Jump to content

Why is a shutter still used in digital photography?


Daniel Smith

Recommended Posts

Hi.

 

Probably a silly question, but why are shutters still used in digital photography? Surely an eletronic servo of some kind could allow for signal intake at certain intervals.

 

anyone?

tnx.

Dan.

 

I'm not actually anything close to an expert on the subject, but one reason that springs to mind for still photography is that the only way to do true single lens reflex is with a mirror that obstructs the imaging plane.

 

Do digital video cameras still have a physical shutter?

Link to comment
Share on other sites

  • Premium Member
Do digital video cameras still have a physical shutter?

Some do (Dalsa, D-21, Viper), most don't. Dalsa and D-21 use the mirror shutter to provide a viewfinder image. The Viper guys had another reason, I don't remember what at the moment.

 

 

 

-- J.S.

Link to comment
Share on other sites

Many DSLR's actually don't have a shutter any more.. they do however retain the reflex mirror, which feeds the viewfinder and protects the sensor.

 

..one reason that springs to mind for still photography is that the only way to do true single lens reflex is with a mirror that obstructs the imaging plane...

The Single Lens Reflex system was a solution to a uniquely film based problem - how do you preview exactly what's going to fall on the capture medium, when the medium requires developing before you can see anything..?

 

Given the improvements in live view, combined with better resolution LCD's, I think the optical viewfinder is likely to be phased out of DSLRs... and the viewfinder will simply be a high resolution LCD screen previewing the sensor.

 

On the other hand it would be nice if Digital SLR manufacturers would stop compromising on their DSLR mirror mechanisms! I personally like the optical viewfinder, but it's frustrating to have the mirror fail.. who ever had a mechanical failure after only 100,000 clicks with a film SLR's?! My old Pentax K1000 was wearing out in all sorts of places, but the original shutter / mirror mechanism was still in great shape.

Link to comment
Share on other sites

On the other hand it would be nice if Digital SLR manufacturers would stop compromising on their DSLR mirror mechanisms! I personally like the optical viewfinder, but it's frustrating to have the mirror fail.. who ever had a mechanical failure after only 100,000 clicks with a film SLR's?! My old Pentax K1000 was wearing out in all sorts of places, but the original shutter / mirror mechanism was still in great shape.

 

Hey Daniel,

 

AMEN!

 

A lot of stopmotion animation is done with DSLR's these days, and that's one of the biggest complaints. After 100,000 exposures (if you're lucky!) you're basically stuck having to buy another camera.

 

Planned obsolescence… :angry:

 

 

Jay

Link to comment
Share on other sites

Hmm, battery saving? No, that is not really an issue, as all the D-cinema cameras need to have a stable internal temperature in order to make a consistent image and all need boot-up time.

The Arri D21 needs about 20 seconds to boot up and warms the sensor to 32 degrees Celsius (or cools down) immediately, otherwise it is not ready to shoot at all.

Besides, the spinning mirror motor has a very low usage compared to all the HD-circuitry.

 

No, one of the reasons is that a CCD or Cmos has to clear its pixels from data, to start all over again for the next frame exposure.

This is done, either by electronically "blanking" or clearing (restore to 0 volts) or a combination of that electronic blanking with a real black-out from the mirror in front of the CCD.

 

Some argue that the last method gives a motion blur that is more equal to what film sees..... and therefore looks less....video.

 

When the shutter is shutter opening is more than 180 degrees we get more motion blur. Electronic shutters can open as wide as 359.1 degrees, which gives an extreme motion blur but also one stop light GAIN.

 

An Arri D21 or Dalsa is limited to the physical size of the shutter, 180 degrees. which can only be narrowed.

Link to comment
Share on other sites

  • Premium Member

No no; rob, I mean more so for when you're squaring up a new set up and the like; well before you'd turn the camera on; just peeking through the viewfinder. while you tweak before actors get on set etc. I mean i even leave my Arri sr3 off for that kind of stuff... no need for me to drain battery when i'll set up a shot and still be waiting for awhile before people filter into set.

Link to comment
Share on other sites

Hey Daniel, another big reason for the use of a mirror in dSLR's is due to heat noise of the sensor. The longer a sensor is activated, the more it heats up...and of course, there's the extra noise. Personally I like looking at the actual light through the mirror as opposed to the pixels of a LCD.

 

I talked to Nikon about the implications the heat issue and the D90's video mode. I questioned if they had found a way to counteract that for longer shots using the D90. The tech guy told me that the video would get progressively more noise as the shot length grew. However, he stated that it would not be readily noticeable unless you cut the end of a long take with the beginning of another shot. So there still is no fix to the time/heat/noise issue...probably not until they get miniaturized, water-cooled heat-sinks behind the sensor (just my guess).

Link to comment
Share on other sites

  • Premium Member
Hi.

 

Probably a silly question, but why are shutters still used in digital photography? Surely an eletronic servo of some kind could allow for signal intake at certain intervals.

 

anyone?

tnx.

Dan.

Shutters are only required for CCD sensors.

 

When a frame is captured by a CCD or CMOS sensor, the variations in light intensity are recorded as a pattern of electrostatic charges on the photosites, which are basically microscopic solar cells.

 

With a CMOS sensor, the readout circuitry simply reads all the voltages and stores all numbers in the video processor's RAM chips. When the readout is finished, the cells are all shorted out to reset the charges for the next frame.

 

With a CCD sensor the pattern of charges has to be physically moved across the chip for the readout process, and this would have the same smearing effect as if film was moved without a shutter.

 

The most common way to overcome this is to have a second set of storage cells mounted directly below the light gathering ones, and first dump all the charges into those berfore reading them out. Because they are then in the "dark" the incoming light shouldn't affect them. Unfortunately silicon is slightly transparent to red light, and that is what causeas the well-known "Vertical Smear" you see on video shots at night when there are bright lights in the picture.

 

A mechanical shutter completely eliminates this problem, and also allows the entire silicon surface to be used for gathering light, rather than having to sacrifice some of it for the anti-smear circuitry. This is why the Dalsa Origin has such an excellent low-light performance.

 

The "no-power" viewfinder function is simply a happy by-product of this. Apart from allowing you to frame shots with the power off, it also allows critical focus without needing an expensive monitor.

Link to comment
Share on other sites

Shutters are only required for CCD sensors.

 

When a frame is captured by a CCD or CMOS sensor, the variations in light intensity are recorded as a pattern of electrostatic charges on the photosites, which are basically microscopic solar cells.

 

With a CMOS sensor, the readout circuitry simply reads all the voltages and stores all numbers in the video processor's RAM chips. When the readout is finished, the cells are all shorted out to reset the charges for the next frame.

 

With a CCD sensor the pattern of charges has to be physically moved across the chip for the readout process, and this would have the same smearing effect as if film was moved without a shutter.

 

The most common way to overcome this is to have a second set of storage cells mounted directly below the light gathering ones, and first dump all the charges into those berfore reading them out. Because they are then in the "dark" the incoming light shouldn't affect them. Unfortunately silicon is slightly transparent to red light, and that is what causeas the well-known "Vertical Smear" you see on video shots at night when there are bright lights in the picture.

 

A mechanical shutter completely eliminates this problem, and also allows the entire silicon surface to be used for gathering light, rather than having to sacrifice some of it for the anti-smear circuitry. This is why the Dalsa Origin has such an excellent low-light performance.

 

The "no-power" viewfinder function is simply a happy by-product of this. Apart from allowing you to frame shots with the power off, it also allows critical focus without needing an expensive monitor.

 

Fascinating! Thank you!

Link to comment
Share on other sites

Hey Daniel, another big reason for the use of a mirror in dSLR's is due to heat noise of the sensor. The longer a sensor is activated, the more it heats up...and of course, there's the extra noise. Personally I like looking at the actual light through the mirror as opposed to the pixels of a LCD.

 

I talked to Nikon about the implications the heat issue and the D90's video mode. I questioned if they had found a way to counteract that for longer shots using the D90. The tech guy told me that the video would get progressively more noise as the shot length grew. However, he stated that it would not be readily noticeable unless you cut the end of a long take with the beginning of another shot. So there still is no fix to the time/heat/noise issue...probably not until they get miniaturized, water-cooled heat-sinks behind the sensor (just my guess).

 

 

Interesting, how different companies have different solutions and problems, with basically similar components.

 

I have used a Visario Speedcam extensively, it has a Cmos imager and goes up to 1000 fps full resolution.

This camera has to have an internal working temperature of 50-55 degrees Celsius, or else the noise and artifacts become visible.

http://optics.org/cws/product/P000003259

 

The Arri D21 has a Peltier element behind the sensor, stabilizing the temperature on exactly 32 degrees.

http://en.wikipedia.org/wiki/Peltier_effect#Peltier_effect

Link to comment
Share on other sites

No no; rob, I mean more so for when you're squaring up a new set up and the like; well before you'd turn the camera on; just peeking through the viewfinder. while you tweak before actors get on set etc. I mean i even leave my Arri sr3 off for that kind of stuff... no need for me to drain battery when i'll set up a shot and still be waiting for awhile before people filter into set.

 

 

I understand you, but with D-Cinema cameras it is a different deal.

Of course the camera man can look through immediately, but the whole issue with D-Cinema is that you (or better, the video technician / DIT) has to monitor the status of the camera constantly. Switching on just before you start to rehearse or shoot is just not an option.

 

The amount of menus and variables and options is so great that if you do not check them regularly (or better, every setup) you might find that you recorded something else then what you saw on the groundglass.

 

Not only that, it is the only way to check if the right signal is going to your recording unit, which is for instance a Sony SRW-1 deck with processor. If you shoot 4:4:4, and one of the 2 cables is badly connected or damaged or by mistake longer than the other, you end up with something else on tape.....

 

D-Cinema is definitely more demanding than shooting on film, from a technical/electronic point of view. IMHO.

Link to comment
Share on other sites

  • 4 weeks later...
  • Premium Member
This may be a dumb question: what is the visible difference in quality between CMOS and CCD censors, if there is any?

There's very little difference, since they both convert photons to electrons using silicon in pretty much the same way.

 

CCD's -- especially cheap ones -- can get streaks trailing from points that are extremely overexposed, because of the way they shift the image off one edge of the chip. CMOS, because of the processing being done right there on the chip, doesn't have as large a percentage of the surface light sensitive -- a lower fill factor.

 

 

 

 

-- J.S.

Link to comment
Share on other sites

Playing Devil's Advocate, why not?

 

Why would you not want to protect the sensor from light in-between periods of exposure?

 

What advantage does replacing a cheap, simple, efficient method of controlling exposure to light with an expensive system that has to be programmed into the sensor have?

Link to comment
Share on other sites

This is a really interesting discussion and something I've thought about a lot since I'm such a goddamn dinosaur with my silly film cameras. It's cool to finally start to understand this stuff from a better standpoint technically, AND to want to learn more. Yay, thanks guys!

Link to comment
Share on other sites

  • Premium Member
Playing Devil's Advocate, why not?

 

Why would you not want to protect the sensor from light in-between periods of exposure?

 

What advantage does replacing a cheap, simple, efficient method of controlling exposure to light with an expensive system that has to be programmed into the sensor have?

 

Regarding the first line, it's a cost/benefit analysis. Why would the sensor need protection? The most vulnerable sensors would be the single chip color filter arrays (Genesis, F-35, or Bayer). But the dyes used are stable enough that the cameras will be antiques before fading becomes an issue. A shutter would only "protect" the chip in the brief time between frames. You can't leave it closed during setup, or you'd have nothing on the monitor to use. You could put a lens cap on during any down time if you're worried.

 

A mechanical shutter is an expensive extra subsystem for an electronic camera. It would also need to be latched open in order to use the near-360 degree shutter angles that are possible with digital cameras. Having it doesn't save any money on the electronic or firmware side. It does solve the CCD streaking problem, though. It wouldn't help CMOS at all.

 

 

 

-- J.S.

Link to comment
Share on other sites

  • 2 weeks later...

Ok thanks.

 

 

Another question, this is slightly vague but I'll try to explain it as best I can.

 

In audio balanced connectors are used to shield the signal from common mode interference, CCD's are also vulnerable to this sort of interference.

It's not just gain adjustments that fires up noise.

 

So, I was wondering if CCD's have any similar technology that shields it from this interference? Could the capacitors somehow be connected to the system with two phases 180 degrees opposite?

 

Sorry for the poncey question, but I was delving into balanced and un-balanced signals the other day and wondered if the same principles could be used on CCD's.

Edited by Daniel Ashley-Smith
Link to comment
Share on other sites

In audio balanced connectors are used to shield the signal from common mode interference, CCD's are also vulnerable to this sort of interference.

It's not just gain adjustments that fires up noise.

 

So, I was wondering if CCD's have any similar technology that shields it from this interference? Could the capacitors somehow be connected to the system with two phases 180 degrees opposite?

 

Imaging systems use some notions that are not exactly similar to those in audio domain but still provide some sort of noise immunity, such as Correlated Double Sampling.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...