• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

The "high-speed sync" mode of Canon Speedlite flash units

Doug Kerr

Well-known member
Today I would like to discuss the mode generally described as "high-speed sync" available in most Canon Speedlite flash units.

Background - Focal plane shutters

All Canon EOS digital SLR camera bodies use focal plane shutters. In such a shutter, when idle, the sensor is blocked by an opaque "curtain" (in particular the one known as the "first curtain).
It is often spoken of as the "front curtain".​

Consider first a fairly long exposure (perhaps 1/60 sec).
When the exposure commences, the first curtain moves quickly downward, exposing the sensor.
The curtain is generally made of two or three overlapping blades; it collapses when being withdrawn so that such a large "garage" is not needed for it at the bottom of the shutter.​
After the desired exposure time, a second curtain (sometimes, "rear curtain") descends to cover the sensor, terminating the exposure.
This curtain is also generally made of two or three overlapping blades, again so that such a large "garage" is not needed for it at the top of the shutter. It unfolds as it descends.​
The curtains take a finite time to uncover and then cover the sensor, typically on the order of 3 ms for EOS digital cameras. If we consider a desired exposure time of, for example, 1/500 second (2 ms), the second curtain will begin its descent before the first curtain has completely cleared the sensor. If we now consider an exposure time of 1/10,000 sec (0.1 ms), the second curtain will begin its descent when the leading edge of the first curtain has only completed about 1/30 of its traverse across the sensor.

The maneuver is often described as a "slit" traveling across the sensor. Nevertheless (in this last example), this results in each spot on the sensor being exposed for 0.1 ms, just as we wish.

Since the scene is, in a sense, being "scanned" over a period of perhaps 3 ms, if the subject is moving laterally, the image will be "skewed", in the way cartoonists draw bicycles (the wheels leaning forward) to make it seem as if they are traveling rapidly.

In fact, the reason that the shutter curtains travel downward, rather than upward, is so that this "skewing" indeed follows the cartoonist's convention: the top of the wheels leaning forward (recall that the image is inverted with respect to the scene itself; the "slit" travels upward across the scene.)​

Flash photography and "X-sync"

The luminous output pulse of a photoflash unit is very short (1 ms of less for camera-mounted units). For the illumination of the entire scene to be effective, the flash pulse must occur while the shutter has completely cleared the sensor. With a focal-plane shutter, the usual approach is to trigger the flash unit as soon as the first curtain is completely open (assuming that happens at all).

This timing is called "X-sync". The term goes back to the use of photoflash lamps. [For the next little while, think of non-focal-plane shutters.] They took a short while to "come up to full luminous output" as the fire spread across their incendiary charge. To best deal with this, shutters that had flash synchronizing contacts included a small timer. When the shutter lever was tripped, the flash trigger contact closed promptly, but the timer delayed the opening of the shutter proper to wait for the "peak" of the flash output. Different "classes" of flash lamp had different general ranges of rise time, and shutters often had separate settings for each (controlling the exact delay of the timer), indicated by letters representing the time class. Only the "medium" class ("M"), which called for a delay of perhaps 20 ms, was in general use with professional and advanced amateur photographers.

Indeed, for longer shutter speeds, it was optimum to delay the shutter opening less than the time to peak to take advantage of more of the overall luminous output of the flash lamp, and sophisticated shutters had a "program" for controlling the shutter delay based on shutter speed setting.​

When electronic flash came into use, it was realized that existing shutters, which delayed opening until typically 20 milliseconds after the flash unit had been triggered ("M sync"), would miss the flash event entirely. Thus, new shutters were introduced, in which the delay selector has a new setting, "X", evocative of no delay in shutter opening.

There was in fact slightly more to it: in the "X" mode the shutter trigger contact had to be delayed so it did not close until the shutter blades were actually open.​

Today, the term "X-sync" is not only used to refer to a certain timing of the flash triggering but as well to the entire basic concept of dealing with a nearly-instantaneous flash output pulse.

Back to focal plane shutters

With a focal-plane shutter, for shorter speeds greater than a certain value, the curtains do not at any instant clear the entire sensor. Accordingly, the is no time at which it is appropriate to fire an electronic flash so that the entire scene will receive its benefit in the exposure. As a consequence, EOS cameras preclude setting the shutter speed faster than a certain predetermined value when a flash unit is in use (and the camera knows it).

Because of the use of the term "X-sync" to refer to the entire concept of using an essentially instantaneous flash output pulse, this shutter speed is known as the "X-sync limit". This is of course not "the shutter speed above which X-sync is inappropriate"; it is the speed above which the kind of flash operation that involves X-sync becomes inappropriate.

Focal-plane flash

Often, it would be desirable to use shutter speeds above the X-sync limit with flash.

Reasons include, in a situation with substantial ambient light:

• to avoid motion blur of the parts of the scene exposed by the ambient light.
• to avoid overexposure of the parts of the scene exposed by the ambient light.

To accommodate this, the Canon system includes, in essentially all Speedlite flash units (with matching provisions in all EOS camera bodies) a mode principally labeled "high-speed sync". The term is (as I will discuss shortly) not apt. A secondary label, "FP (focal plane) flash", is much more apt.

In this mode, the luminous output pulse from the flash unit is prolonged, lasting perhaps as long as 10 ms or so, and being approximately of the same potency over that period. As a result, as the scene is "scanned" by the "moving slit effect" of the focal-plane shutter, all parts are illuminated.

It is popular to describe the luminous output of the flash unit in this mode as being a "rapid series of very short luminous pulses". Actually, based on the best information I have (I will discuss that at the end), the current to the flash tube is indeed delivered in a series of short pulses of modest duty cycle (so as to "meter out" the energy in the storage capacitor over the overall burst time). However, the tube continues to glow for a while after the end of each current pulse (after all, the gas in it is "afire"). Thus what we have is a continuous luminous output but one which does vary up and down a bit at the current pulse rate (like DC voltage with "ripple").

In the one situation for which I have definite information (a Speedlite 380EX flash unit), the current pulse rate is about 40 kHz.

Synchronization

Although the most common label for this mode is "high-speed sync", its pivotal feature is not a different type of synchronization, but rather a gigantic difference in the nature of the luminous output pulse (its duration now being perhaps ten times what it is in the normal mode).

Nevertheless, a different approach to synchronization must be employed here (which some feel justifies the "label").

With the normal flash mode, X-sync is used, in which the flash unit is triggered as soon as the first curtain of the shutter is completely open.

In the "high-speed sync" mode, FP-sync is used (although we never hear that mentioned), in which the flash unit is triggered just before the first shutter curtain begins to open.

X-sync above the X-sync limit

If we have a flash unit connected to an EOS camera via the PC sync jack, or even a non-Canon flash unit in the shoe, the camera does not know that a flash unit is present, and thus it does not impose a limit on the shutter speed that can be set.

However, above the X-sync limit, the camera does not decline to provide a flash sync contact closure (at least in modern EOS bodies). It is consistently on an X-sync basis; that is, the contact closes (as near as I can determine), shortly after the first shutter curtain is fully open.

Of course if we draw upon this with shutter speeds very much above the "X-sync limit" for the camera (which, as a matter of fact, is chosen based on the time required for the shutter blades to complete their transit across the frame), we will get an exposure in which (at least with respect to illumination by the flash) only the upper portion of the resulting image is exposed, the height of the exposed zone decreasing as the shutter speed increases, becoming zero above a certain point (because the flash trigger does not occur precisely as the first curtain clears the fame, but an intentionally short time later, for the sake of "margin").



Time for breakfast now. It's Sunday, so that means three kinds of fruit, scrambled eggs, bacon, hashed brown potatoes, 12 IU of Humalog insulin, intelligence briefing from the New York Times and elsewhere, political commentary on TV, and of course a discussion of many aspects of life with the fabulous Carla Red Fox. (She will have three more great-grandchildren by December.)

Best regards,

Doug
 
Last edited:

Doug Kerr

Well-known member
There is an interesting sidelight on the matter of the timer used in shutters to provide flash sync delay (as for "M" class flash lamps).

A while after this type of "synchro" shutter came into use, a manufacturer realized that a desirable feature, the "self timer", could be implemented at little additional cost by "hijacking" the flash sync timer, changing its configuration so that the delay would become much greater.

Of course, having done that, one could not have flash sync delay, so if flash were used in a "self-portrait", all that was available was "simple" sync (in which the flash was triggered just before shutter operation) or X-sync (in which it was triggered when the shutter blades were fully open. That of course, for shorter shutter speeds, resulted in a decline in flash exposure, but one just lived with that.

Not surprisingly, with this scheme, the self-timer mode was usually engaged by a position on the sync selector. It was normally labeled "V" (the earliest shutters to offer this were made in Germany; "V" stands for vorlauf, literally "before run", or more idiomatically, "pre-run"). Thus, the sync selector typically had the positions "M", "X", and "V".

Best regards,

Doug
 

Doug Kerr

Well-known member
The most definitive information I have on the details of the "high-speed sync" mode for Canon Speedlite flash units is out-of-date, but is very valuable nevertheless.

It comes from measurements made in 1997 by Jeremy Stein of Albuquerque, reported on the photo information site (frozen as of about 1999) of Toomas Tamm, then in Helsinki (now with the Tallinn University of Technology, in Estonia).

Stein took oscillograph traces of the luminous output of a Canon Speedlite 380EX flash unit operating on a Canon EOS Elan II body (film).

The results can be found here:

http://www.chem.helsinki.fi/~toomas/photo/flash-discharge/hispeed.html

The same series includes a page of information on the pulse width of the Canon Speedlite 430EZ flash unit as a function of different (manually-set) output levels (also done in 1997):

http://www.chem.helsinki.fi/~toomas/photo/flash-discharge/regular.html

A third page shows the effect of firing the flash unit with different levels of charge in the storage capacitor (for both 430EZ and 380EX units):

http://www.chem.helsinki.fi/~toomas/photo/flash-discharge/redwait.html

I suspect that this information is broadly applicable to contemporary Speedlite flash units, but I have no detailed confirmation of that.

Best regards,

Doug
 
Top