What's the use of a physical shutter any more?

[MarkAlsop]

Reading this thread got me to thinking about other possible approaches that might be better suited to still photography than simply eliminating the mechanical shutter. Why not replace it, as a discrete device, with a solid-state shutter. LCD will not work for a number of reasons: Too slow, not transparent enough, and blocks too much light.

Remember that the science-fiction of the Star Trek communicator of the 1960’s has become the reality of the cell phone today and the iPhone is pushing into the area of the Tricorder. So…

First, consider that a photon is just distinct amount of energy at a specific wavelength traveling in a straight line. At the surface of a transparent solid the photon’s energy is absorbed by a molecule of that solid material. Because the molecules of a transparent material cannot hold that energy (has to do with the electron shells) it is transmitted to an adjacent molecule in the solid. This process happens over and over until the photo-energy reaches the other side of the solid where it is released (emitted).

I will ignore the issues having to do with refraction and diffraction since this will only concern light striking the surface at 90 degrees or parallel to the normal.

Now, suppose we could create a crystalline substance that was completely transparent in one state and completely opaque in another state. Further, suppose that the substance could be rapidly switched between these two states by the application of a current.

We turn crystal vibrators on and off in this way. Perhaps one of the methods for impregnating the surface of a crystal with atoms not associated with the basic crystal structure could be employed such that the application of a current to the crystal changed the energy absorption characteristics of the crystal molecules enough that they absorbed or reflected photon energy instead of passing to the next crystal molecule.

Such a substance would be a perfect solid-state shutter and the SLR could have it both ways: No moving components to wear out and still have a shutter so that the image sensor could be optimized for recording light without the need to perform the function of a shutter as well and we as photographers could still have complete control over the duration of an exposure.

[David_Buzzard]

The D1 series Nikons used electronic sensor timing above 1/250th. That meant that you really had an unlimited flash sync. About 1/1250, you started to lose flash power, as the shutter speed was shorter than the flash duration, but even at that speed, you had no shutter clipping of the image. While I don't believe that you can do the same thing with a CMOS chip, I can't believe that they didn't put the same thing into the CCD equiped D200.

If Nikon is planning a new model, the thing that will probably get them more business than anything would be the ablity to do full power flash syncing at 1/1000th, or even 1/500th of a second.

David Buzzard's Technical Blog

[Scott_Witcher]

I'm with Mr. Buzzard on this one!

The biggest advantage to my D70s in my opinion is the electronic shutter and the 1/1250 flash ability. I am no expert but I can't see why this could not work with CMOS as the sensor is grabbing timed amounts of light from longer shutter speeds. Also, if the sensor is using timed amounts of light and is independant of the shutter which moves at say 1/500 or slower, you could then leave the shutter open during a burst, you would have less wear on the shutter and only have to worry about the mirror. A longer life shutter would be really nice.

[peter s]

Originally Posted by Scott_Witcher

The biggest advantage to my D70s in my opinion is the electronic shutter and the 1/1250 flash ability. I am no expert but I can't see why this could not work with CMOS as the sensor is grabbing timed amounts of light from longer shutter speeds.

I'm sure it could but an interline ccd (like the one in D70s) needs to transfer the charge in the pixel to a hidden "pixel". That's how the electronic shutter works and because of that each pixel has less effective surface area (fill factor). And a smaller pixel always means more noise.

Cheers,
Peter

[washington678]

Thanks for the feedback. That is very useful


pret auto