[stevebrot]

Does anyone make cameras with a shutter which really opens for just 1/3000?

Short answer is...No, not for the full frame for the full 1/3000s*. With electronic, it may be possible to come close, but even then there is not full synchrony across the frame.

This question is approachable via a mind exercise with the eventual conclusion that a shutter may approach full synchrony across the frame only as the frame size approaches zero. Of course nobody would indulge in such a mind exercise unless they were overthinking the realm of possible use cases. For the majority of use cases existing shutters represent a reasonable compromise where every point within the frame receives the same exposure, if not at exactly the same time.**

FWIW, freezing motion is sort of a happy side-effect of a faster shutter speed and is never fully effective if you look close enough (i.e. have capture media and optics with high enough resolution). If the image of the object shifts more than one pixel during the exposure, it is technically not "frozen".


Steve

(...wasn't this discussed in depth in the other thread?...)

* Even with leaf shutters it is possible for an object to traverse the frame, yet not be captured by the film/sensor.

** After all, the aim is not to define the same range of time (point "a" to point "b") across the frame. Rather, it is to provide the same period (x # of seconds) for equivalent exposure.

---

Last edited by stevebrot; 09-21-2015 at 12:05 PM.

[stevebrot]

This question is approachable via a mind exercise with the eventual conclusion that a shutter may approach full synchrony across the frame only as the frame size approaches zero.

I suspect there might be some sort of quantum effect where the shutter and the frame do not exist in the same place at the same point of time...


Steve

[Lowell Goudge]

I posted a question here recently about how the shutter works, and it just forms a variable-width slot which moves across the frame.

So speeds like 1/3000 are illusory when it comes to freezing motion!

I have found that once you get to say 1/800 there is NO further advantage in going faster if the reason for the high shutter speed is camera movement.

Does anyone make cameras with a shutter which really opens for just 1/3000?

Other Pentax Q has an iris shutter built into each lens. They run at 1/2000 max, and the Q-PK adaptor has a built in shutter that runs at 1/1000. These are real iris shutters, so among other things you can flash sync at full shutter speeds

[stevebrot]

Other Pentax Q has an iris shutter built into each lens. They run at 1/2000 max, and the Q-PK adaptor has a built in shutter that runs at 1/1000. These are real iris shutters, so among other things you can flash sync at full shutter speeds

Even leaf/iris shutters are not open across the aperture at a single instant. They are built so that the center has the same amount of exposure as the edges.


Steve

[peterh337]

This is hypothetical but you could make an electronic shutter using the same principles as used in welding goggles. A similar technology is used by the military, to protect pilots' eyes from the flash of a (distant) nuclear explosion. Operating time is tens of nanoseconds. Similar technology, very leading-edge and very guarded, is used to see through cloud/fog, by sending out a picosecond-long light flash and using a picosecond-speed "shutter" which opens a specific time later, so you select the exact distance ahead from which the light is being returned.

You lose a lot of light but it would be a solution to getting a "global shutter" at say 1/5000 or faster. The "closed" condition is not totally opaque so one would still need the 1/180 curtain, and one would activate the electronic shutter in the middle of the opening of the curtain.

The basic point I was trying to make is that my experience is that speeds like 1/3000 don't seem to do anything useful when it comes to freezing camera or subject motion. I am just working through about 1500 pics (K3, DNG) from a trip and it's obvious that anything above about 1/1000 does almost nothing useful. Maybe there is a small effect but it's for sure very small and not even a fraction of what one might expect as a pro rata of the shutter speed. The reason is obvious: the "1/3000" shutter is actually a 1/180 curtain with a slot in it!

[osv]

The basic point I was trying to make is that my experience is that speeds like 1/3000 don't seem to do anything useful when it comes to freezing camera or subject motion.

that would be wrong, as several people have already pointed out.

I am just working through about 1500 pics (K3, DNG) from a trip and it's obvious that anything above about 1/1000 does almost nothing useful.

did it ever occur to you that could be due to the type of things that you are shooting?

click this pic through to flickr, and look at the 1920x1080 version... the board is completely out of the water, his front foot is barely touching it, do you really think that this level of subject motion could be frozen at 1/1000th? do you see any artifacting due to your shutter slot theory?

[stevebrot]

The basic point I was trying to make is that my experience is that speeds like 1/3000 don't seem to do anything useful when it comes to freezing camera or subject motion. I am just working through about 1500 pics (K3, DNG) from a trip and it's obvious that anything above about 1/1000 does almost nothing useful. Maybe there is a small effect but it's for sure very small and not even a fraction of what one might expect as a pro rata of the shutter speed. The reason is obvious: the "1/3000" shutter is actually a 1/180 curtain with a slot in it!

????

At 1/3000 second, no portion of the frame was exposed for more than 1/3000 second. I would suggest that you are looking too close. During that 1/3000s a lot can happen. A little math and you should be able to calculate the image point shift present for even moderate subject motion.* I bet you will find that it will span several pixels and be readily apparent as blur when viewing a full resolution crop.

FWIW, I don't know that I have ever had any trouble freezing subject motion, not that doing so is usually desirable. As noted in one of the earlier comments, the traditional way this is done for fast motion is with a short duration flash in darkened room and even then there may be some blur because no flash is short enough if you look closely enough.


Steve

* Of course, with a moving slit shutter the ability to freeze depends on the direction of subject point motion within the frame. If the motion at is in the direction of curtain travel and greater than 2.8m/s (the curtain speed) there will be blur regardless of shutter speed. With a leaf shutter the same thing happens, but with different directional dynamics.

[DeadJohn]

I shot the Blue Angles sometime last year. I found that 1/4000 was too slow for them. Needed 1/5000 to freeze the motion...

For freezing motion on fast objects, panning is another technique that doesn't need such a fast shutter (and therefore helps manage ISO). Panning also adds to the blurred background effect if you like that.

[tuco]

640 Hz is more shutter speed than anyone will ever need. Um, wait, I mean computer memory.

[Digitalis]

640 Hz is more shutter speed than anyone will ever need.

I see what you did there...

[savoche]

640 Hz is more shutter speed than anyone will ever need.

Ah, Mr Gates, I didn't know you were a Pentax shooter

[clackers]

Ah, Mr Gates, I didn't know you were a Pentax shooter

Surely a Canon man! :-)

[wombat2go]

Even leaf/iris shutters are not open across the aperture at a single instant. They are built so that the center has the same amount of exposure as the edges.

https://app.box.com/s/6d06x1j0n6dm5et3jmypxpijpscdym8a

[stevebrot]

https://app.box.com/s/6d06x1j0n6dm5et3jmypxpijpscdym8a

Please interpret. I took a gander at the note sheet (image #5) and believe I know what they show, but don't want to jump to conclusions.


Steve

[SpecialK]

I suspect there might be some sort of quantum effect where the shutter and the frame do not exist in the same place at the same point of time...


Steve

It's the Eisenstaedt Uncertainty Principle.