make sense that a continuous cooling would be necessary - I'm sure there are others who could add more intelligent information.
Does anybody have a good explantion of thermal noise?

Des

-- Dan Desjardins (dan.desjardins@avstarnews.com), June 19, 2000.