Almost all of the DSLR cameras have focal plane shutters covering the film or sensor plane like a curtain (shutter in the camera body instead of in each lens). Focal plane shutters have been used in the better 35 mm cameras since the first Leica in 1925 (and FP shutters are older than that), because they provide superior timing of fast shutters. The really big deal about focal plane shutters is their precise accuracy. They use the same motor, same gears, same curtain speed, for EVERY shutter speed, be it 30 seconds or 1/8000 second. No complicated speed switching or gearing. One simple precise mechanism, with one mode and one curtain speed. One curtain opens the shutter, and a second curtain closes it. The exposure is controlled by the starting time of the second curtain. The accuracy is a big deal. And today, timing between the curtains is quartz crystal clock controlled, providing very fast and accurate 1/8000 second speed, and also a very slow 30 second speed (both are very difficult for a coiled spring, those older shutters only went to one second). This focal plane system is great for interchangeable lenses, because it only needs one good shutter in the body, instead of a shutter in every lens. And this is a better shutter, but FP shutters do have the one limitation of maximum shutter sync speed with flash (this page).
The duration of electronic flash is extremely brief, but the shutter must be fully open first, so that all areas of the frame are exposed simultaneously. At focal plane shutter speeds faster than maximum sync speed, the shutter is not fully open, and we see a black unexposed band at the top or bottom of our flash pictures (focal plane shutters are vertical travel today). If you rotate your camera to portrait orientation, you will of course see this dark band at one side.
So flash sync speed is an issue, often limited to 1/200 second shutter speed in many cameras. We'll show that issue first before the text descriptions.
Examples: These Nikon D300 and D800 (and the D7xxx family too) use different shutters with different life expectancies, but these (faster) better grade shutters offer sync at 1/320 second. The specs say 1/250 second, but they work well at 1/320 second. There is a warning in the manual, since 1/320 might cut off a little of a speedlight at its longer full power level.
All images are always inverted, the lens projects images upside down on the sensor, so the actual travel is the opposite direction (and note that the focal plane shutters on these two camera models are moving in different directions). The illuminated area is the traveling open slit width, which is fully open over full frame only up to maximum sync speed. At faster shutter speeds, the rear curtain begins closing sooner to make the open slit be more narrow. The speedlight flash is fast enough to stop the moving slit. These pictures simply show that open slit.
Many camera shutters only sync flash at up to 1/200 second (specs, Flash Sync Speed, also called X sync). Any time you see a problem like this, just reduce your shutter speed, so that it does not exceed maximum flash sync speed (in your camera's specs). This is what maximum sync speed is, it is a shutter situation, about when that shutter is fully open. Maximum sync speed is NOT a property of the electronic flash. Instead it is simply the fastest focal plane shutter speed that the front curtain is fully open, and the rear curtain has not started to close (slit is open wide enough to provide a full open frame for the flash). But sometimes radio triggers add a delay which causes the same maximum sync problems (to diagnose, try flash with and without the radio trigger. Also then it might just be time for new batteries in the radio trigger).
The band may not be fully black, you might see something in it if there is enough ambient to show it, but the band area will block the flash. The darker band may not always even be very noticeable in bright sun. FWIW, if this were a FX camera shooting DX crop, that might crop a narrow dark band off, so it might get away with a slightly faster shutter speed than if shooting full frame. The two shutters appear about the same speed here, but the FX shutter has to travel 1.5x greater distance than the DX shutter.
And yes, this example also shows the difference in the DX and FX view with the same lens - The DX view is more telephoto (cropped and enlarged more), and the FX view is more wide angle (not cropped).
These samples were a SB-800 speedlight flash (fast flash - 1/17800 second at 1/32 power, which stops the shutter travel too). The flash exposure is not affected by shutter speed, where we are able to see it. However, this is of course regular flash mode, and is NOT Auto FP flash mode (next page). This flash was not in the hot shoe, and was instead triggered with a PC cord - so that the camera was unaware any flash was present, so it allowed shutter speeds faster than maximum sync speed.
If the Nikon camera can recognize that a flash is present, it will prevent exceeding maximum sync shutter speed with a flash, even in camera manual mode. If the flash is off camera or the brand has no communication with the camera (no flash known present), then no shutter speed is prevented.
An undetected flash (off camera on radio trigger or PC Sync cable, or a non-system flash (off brand or studio flash, etc) will not prevent this issue from occurring. I used a PC sync cord off camera to be able to take the too-fast shutter speeds above. The point is, with the exception of HSS flash mode, the shutter speed with flash should not exceed the maximum shutter sync speed of a focal plane shutter (mostly on DSLR).
A focal plane shutter is two moving shutter curtains in front of the sensor (like two roller window blinds). At slower shutter speeds, one curtain opens (to expose the sensor to light from lens), and after a timed duration, the second curtain closes to cover it. That makes the full frame open to the lens for the duration, which allows the instantaneous flash to work. Flash is fast, and it can occur anywhere within this duration, but normal mode is early front curtain sync, and we also have later rear curtain sync for the blur trails with slow shutter speed.
But to implement the faster shutter speeds, the second curtain starts closing before the first is fully open, tracking together to create a narrow timed slit between the two curtains, the open slit moving across the frame (down the frame today, the shorter travel is faster). This narrow slit of course reduces exposure under it, to be a faster shutter speed equivalent. And a narrow slit cannot expose the full frame area from a flash of light.
A few descriptions of focal plane shutters are available. The Wikipedia link has an interesting section on history and improvements. A very interesting photograph (a classic) using a focal plane shutter is this 1912 picture by Jacques Henri Lartigue. The shutter slit's travel motion can cause distortion of a moving subject... different parts of the frame are exposed at slightly different times, when things may have moved. The shutter slit was moving down the picture there (upwards in the camera, but we see it after inversion). The spectators are leaning left because the 4x5 camera was being panned right, following the race car. Their feet were exposed earlier than their heads, at slightly different times, due to the moving slit, and the panned camera. The car was going faster yet, so it leans the other way. This classic picture is responsible for us imagining speeding wheels as being slanted ovals, at least in cartoons. Shutters are much faster today, but the fastest motion can still be an issue (see the FP flash picture of the grinding disk on next page). The narrow open slit can provide a very short (fast) exposure, but it takes much longer for this slit to travel across all of the frame (and the subject might be moving fast). The curtains cannot open or close instantly.
Shutter speeds not exceeding than maximum sync speed (a longer shutter duration) : The shutter can be left open for any long exposure, minutes even if using Bulb shutter setting. (Here comes the tricky part). But the shortest time that the shutter can be actually fully open for flash (none of the full frame is blocked by either curtain) is near zero... just enough for the flash - meaning when the second curtain starts closing immediately just as the first curtain reaches fully open. The flash can be fired during that instant that both curtains were fully open (so that the flash can expose the full frame). This zero time fully open case makes the exposure duration be the opening curtain travel time (to be evened out by the closing curtain travel time, making the exposure sum be one curtains time). This also makes one curtains travel time be equal to the Maximum Sync Speed by definition (becomes obvious after you think about it... this shortest travel duration is the maximum shutter speed that still has an instant of fully open shutter). So faster moving shutter mechanisms is what also creates faster Maximum Sync Speeds, being the maximum speed that the shutter curtain can be fully opened for flash. Modern shutters travel the short dimension vertically down (or up) the frame to be faster than travel across the longer dimension of the frame.
However, the total shutter travel time is the travel of the first curtain opening, and then the second curtain closing, or at minimum 2x the travel time of one curtain. Note this is just shutter motion, and NOT exposure time. HSS flash has to stay on illuminating for that total time (the shorter 2nd curtain delay to make the narrow slit of a fast shutter speed reduces that sum only slightly). But (for faster shutter speeds when it becomes more significant), the leading edge is exposed earlier in time than is the trailing edge, which with fast shutter speeds can distort motion captured during that time (not distorted by the instantaneous flash, but by the continuous ambient).
Shutter speeds faster than maximum sync speed (shorter shutter duration) : Faster shutter speeds start closing the second curtain before the first curtain is fully open.. The rear curtain closing is following behind the front curtain opening. The frame is never fully open. This exposes only a narrow open slit width (i.e., the timing between the two curtains is the exposure time, which is: duration = slit width / velocity). Flash cannot work then, since the frame is never fully open to pass it. The overall two curtain travel movement takes longer than the exposure from the narrow slit, but the actual exposure time is the time the open slit is over any area (the overall travel movement is always a constant travel time). The point is, the same remarkable precision mechanism moves both curtains, creating and moving a virtual narrow slit of opening across the frame at a fixed speed, to provide precise fast exposures under that slit. Regardless of shutter speed, the geared motor travel speed of the two curtains is always at one fixed speed, which can be made very accurate. This means however, that the exposures of different parts of the frame under this moving slit does occur at slightly different times (as the open slit moves across), but the actual exposure under the width of this moving open slit can be very fast, and very accurate.
The downside for flash is at these faster shutter speeds, the shutter opening (of light to the sensor) is only under this narrow slit. An instantaneous flash then only illuminates what is under that narrow slit at that instant, and the rest of the frame is not exposed to the flash. We can only use flash at shutter speeds slow enough that the entire frame is open at the same time, to let the fast flash expose the full frame area. This can be a problem (for flash in bright ambient like sunlight, but not much issue indoors), since the camera with a focal plane shutter normally can not let us set shutter speeds faster than maximum sync speed.
The flash is relatively nearly instantaneous, but it has to be synced to trigger when the shutter is fully open. For focal plane shutters, the fastest shutter speed at which the full frame is all fully open at any one instant (to allow instantaneous flash to fully cover the entire frame), is called the Maximum Sync Speed. The Nikon spec charts in rear of camera manuals call it Maximum Flash Sync Speed, and I often tend to say Maximum Shutter Sync Speed (all the same thing, sync of the flash to the fully open shutter). Flash sync is not possible if the focal plane shutter speed is faster than this limit (because some of the frame would be covered by the closing second rear shutter, and so unexposed, causing a dark band in picture). Some cameras use leaf shutters in each lens, or electronic shutters in some camera sensor chips, and these can sync faster than focal plane shutters (because there is always a fully open instant).
Until up into the 1980s, the fastest focal plane shutter maximum sync speed was 1/60 or 1/80 second. Today, shutters are faster, and 1/200 second (5 ms) is common today. A few models can do 1/250 second (4 ms) or 1/320 second (3.125 ms). Full frame cameras must travel half again farther than DX, but the vertical shutter travel today covers 1/3 less distance than horizontal. FX covering 24 mm in 3.125 ms travels 7.7 meters/second. DX covering 16 mm in 5 ms travels 3.2 meters/second. There is no Minimum shutter speed - any slower shutter speed can always be used with flash (however Nikon cameras probably arbitrarily force a 1/60 second Minimum Shutter Speed With Flash in camera A or P or Auto modes, E2 menu on some models. Slow Sync or Rear Curtain Sync will bypass that limit to allow the slow metered shutter speed.)
The Nikon specs clearly say Maximum Sync is still 1/250 second on their faster shutters, but some of these models provide a mode to allow 1/320 second sync, warning it might suffer reduced range of the flash. My guess is this is not about sync, but means the fast shutter could cut off slow flashes, like some large studio flash at low power, or speedlights at maximum power (but any speedlight power level lower than maximum should not be any issue). The speedlight T.1 flash duration at maximum will about 3x longer than T.5 (typically around 1/300 second). Maximum sync speed is normally not a speed problem (at normal ISO and apertures, when the ambient is not contributing much), because the flash is faster, so maximum sync speed is perfectly fine. The speed concern is about daylight level bright ambient, unable to use faster shutter speed.
Even in camera Manual mode, the camera will not allow you to choose these fast shutter speeds if it knows the flash is present (if flash is on the hot shoe for communication). But a PC sync cord will not know any flash is present, and this example above used a PC sync cord (flash not in hot shoe), to prevent communication, to fool the camera into allowing any shutter speed. There is of course no reason to do that, other than to show the problem here. A DX crop exposed on a full frame camera may crop off a small dark band, so it might get away with a slightly faster shutter speed than if shooting full frame.
Historically in the distant past, the only way to get around this sync problem with a faster shutter was to use a special longer burning flashbulb (named FP sync) that stays fully illuminated for the full shutter travel time - more like continuous light for the shutter duration. Auto FP flash is one such similar solution today (next page).
The shutter speed limit of Maximum Sync Speed is really no big deal for flash indoors. The flash is fast itself, so in dim ambient light (where we need flash), it matters less what shutter speed is, the flash is faster. The dim ambient is too dim to blur motion seen by a slow shutter, and indoors (if low ISO), we can open the aperture as wide as we wish without concern about overexposing the ambient. But this maximum sync speed becomes a much bigger problem when using fill flash in bright sun (because overwhelming bright sun cannot be ignored). The speedlight is extremely fast in dim light with no ambient to blur anything, but the bright sun is continuous, not fast.
The Sunny 16 Rule says at ISO 200, exposure in bright sunlight is 1/200 second at f/16. Bright sun does not vary, so this is the norm we expect. And we pretty much have to correctly expose the daylight scene - we cannot ignore the sun like we can ignore dim indoor illumination. Continuous light (like sunlight) has no shutter sync requirement, but when we add a flash for fill, now there is one. The topic of fill flash in sunlight is covered in Part 4.
With flash in sunlight, we cannot use faster equivalent exposures, like 1/400 second at f/11, or 1/800 at f/8 - because the focal plane shutter's maximum sync speed is 1/200 or 1/250. That means, if using flash in bright sun at ISO 200, the camera exposure WILL NECESSARILY BE around 1/200 second at f/16 (Sunny 16, but cannot exceed sync speed). The f/16 requires a lot of flash power, but otherwise this works fine, unless you just craved to use f/2.8 out there. Camera P mode knows all about this, and has ability to set both shutter speed AND aperture, and so it is a good choice for fill flash in bright sun. But if you use camera A mode, and set f/4 out there without thinking, the camera will just fuss warning HI at you, until you set near f/16 so it can work in bright sun (at the limited sync speed).
Wishful thinking, but our dream is that if somehow we could increase shutter speed, we could open the aperture, for reduced depth of field, or to an equivalent exposure for daylight which lets the regular flash mode work at lower power level, without affecting either exposure. Or, a faster shutter speed could help the flash to "overpower" the sun, reducing the sun without affecting the flash - if we could, but we cannot (in these bright sun cases.). We are up against the maximum shutter sync speed wall. FWIW, using lower ISO, or using a Neutral Density filter, can allow a wider aperture in that case, but that's all it does (maximum shutter sync speed is still enforced). But these affect both the flash and the sun equally, so these do not change the balance between flash power and sun. We can only use more flash power to affect the balance, and then exposing the higher flash power might end up at f/32, which does decrease the sun that needs f/16, but we still need Maximum shutter speed sync. Flash in bright sun is a special case.
Flash sync speed is really only an issue in bright sun. It is unimportant in dimmer light, where we need flash. So we might as well get used to Maximum Shutter Sync Speed, because it has always been this way, since the first Leica 35 mm camera in 1925 (focal plane shutter). The first Nikon F (1959) sync'd flash at 1/60 second. The situation is better today than it has ever been before.
One workaround in sunlight is to underexpose the sunlight ambient by a couple of stops, so the flash can stop the motion without the ambient blurring it again. Flash becomes main light instead of fill. Sports action likes this, it highlights the subject better anyway. The speedlight is faster than the shutter, and reducing the sun's exposure causes less blurring of action. And speedlights are the way High Speed Photography is done.
A couple of confusion factors which do not change the basics:
There are mechanical focal plane shutters, and there are electronic shutters.
If we had the electronic CCD chip shutter (no mechanical motion involved), then these can sync flash at the fastest shutter speed. The CCD sensor is simply electronically enabled and disabled, which is also used as a shutter, even like at 1/4000 second. However, the faster shutter speeds could be fast enough to truncate the longer flash duration, reducing effective light from the full power flash (but speedlights become even much faster than the shutter, at lower power levels, in more dim light so the ambient won't blur anything).
The Nikon D40, D50, D70 cameras did use the CCD electronic shutter, but if these models recognize the flash is present, they still limit the maximum shutter sync speed to 1/500 second. However, with these electronic shutters, if you can break the communication to fool the camera so that it does not recognize a flash is present (for example, if you use a PC sync cord), then the camera firmware will not limit shutter speed at all. You can use any faster shutter second then (only on these models with this electronic shutter). The problem is not about syncing the electronic flash, the focal plane problem is about opening the shutter wide, to get the flash through it.
Note that just because these CCD situations might sync the flash with a fast electronic shutter, it does not mean there is no effect on the flash. A fast shutter surely does truncate the longer full power flash duration and reduce the useful flash exposure. However, lower speedlight power levels are much faster than any shutter, not likely affected.
It is difficult to define how to measure actual flash duration. Flash is a fast pulse which then decays relatively slowly, trailing off to zero, and it is hard to agree when it effectively finishes. When does the gradual trail-off stop being effective? (10%? 5%? 1%?) The standard method for published flash duration is called t.5, which measures the time that the flash is stronger than 50% of its peak intensity. This is an engineering convention, convenient for engineers, but is not a photographer's convention. Because 50% intensity is only one stop down, and is still rather bright, so photographically, this means the t.5 specification number is actually about three times faster than we realistically see in our pictures - the useful 10% limit of the flash (called t.1, measuring 90% of the power) is about 3 times longer than the conventional 50% spec number for duration - BUT the definition of this limit of usability is really difficult to specify. 10% is not much to a photographer, less than our smallest setting change of 1/3 stop, but 50% is a full stop.
Camera speedlight units are a big exception, and are "different", being extremely fast at lower power levels (the name "speedlight"). Their 1/32 power level may have an actual duration of 1/20,000 second, because speedights reduce power by truncating the flash duration. This curve of the truncated pulse has steep sides going up and down, and so t.5 really hardly applies, EXCEPT when at their full power level when t.5 and t.1 are probably the same (just saying, the truncated lower power speed specs are more unequivocally precise regarding what actually happens). But otherwise, speaking of the speedlight full power level (and studio lights), then in general, a t.1 time is more meaningful for photographers, which duration is mathematically about three times slower than the t.5 stated by the spec, and so the full power duration may approach normal maximum shutter sync speeds closer than we may realize. If that may be not be clear, please see the actual description at High Speed Flash.
The subject of this page. Flash is limited by a Maximum Shutter Sync Speed, typically around 1/200 second maximum. Start at the top of this page. :)
These focal plane cameras often offer a FP High Speed Sync mode (Nikon calls it Auto FP, Canon calls it HSS), which in fact does allow any shutter speed faster than the maximum sync speed possible. This is actually a feature in the flash unit, which drastically changes the flash behavior, but the camera must select it, and trigger it slightly earlier (it is a "system", flash and camera, more at top end).
Auto FP mode is not actually the same thing as "flash", but Auto FP mode does offer one way to increase shutter speed with flash, for this bright sun case. (next)
Auto FP flash is a big subject, continued on next page.