How does the 5D shutter speed work?

[dave_hickey]

<p>Hi all,</p>

<p>I have bought a Speedlight 580exII and have read on-line, that the sync speed to my 5D is 1/200 second, because the shutter curtains cannot travel across the full frame sensor on my 5D any quicker than that - I have also read that this speed is more like 1/160 second in reality.</p>

<p>But my two questions are these: if the above maximum speed for the sake of argument is 1/200 second, then how come I can shoot without a flash at 1/1000 second? How can it give me a faster shutter speed than the shutter can physically provide?</p>

<p>And, if this means the data is simply read from the sensor at the 1/1000 second speed and isn't therefore determined by the actual speed of the shutter curtains moving other than having to wait until they are fully open, then why can't this speed be available when I use the flash, bearing in mind the flash can operate at 1/1000 of a second and apparently, so can the sensor?</p>

<p>I am totally confused - as the reason I may wish to do this faster speed, means I can have the lens wide open and control the ambient light with speed, rather than having the lens stopped down which then gives me less control over the depth of field.</p>

<p>Help...</p>

<p>Thanks for any advice.</p>

<p>Dave</p>

[zml]

<p>The shutter has to be fully open when the flash fires or you'll get a partially unexposed frame. You may want to read for instance <a href="http://en.wikipedia.org/wiki/Focal_plane_shutter">this</a> and <a href="http://www.mir.com.my/rb/photography/fototech/focalplane/index.htm">this</a> and above all <a href="http://photonotes.org/articles/eos-flash/">this</a> for more on that topic.</p>

[hjoseph7]

<p><em>"if the above maximum speed for the sake of argument is 1/200 second, then how come I can shoot without a flash at 1/1000 second?if the above maximum speed for the sake of argument is 1/200 second, then how come I can shoot without a flash at 1/1000 second?"</em><br>

You have to set your flash(580ex II) to "HP" high-speed flash, then you can shoot at higher speeds, check your Canon 580ex II manual for more info.</p>

[alec_myers]

<p>Very briefly...<br>

The two shutter curtains take a finite time to traverse the sensor plane, and the exposure takes place in the time gap between the first curtain uncovering and the second re-covering the sensor.</p>

<p>At high shutter speeds - higher than the maximum flash sync speed - the second curtain is released before the first curtain has finished its travel. The two curtains move together, and the exposure is in the form of a moving slit of light that exposes each part of the sensor for the same duration, but never all of the sensor at any one instant.</p>

<p>Since at these shutter speeds the whole sensor isn't uncovered at one point in time you can't use a single flash to expose all of it.</p>

<p>If it's any help, bear in mind that calling it a "shutter speed" is misleading. The shutter curtains always travel at the same speed which is as fast as the camera designer can make them go. It's the time interval between the two curtains that changes with "shutter speed", so probably calling it shutter duration makes more sense.</p>

<p> </p>

[michael_ziegler2]

<p>Alec, "so probably calling it shutter duration makes more sense."<br>

I believe thats why camera manufacturers call it Tv or "time value"</p>

[edward_rapier]

<p>In addition to Alec's comments - in regular flash mode where the shutter speed is below the sync speed, the flash waits until the entire sensor is exposed and then emits one powerful burst of light. In FP mode the flash emits a series of pulses of light, like a strobe light, as the slit moves across the sensor. The series of pulses are too quick for the human eye to distinguish and look like one burst. The sensor is exposed one section at a time as the flash pulses while the gap between the curtains moves across the sensor. </p>

[alec_myers]

<p>"I believe thats why camera manufacturers call it Tv or "time value"" - Good point! I hadn't thought of that. Pity it hasn't caught on in general parlance though.</p>

[davebell]

<p>To answer the OP's question - it works very well indeed. No problems reported.</p>

[paul_b._davis1]

<p>One other important point--When you're operating in High-Speed Sync mode, the flash becomes less powerful and thus has a lower effective range.</p>

<p>Usually this means that you'll have about 1/3 of the power of normal flash mode.</p>

<p>-Paul B. Davis</p>

[alec_myers]

<p>"Usually this means that you'll have about 1/3 of the power of normal flash mode."</p>

<p>I believe that, rather than being a fixed ratio of normal mode, the power drop is heavily dependent on the shutter duration. As the shutter duration decreases, a smaller and smaller fraction of the light makes it through the narrowing slit and more gets wasted falling on the back of the shutter curtains. Hence an increase in the percieved power drop.</p>

[paul_b._davis1]

<p>You're correct Alec, it's not a fixed ratio, but the rule of thumb that I'm familiar with is "usually 1/3." The canon manual for the 550 EX II says that it's inversely proportional to shutter speed: More speed = less power. That's probably a more reasonable description. :)</p>

<p>The important thing to know about High-Speed Sync, though, is that the power drop is not <em>perceived</em> --It's real. You've accurately described most of the mechanics, but there's a piece of key information missing:</p>

<p>When a modern flash fires, it sends a single pulse of light for a <strong>very</strong> short period of time. Very short, in this case, means something faster than 1/1000th of a second.</p>

<p>When the shutter is not able to fully open and travels across the medium as a slit (to work at higher shutter speeds), the result of this short pulse of light would be a single, well-exposed band of light at the position the slit was at when the pulse fired.</p>

<p>The solution for this problem is two-fold: </p>

<p>First, declare a maximum speed on the body at which the shutter is fully opened and flash can be used normally. This is called the maximum sync speed, or the X-sync speed. The flash is allowed to operate normally at or below this speed, since the medium is fully and evenly exposed to the world during these timings.</p>

<p>Second, for speeds where the shutter is never fully opened, we fall back on an age old photographer's trick: Make your bulb burn longer. If the bulb is burning for the entire time that the shutter opening traverses the medium, then the medium receives an equal amount of flash from top to bottom.</p>

<p>The trade off with modern flash equipment, though, is that we're no longer working with single use bulbs that are designed for a long burn; we're still using that single-pulse flash that's designed to go off once, and very quickly at that. The workaround then, is to pulse the flash very rapidly (and evenly), for the duration of the shutter movement.</p>

<p>This rapid pulsing means your flash is actually firing 50,000 times a second in High-Speed Sync mode, which ultimately nets out to a loss of overall power in your output.</p>

<p>-Paul B. Davis</p>

[alec_myers]

<p>True... but Paul... I think y<em>ou</em> missed a key piece of information ... it's not the fact that the tube is firing at 50kHz that causes the power loss - it's the fact that because it has to fire for the entire duration of the curtain travel, only a fraction of the light hits the sensor.</p>

<p>I'm not sure how a "usually 1/3" rule of thumb can usefully apply when the flash useful-power drop between a shutter speed of say 1/500 and 1/8000 will change by a factor of 16 ... so "1/3" should really be more like "anywhere down to 1/20 or even less."</p>

[paul_b._davis1]

<p><em>I'm not sure how a "usually 1/3" rule of thumb can usefully apply when the flash useful-power drop between a shutter speed of say 1/500 and 1/8000 will change by a factor of 16 ... so "1/3" should really be more like "anywhere down to 1/20 or even less."</em></p>

<p>Agreed--It's a gross oversimplification. Let's drop it.</p>

<p><em>...it's not the fact that the tube is firing at 50kHz that causes the power loss - it's the fact that because it has to fire for the entire duration of the curtain travel, only a fraction of the light hits the sensor.</em><br>

I find it hard to believe that a high-voltage system designed for the sub-millisecond "pop" of a flash tube can maintain normal voltage levels over a sustained burst. I've poked around for a better explanation of the power loss, but I haven't been able to find anything more specific than this, which seems to imply that there <em>is</em> a power reduction:</p>

<p><a href="http://web.canon.jp/imaging/flashwork/ettl2/high/index.html">http://web.canon.jp/imaging/flashwork/ettl2/high/index.html</a></p>

<p>The other reason I suspect that there's a power drop involved is the loss of effective <em>range</em> . If something can be illuminated at a maximum of 10 feet with normal power, but only at 5 feet with high-speed power, what is the cause of the loss of range?</p>

<p>Is it really just that there's less exposure time overall (so farther light doesn't register at that speed), or is there truly a loss of power? I don't think we can reliably reach a conclusion based on the evidence that we have here.</p>

<p>-Paul B. Davis</p>

[dave_hickey]

<p>Hi Guys and thanks.<br>

What a full and interesting set of replies to my question.<br>

So basically, if I set my flash to high speed sync mode for anything less than 1/200 sec, the flash will pulse for that shutter duration or timer value, but the down side is that I will lose approximately 2/3 of the power in the flash by doing so, which I assume I can then compensate for by moving the flash in closer to the subject, to increase the apparent size and power of the flash in respect to the subject.<br>

Got it - cheers.<br>

Dave Hickey</p>