Science Forums 2

[CraigD]

Originally Posted by freeztar

Could it be, Craig, that you did not find or identify the shutter mechanism in the dig cams you disassembled? Or perhaps this is a myth of epic proportions?

Honestly, I've never considered the fact that a dig cam might not have a shutter. But now that I'm thinking about it, wouldn't this be easy enough to control on the CCD? Perhaps this is what they mean by electronic "shutter"?

I’ve done a bit of web browsing, and while not quite getting down to the level of actual parts specifications, am pretty sure that low price cameras, including the kind built into mobile phones and on keyring fobs, don’t have mechanical shutters. Instead, they’re purely electronic – when their button is pressed, they “drain” (the charge of each pixel of) their CCDs, then, when a timer reaches the fixed or calculated exposure time, convert each pixel’s charge to a (digital) number, process them into jpg format, and write the file to their internal memory (flash, I assume), or, if they have them, removable card.

As best I can tell, decent snapshot, midrange, and high-end digital cameras all have mechanical shutters. DSLR cameras, as their acronym suggests, also have a mirror that sends the lensed image through a viewfinder eyepiece, then “reflexively” retracts just before the shutter opens to expose the CCD, just like SLR film cameras. Some cameras, such as ones based on the recent “Micro 4/3" system”, keep the lenses, shutter, and CCD of a DSLR camera, but eliminate the mirror and optical viewfinder, instead keeping the shutter open and using the CCD – what the Micro 4/3 specification calls “live view”.

Because CCDs need to be drained in actual darkness to be as accurate as possible, these cameras produce lower quality images than ones with actual mechanical shutters.

Having looked at a picture of a disassembled Lumix camera, I’m pretty sure my wife’s DMC-FZ8 has a mechanical shutter, which I can’t hear because it’s too quiet, and sealed inside the camera body. My initial impression that it didn’t was, I think, wrong.

The best digital cameras, like the best film or plate cameras, have focal plane shutters, “curtains” that open vertically or horizontally to expose either the entire CCD, film frame, or place, or a partial strip of it that moves across the entire surface. For fast shutter speeds, it doesn’t matter much if each line of the surface isn’t exposed at precisely the same instant, allowing focal plane shutters to have effective exposure durations than the total time the shutter is moving.

The “digital shutter” alexander describes is, I think, accomplished by using the mechanical shutter to allow a good draining of the CCD pixels, then a computer-controlled timer to measure the charge of the CCD pixels before the mechanical shutter has actually blocked light from reaching them. Synchronizes with the movement of the shutter curtain, this can give very fast effective shutter speeds.

In short, the mechanical shutter isn’t used for timing, but to assure that the CCDs perform as accurately as possible.

Wandering into the realm of extreme photography, rapatronic cameras, which have been around since at least the 1940s, have shutters based on the Kerr effect that can achieve exposure times of around 1/100,000,000th sec. Even with extraordinarily sensitive photographic plates, they need very bright illumination, though as far as I know they’re used only to photograph nuclear explosions, which are plenty bright, so this isn’t a major problem. Since the Kerr effect is wavelength specific, single-shutter cameras based on it wouldn’t be much good for ordinary photography, as they’d “clip” significant parts of the visible spectrum. To take advantage of their super fast shutter speed, you’d also possibly need to use a nuclear bomb as a flash, so I wouldn’t be surprised if Kerr effect shutters aren’t on the consumer market anytime soon.