This article is a continuation of a series of articles on issues raised earlier in the articles'Shutter noise'and'Disadvantages of modern DSLR cameras'.
Mechanical shutter
Modern digital cameras use focal shutters of a curtain-slot type with a vertical stroke of the blinds. This means that such a shutter is located immediately in front of the camera’s matrix, consists of shutters that move vertically (usually from top to bottom and back).
The following illustrates how the shutter is released:
1 video.
Pay attention to how much shakes the mirror after lifting and returning, as well as how shutter curtains tremble monstrously... The video shows that the shutter curtains consist of several parts (the so-called slats or 'louvers').
2 video.
In this video, you can notice the gap that forms during the movement of the shutter curtains.
3 Videos.
Full-format camera and cropped camera.
4 video.
Shakes not only the mirror and shutter shutters, but also the aperture blades.
And a little discussion about the shutter, using the camera as an example Nikon D80.
Exposure The synchronization of this camera is 1/200 second. This means that it is precisely this period of time that the shutter curtains need to travel a distance equal to the height of the matrix.
If you need to shoot at shutter speeds slower or equal shutter speed synchronization, then the shutter will work as follows:
- The first curtain opens, it takes 1/200 second.
- Held Exposition, at this time, the matrix remains fully open. Take the shutter speed 1/60 second as an example. The second curtain starts its movement 1/60 second after the start of the movement of the first curtain.
- The second curtain closes, it takes 1/200 second.
- The curtains rise together to their initial position.
At such shutter speeds, it is easy to synchronize the flash and shutter. Usually, the flash fires after the first curtain (as soon as the shutter fully opens the matrix), or before the second curtain begins to move (before the shutter closes). For example, the impulse of my Nikon flash SB-910 has a duration from 1 \ 800 s to 1 \ 40.000 s depending on the power. When the flash fires, the camera’s sensor is fully open and there is no synchronization problem.
If you need to shoot faster excerpts synchronization, then the shutter will work as follows:
- The first curtain opens.
- The second curtain does not expect a full opening of the matrix and begins its movement after the first. The delay of the second curtain determines the time exposure. Take, for example, the shortest shutter speed allowed for Nikon D80 - 1/4000 p. In this case, the second curtain begins its movement in 1/4000 s after the start of the first curtain movement, and thus the two curtains move together, forming a moving slit, which produces the exposure.
- The curtains rise together to their starting position.
It is difficult to synchronize the flash with the shutter at such speeds. If the flash fires only at a certain moment, then in the picture we will get a strip, which is formed by the shutter slit. To get around this limitation, high-speed sync flash units are used that “light” the entire time both shutters move to avoid streaking.
Interestingly, if we shoot for 1/60 of a second, then the shutter actually needs much more time for its work. So, it takes 1/60 s to lower the first curtain and wait for the second, 1/200 s to move the second curtain and at least another 1/200 s to lift both curtains to their original position (perfect case, in reality, you need more time). Total 1/60 + 1/200 + 1/200 = 2/75 s. If you remove the restrictions on the operation of the mirror, aperture and processor of the camera, then in one second under ideal conditions it will be possible to take no more than 38 frames, and this is mechanical restriction of burst shooting.
At the same time, cameras that use the electronic shutter, which does not need to spend time on the movement of the curtains, can now easily shoot at 60 frames per second in photo mode (as an example, look at Nikon 1 J1). Just imagine how useful it would be for photojournalists and sports photographers to photograph certain events at such tremendous speed. For example, the fastest DSLR for 2014, the Canon 1DX, shoots at a maximum of 14 frames per second, which is 4 times lower than the 60 fps of some electronic shutter mirrorless cameras. The only problem is that modern cameras with an electronic shutter have their drawbacks, for example, they suffer from 'rolling shutter', etc. and for now, one can only dream of an electronic shutter that has the positive qualities of a mechanical shutter and a huge shooting speed.
By the way, the “real” speed of the shutter curtains is easy to calculate. Matrix height Nikon D80 is 15,8 mm, the shutter passes this distance in 1/200 second, and its speed is 3,16 m / s or 11,38 km / h, which is quite a bit :)
Thank you for attention. Arkady Shapoval.
Good article! Very clear and understandable!
“The first curtain opens, it takes 1/200 of a second.
An exposure is carried out, at which time the matrix remains completely open. Take the shutter speed 1/60 second as an example. The second curtain starts its movement 1/60 second after the start of the movement of the first curtain.
The second curtain closes, this takes 1/200 second.
The curtains go up together to the starting position. "
Arkady, the clockwork is waiting not 1/60, but 1 / 60-1 / 200.
In FED cameras 3, 4, 5, the sync speed is 1/30, and so there the clock mechanism at a shutter speed of 1/15 is set to 1/30, and at 1/8 to 3/30. A total of 3/30 delay times + 1/30 sync time gives 4/30 which gives 1/8 (roughly speaking). Since the days of the FEDs, little has changed.
I apologize in advance, maybe I misunderstood something :)
Let's take instead of 1/60 the limiting case - 1/200, according to your logic, the second curtain would start its stroke 1 / 200-1 / 200 = 0 seconds, after the first one, so we would just get a completely dark picture, since there would be no the gap between the curtains.
"Arkady, the clockwork is WAITING not 1/60, but 1 / 60-1 / 200."
At 1/200, the clock mechanism is WAITING 1 / 200-1 / 200 = 0, that is, this is the sync speed. Shutter speed is calculated as a clockwork + sync speed
"A total of 3/30 delay time + 1/30 sync time gives 4/30 which gives 1/8"
from our example 0 + 1/200 = 1/200.
You have to admit your mistakes and get better, and not get out.
Thanks for the review, good shots :)
I always correct, if a mistake was made, do not breed demagoguery. So far, the errors that I wrote do not see. Perhaps try to explain in more detail.
I don’t know how to explain it easier :) There is nowhere easier.
http://img9.joyreactor.cc/pics/comment/%D0%9A%D0%BE%D0%BC%D0%B8%D0%BA%D1%81%D1%8B-%D0%AF%D0%BE-%D0%9C%D0%B8%D0%BD%D1%8C-fuuuu-%D0%BC%D0%B0%D1%82%D0%B0%D0%BD-30119.jpeg
So you don't know. On business, please note that I have written "The second curtain will begin to move in 1/60 second after the start of the first curtain." - exactly after the start of the movement of the first curtain, and not after the first curtain opens, otherwise it would be as you indicated - 1 / 60-1 / 200.
Yes, I deeply apologize. Wrong. Thank you for the clarification :)
Good. In any case, thank you for sharing your observations of the old film cameras :) If it's not difficult, share your experience, how long were FEDs with fabric shutters?
Cloth shutter as lucky. For example, at home there is a father’s Zenith-E (~ 1977), so there is also a fabric shutter, and after 20 films, the shutter stopped closing normally. A few more films and he tore. But this is not an indicator, since most likely there was a factory marriage.
Soviet fabric shutters (Barnack shutter) were flawless until the early 70s of the last century (a characteristic feature of cameras: a 3/8 inch tripod thread). In terms of workmanship, they were not inferior to Leica shutters. My Zenit-3 made in USSR in 1962 (purchased in 1965) is still in working order, and not only the curtains, but also the mirror drive mechanism with a “native” nylon thread. The average load was about 10 films per year. Thus, the shutter withstood about 18000 operations.
1 / 60-200 will give not 1/60, but just 1 / 60-1 / 200 - this is the time between the full opening of the first curtain and the beginning of the movement of the second, and not between the beginning of the movement of the first and second, there is just 1/60
If you dream of a completely “digital” camera, you can replace the shutter, remove the mirror ... but what about the diaphragm? let's say the amount of light we can control the ISO, but the depth of field? is there any option to control depth of field without aperture?
How about the principle of the diaphragm as described by me here - https://radojuva.com.ua/2012/12/dslr-mechanics-future/
A liquid crystal diaphragm, which would not all darken, but areas darken with rings. It would then be possible to realize bokeh in the form of rings (as on mirrored lenses), if desired, by dimming the middle.
Paid plugins for various patterns would appear :)
LCD elements, even in an inactive state, will significantly increase the otic density of the lens (LCD material + some kind of polarizing filter for obtaining opaque zones), therefore, this method will be relevant exclusively in the amateur sector, since there can be no talk of a full lens "drawing" ...
d.martyn, you are wrong. Already now there are glasses that have a very high transparency in the inactive state, and the darkening is achieved by “liquid ink”, which in the inactive state is concentrated on the outer circumference of the glass. Such glass is absolutely no different from a regular lens. In lenses, from 3 to 15 optical elements are used anyway, and from this the quality of their pattern is not lost.
Igor, a strange thesis. In comparison with what quality is not lost? I generally believe that an ideal lens should be without lenses. True, unfortunately, the laws of physics do not agree with me.
In video 4, Canon’s aperture doesn’t shake as much as Nikon’s, is it good or bad for the lens aperture?
Vibration means insufficient fitting or balancing of parts, and hence increased stress on the mechanics. And if there are backlashes, then this is shock loading. Could this be good for aperture? Definitely not.
Kenon and Nikon have fundamentally different principles of diaphragm control (namely the jump rope). Nikon (both Pentax and Sonya) has a kinematic connection with the camera, and Kenon has a drive in the lens itself. That is why Kenon is “softer”.
On cheap cameras, the shutter and diaphragm are one unit. just the central two-leaf (well, for example ...) shutter does not close completely, but to the desired aperture.
Hello everyone! I took my first steps in photography with my neighbor's “Kiev”. I remember that the camera was a rangefinder and focusing was carried out with a wheel on the top of the camera. So, it had a shutter metal shutter. The fastest shutter speed was 1/1250. curtain shutter, very short. I did not have to use such a shutter speed. On my own household there were FED 5V, with Olympic symbols and ZenitET. They had rag shutters. In those days, few 36 frames were shot that needed to be processed. About the resource they can probably say Pro. I shot Zenith for about 15 years and lately at long exposures (up to 125) the second curtain wedged up. I think because of the dirt. There was a central shutter on Smena-8m (I can be wrong in the name). These shutters seemed to be synchronized at any shutter speed. Curtains for 1 \ 25-1 \ 30.
Wedges due to dirt, grease thickening, and wear / breakage.
The resource of Soviet cameras was rather low, especially when compared to modern cameras. I don't remember the numbers, but after several thousand they flew out at full speed. I myself have a couple of Soviet cameras with faulty shutters, and one friend on business trips with her own killed more than one shutter - there are several Zenits with jammed shutters in her house. They broke down precisely on the basis of several thousand frames.
So, in modern DSLRs, the shutter life is higher than old mechanical cameras, apparently much less mechanics (the electric drive has eliminated a lot, and most importantly the shutter speed setting mechanism).
But I wonder how old you are? What boldly talk about the resource of Soviet cameras. After all, the USSR has not existed for 23 years.
I'll put in my 5 cents :)
According to the passport, the zeniths E, ET, EM, TTL, 11, 12 have a resource of 500 rolls of film for 36 frames, let’s keep a margin that you can make up to 40 frames on a film and play a margin, let even 1000 rolls, everything is exactly small . The reason is very simple, firstly the hackish production of mechanics, secondly, the mechanics themselves are extremely oak, which only means that the curtains have a very large tensile load at the time of start and stop, the mirror is driven by the same mechanism of curtains, etc.
Now on the used market, 60-70% of zeniths are either dead or half dead or will die very soon.
Another thing is FED, in FED, first of all, very simple mechanics. Fewer details, less chance of breakdown, secondly, in FEDs there is a damper and shock absorption for the curtains, the curtains themselves are better rubberized. I never found information about the resource, but I saw only one dead FED, and then after a lot of sand got into it, all the rest, including those on the used market, are workers.
It is easy to check on the curtain amortization account:
take the zenith cock the shutter with your finger, raise the mirror to the stop and click what we hear, clanging sound with pop, the zenith curtains work this way.
In the Fed, a humbling sound is barely audible.
Stupid question about age. What is the difference, how old I am and how long the USSR does not exist, if there are facts of using Soviet technology? And if there is passport data for cameras in which very modest numbers are indicated?
The acquaintance I mentioned actively used these cameras just in Soviet times. I didn’t kill a single camera because I used them a little and didn’t “shoot” even half of the resource.
Your conclusions about the unreliability and low service life are really true for Soviet SLRs, regardless of the model and type of shutter. The most valuable acquaintance of the Soviet professional master in repairing cameras (thank God! - though they did the lenses qualitatively). But your conclusions are completely wrong with respect to imported mechanical and electronic-mechanical DSLRs. This is generally a different world, a different level of design and assembly technology! This is simply an unkillable technique, and it’s pointless to count the number of operation cycles for it: neither patience nor life is enough. Click, for example, Nikon FM-2 or Olympus OM-2, and you will understand everything.
I can't find where, but on some foreign site I read about the beginning of the production of special glass, which is somewhat similar to the technology of "electronic ink" - that is, under the action of a current it becomes either completely transparent or completely black and does not transmit light. The rate of change is 1/20000 of a second. It is planned to use it instead of a shutter, but a mirror is still necessary.
The first “ones” of both companies are absolutely “champions” in speed: they work out 1/16000, yak from the ground :)
The shutter speed is 1/500, therefore, with the same minimum gap, instead of 1/8000, 1/16000 is obtained :)
Well yes. For "athletes" - that’s it.
In theory, it is now technically possible to make a DSLR without mechanics. Sony makes his mirrors with a translucent fixed mirror. Electronic shutter technology, as the author wrote, is also available. Well, to make a liquid crystal diaphragm is probably now also possible.
But this is why manufacturers do not make them this is a separate issue.
What's the point? There are mirrorless cameras. It is still, they are positioned as advanced soap dishes. But for me it’s so over time, when there will be more full-frame, or less cropped than 4/3, they will quietly erase the traditional SLRs.
Mirrorless cameras (for example, Sony NEX) differ from DSLRs only by the absence of a mirror, the rest of the mechanics remain with them (electromagnetic diaphragm and curtain shutter). In the diaphragm, I think there is no big problem, tk. it is in the lens, firstly (that is, it does not affect the wear of the camera itself, and non-mechanical diaphragms will appear - you can limit yourself to buying a lens), and secondly, in my opinion, the resource of the diaphragms is higher than that of the shutters. But the electronic shutter is, of course, good. the mechanics knocks, consumes electricity, and, what is saddest, wears out.
Of course, I am a conservative, but still IMHO: a carcass is a consumable, glass is for a long time.
It depends on which carcass.
if glass is cheaper than a carcass, then you are clearly doing something wrong))))
Still, do not forget about the speed of operation of such compact cameras, even if they are full-frame, yes, even if they are medium format. They focus on sharpness (yes, most likely, they will focus on this method for a long time), they are in contrast using a matrix, and not using a focusing module. I'm not talking about the rest of the body kit ... Yes, and the price is not very pleasant for such compacts.
Phase autofocus is already built into the matrices, while it works in tandem with contrast detection and is obtained quickly (due to phase detection) and accurately (due to contrast detection) - and there is no hemorrhoids with misalignment in DSLRs, when the sensor is focused, but not in the matrix ...
An excellent explanatory review! Finally, a video appeared, this is the next level of your blog Arkady. I won’t comment much and discussed so much. I will only say one thing: if you remove or change the mirror, we will lose a great name, the mirror will not be a mirror but at least a hybrid.
By the way, the "real" speed of the shutter curtains is easy to calculate. The height of the Nikon D80 matrix is 15,8 mm, the shutter travels this distance in 1/200 second, and its speed is 3,16 m / s or 11,38 km / h, which is quite a bit :)
thoughts in hearing are not fatal, but acceleration up to 11 km / h in 1/200 second a person would probably die with such acceleration
“Acceleration up to 11 km / h in 1/200” - 1/200 is essentially the total travel time of the curtain, and the acceleration in much less time.
Hello Arkady.
I have FED NKVD “Trud. commune of the NKVD of the Ukrainian SSR. FE Dzerzhinsky "copy of the watering can I find it difficult to say, presumably 1934-1936. So it is in full working order and the shutter too.
And the neighbor's Zaporozhets is hunchbacked in the garage, on the move and completely serviceable. What is it for? It all depends on the operating time, on the quality of a particular copy, on the operating conditions. The fact that your copy itself is serviceable does not mean anything. Now, if you said that “I shot 10 on it and is like new” - that would be an indicator ...
By the way, it is better not to use such old cameras without prophylaxis. In such a time, the lubricant turns almost into plastic, and the mechanics are working to wear ...
The shutter is curtain.
This is precisely the peculiarity of the Soviet apparatus: each instance is unique and in a certain sense unique. How will they assemble, who will assemble, who and how will make the details, whether at the beginning of the month, at the end of the year, God alone knows! One copy can work for forty years without any problems, the other is full trash, a set of spare parts. Soviet camera equipment has never been stable in terms of product quality, and the reliable operation of a camera personally does not characterize the entire product. For the most part, Soviet cameras were structurally simple, therefore very repairable. This quality has always helped photographers.
The equipment becomes obsolete every six months, and after 3 years it is completely irrelevant. So what's the difference to collectors how much the shutter is worn out?
I'm even confused.
Sync Exposure is the minimum time at which a frame is fully open. right?
If yes, then with a shutter speed of 1/200, the frame is fully open for exactly 1 / 200s. That is, the second curtain starts to move in 1 / 200s from the moment the first frame is fully opened or from the moment the first starts to move?
1/200 from the moment the frame is fully opened first.
Ugh, I thought one thing, wrote another - from the moment the first one began to move (so that each part of the matrix was exposed to 1/200) - at shutter speeds shorter than the second curtain moves from its place before the first one opens the frame.
maybe not quite in the subject, but it has been interesting for a long time, but what is the mechanism of DSLR-video shooting (on DSLRs) - you can also set the shutter speed there - but not fur. shutter ...?
It uses an electronic "shutter". Because of this, the quality of the shooting drops, at least, a Rolling shutter appears (when the image is deformed in dynamic scenes), all sorts of stripes from bright light sources can also come out.
Shutter speed when shooting movies? Do not quite understand. There, the mirror flies up, the shutter opens and remains open until you turn off the video. And the matrix has been working in liveview all this time.
Whether shooting video or liveview, image brightness is adjusted by changing the shutter speed of each frame. Therefore, it turns out that there is a limitation on the minimum illumination during video shooting. the shutter speed cannot be longer than the duration of the frame, which in turn depends on the frame rate. Those. at 30 frames per second, shutter speeds greater than 1/30 cannot be achieved. This is clearly visible on cameras, which in livewiev raise the sensitivity to maximum and we see a noisy but bright picture during viewing (and often the frame rate decreases and the exposure of each frame increases), and after switching to video mode, the sensitivity is set to the maximum working exposure frame rate and the screen darkens.
just, as far as I know, in DSLR video, you can adjust the shutter speed (of each video frame) in the video settings ... - but how is it technically implemented?
You can adjust the shutter speed during video recording or not - it still changes during video recording (manually or automatically, respectively).
When shooting a video, the shutter speed, for example, can be adjusted by the time between reading information from the matrix and resetting it before exposing the next frame. As far as I remember, cameras do something similar, why not do it in the same way.
Great article, everything is clear! Those who read it are worth seeing this youtu.be/ZT1iCze2wIo video for “fixing the material”.
Write (with examples) about the distortion of moving objects, with a shutter speed shorter than the shutter speed.
Guys .. Any movement starts at zero speed with acceleration.
It turns out, on top of the matrix (at the bottom of the picture), where the curtains begin to move, does the exposure last longer ??)
curtains do not move so fast