liquid crystal as shutter, why not?

[jonathanalis]

Hello,

Im a new member, and Im starting studying computational photography.
I read some articles about eletronic and mechanic shutters, the advantages and drawback of both.
The real drawback of eletronic shutters is that they are attached to sensor eletronics, and cause some limitations. The mechanic shutter is independent, but is mechanical(now i think is time to get rid of mecanical stuff if possible, right?)... Also, both are limited in shutting capacity.
I was also studying LCD and how it is being used in projectors.
And I was wondering, why not using an sensor independent eletronic shutter?
LCD painels are eletronically controlled arrays of cells that let the light pass or not, you also choose the intensity.
My ideia is just put the LCD just above the sensors, at the same level as the color filters.
This way, is possible to control the shutter independent from the sensor speed readout, so, any form of exposure is possible.
So, you can implement global shutter, flutter shutter, rolling shutter, per-pixel shutter(this ive seen in a lot of articles, but they simulate it by opening a projector, getting the DMD chip and building a large optical system...), and lots of applications could be possible.
Is it viable?

Thanks in advance,
Jonathan

jonathanalis@gmail.com

[Manfred M]

Mostly because they are not good enough (too slow, poor light transmission). This topic comes up from time to time.

Cheap cameras use purely electronic shutters where data is polled from individual photodetectors.

See this thread for the most recent discussion on this topic: Focal plane shutters

Projectors I've seen all operate in dimmed light conditions, so that should provide some notion regarding the visual efficiency of these devices. Yes, one can always increase the output of the light source (increased output bulbs), with the inherent increase in noise (due to higher airflow to cool the projector) and associated heat. Put a fairly inefficient LCD device in front of a camera sensor, and the only way to improve lower light performance is to crank up the gain, hence more noise from various electronic elements as well as decreased levels of colour resolution.

[Dave Humphries]

Hi Jonathan,

Warm welcome to the CiC forums from me.

I can see what you're trying to achieve, but I know a little about the technical side of this subject and in my view it would make such sensors VERY expensive, due to reduced production yield.

I wouldn't be surprised if there might be specialist sensors already built along those lines for particular needs.

[JHzlwd]

Jonathon,

This is a good idea in principle and the future may very well belong to this type of shutter. In addition to some problems already mentioned in regard to current LCD mechanisms there is the matter of polarization. LCDs achieve variable light transmission by rotating the polarization plane of one layer of the device with respect to another. Used in front of a camera sensor this would differentially alter the brightnesses of different portions of the subject depending upon the extent to which light is polarized from different parts of that subject. For example, an object in the subject field having light predominantly at right angles to that of the active LCD layer would be rendered as nearly black even if it were red or blue. This phenomenon doesn't affect projectors because their source light (in effect the "subject field") is non polarized. In the world of things we see (and photograph), polarized - or somewhat polarized - light is more the rule than the exception.

Best wishes,
jh

[jonathanalis]

Thanks for the answears.

Now I see the problem.
I initially thought about LCD because i could be implemented as a focal plane shutter and would not spend lot of space in camera.
But I admit the idea came from projectors. There are also DPL projectors that use DMD, that are fast moving micromirrors(one for each pixel) in 2 possible orientations. one of them points to a light absorber and other to the projector plane. If a pixel remais more time pointed to projector plane, brighter is the projected pixel, if it remains more time pointed to light absorber, the projected pixel becames darker.

So, why not use a DMD as camera shutter? We avoid the polarization problem and have all the benefits already listed on first post.
But of course the camera arrangement would be more complicated. But it would still have a public, because lots of researches are simulation acquisition like this using an external optical system that include projector mirrors for shuttering the incoming light:
http://www.merl.com/publications/docs/TR2010-103.pdf
http://www.merl.com/publications/docs/TR2011-039.pdf
http://www1.cs.columbia.edu/CAVE/pub...i_ICCV2011.pdf
http://www1.cs.columbia.edu/CAVE/pub...yar_IJCV06.pdf (figure 4b of this article explains well how unpractical it is made, and it would be much more pratical if implemented by a camera manufacturer along with internal camera optical system)

tks

[JHzlwd]

Jonathon,

This is also interesting. Haven't thought it through much but could envision the micro-mirror array deflecting the light from the lens off to the side where the sensor actually resides. Micro mirrors get around the problem of polarization effects and they're fast.

Best wishes,
garth

[Adrian]

I would think, without possessing technical expertise, that it is only worth putting money into developing an improved shutter if the shutter is a material limiting factor in a camera. But is it? A high end camera shutter is capable of 1/8,000 second. The shutter system is extremely durable (100,000 -250,000 actuations). Probably manufacturers are better putting development funds into things that will give competitive advantage through improved picture rendition or much lighter overall weight and bulk, rather than creating a better paperclip.