Single diode camera idea

I thinked to take a photograph with single diode. Some university professors had been thinked the same but they say their technology takes ten seconds to record an image.
They scan the entire photograph with single diode with the help of flexing mylar mirror.
Mylar mirror samples the entire image from one corner to other and cast the one pixel images on to the diode in 10 seconds.

I have an other idea.

If we paint the lens with transparent dyes with inkjet printer and if we make a color mosaic on the lens with different spectrum dyes and than take a single shot with single diode of entire picture with the focused beam and than spectrum analyze the powers of each spectum line on the computer,
And if we know which spectrum band covers which area of the image and if we build the mosaic again


can we take pictures with this technology ?

Best ,

Mustafa Umut Sarac

Istanbul

That is a question posed to an Engineer or Physicist. Any comment in here would be fluff.

Mustafa seems to have placed interesting posts such as this in the past.

pellicle said:

Mustafa seems to have placed interesting posts such as this in the past.

Click to expand...

I never said it wasn't interesting. This is not a forum that lures people with detailed photon and ionic-bonding knowledge. Comments here would likely be of total ignorance.

Now if he asked us if we *wanted* such a thing to exist, a relevant answer could be given.

Mustafa is one crazy dude.

Can you scan a focused image in the field, building a photograph one pixel at a time? Sure.

Is it practical to do so? Not really.

It would take very accurate and seriously high precision mechanics. It would have to scan across the image plane in both directions while controlling overlap down at the micron level. This alone will make it impractical in the field. Just the temperature changes would make this extremely difficult. The power draw to run the mechanics would be non-negligible so you have a battery problem right up front. The time required to scan at a decent resolution would be very long (longer than drum scanning (which itself builds scan files one pixel at a time), which can take more than an hour), which limits the subject matter to that which isn't moving, or introduces some interesting motion artifacts in the scan file. I could go on...

There are digital scan backs already on the market: for example, the Better Light digital scan backs. These scan a line at a time. They have many of the same problems, but scanning a line at a time makes them considerably more manageable, and thus considerably more practical.