• 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!

Helpful Hints/D.I.Y.: High Speed Flash Sync 1/8000 sec f/2.8 Mid Day

Doug Kerr

Well-known member
For what it's worth, with a Canon EOS 40D and Canon Speedlite 580EX II, with the system in the FP mode (called "high-speed sync"), the sync timing at the flash sync PC connector on the 40D is apparently X-sync (trigger closure when the first curtain is completely open).

This was determined by making the 580EX II not contribute substantially to exposure (set to Manual, 1/128 output, and the head pointed away from the test scene) and firing a Vivitar 285HV from the PC connector on the 40D. The partial exposure of the frame, for fast shutter speeds, is the same with the system in FP mode or not.

In older EOS cameras, this output came directly from an electromechanical contact on the shutter, so it is not surprising that it has a consistent timing with respect to shutter operation.​

Note that in the Canon system, with "older" EOS cameras, the system must be put in FP mode by a setting on the dedicated flash unit (e.g., 580EX II) itself, but the camera is of course aware of it. In more modern cameras, including the 40D, the FP setting can be made from the camera, but the intimation is that in that case we make a setting on the flash from the camera, which then advises the camera.

No surprise here, but just another data point on the misthaufen.

The Canon dedicated flash units do not provide a sync output to electrically fire another external flash, so I cannot test the Deal deal.

********

As a matter of gratuitous historical interest, note that the "X" in "X-sync" means "zero shutter opening delay after sync closure". It goes back to the days when, to accommodate the lag in the peaking of flash lamp output, the shutter opening was delayed from sync closure (with fancier shutters, by one of two time delays for different classes of flash lamp - but often just the "M" - medium - delay, suitable for the most common "consumer" class).

When electronic flash was introduced, no such delay was needed, and a setting of "X" on the delay selector meant "delay: disable".

In fact, it was then important that the sync closure was itself delayed until the shutter was fully open, often a new wrinkle in the shutter mechanism design.

Then, the shutters often had a selector lever for "X" vs. "M" delay.

Often that delay timer was hijacked for the "self-timer function", which then was a third setting of the lever - "V" (for vorlauf, German for "pre-run"). In that case the sync timing was "X" whether that's what you wanted or not.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Bart,

The only explanation I can come up with is that the slower declining tail of the discharge curve is being used for the actual exposure.
I think you are right on the money there (see my earlier post today).

Maybe using a sync cable of adequate length . . .
Since the sync output is a low impedance "closure", it is unlikely that the length of the sync cable would influence the trigger instant.

. . . combined with a slightly (too) fast shutter sync introduces just enough delay to miss the discharge peak?

Not sure I follow. Any delay in firing the external flash would go in the other direction - a later firing so that we would see more effect from the main peak.

Note that, in any case, I believe that the typical trigger instant for FP sync occurs a little before the first curtain has begun to actually clear the frame (maybe when the curtain is released).

Best regards,

Doug
 

charlie chipman

New member
Indeed. You do not mention here the output level of the Acute pack, but my guess (from the narrow gradient on the boundary in the image) is that is was set fairly low (thus a short burst).

You are correct, the acute pack was set close to its minimum power, same as in second series.


I think you did not mention here how the Acute pack was being triggered - from the PC connector on D700, or from the PC connector on the SB-800. Based on the overall issue here, I am assuming the latter.

It would not be surprising if, with the system in the FP mode that the SB-800 sync output would be on an FP-sync basis.

But it would not be out of the question that, with the system in the FP mode, the sync output from the D700 (via its PC connector) might be on an FP-sync basis as well. (Recall it is still the camera that determines the proper firing instant for FP flash operation - however it sends that to a dedicated flash.)

The D700 has the same options as the D300 for FP sync at either 1/250 and 1/320, for these tests I had it set to 1/320. I tried a couple shots at the FP 1/250 setting though I did not notice a difference.

For my tests I used both the PC connector on the D700 as well as the PC connector on the SB-800 for each series, the pictures turned out the same from either PC connector. The pictures I displayed here are the set from the D700's sync connector.

I think it is safe to say that when the system is in FP mode the sync output from both the D700 as well as the SB-800 are on an FP-sync basis.



I think you are both right in that we are seeing the declining tail of the flash duration. Bart I do have a couple of very long sync cables, I will see if I can find the right adapters to set them up and see if anything changes.


For what it's worth I tried the same test with my Elinchrom Ranger RX speed AS pack with the A head and the results were not the same, very minimal light in the exposure at any setting if I remember correctly. The Ranger has a very short flash duration (see below) so I suspect this is why. I can post the results if you wish to see them.


Elinchrom Ranger:
Flash duration @ t 0,5; 1/2 A heads
A output: 1/2300s / 1/4310s ~ B output: 1/5120s
 

Doug Kerr

Well-known member
Hi, Charlie,

For my tests I used both the PC connector on the D700 as well as the PC connector on the SB-800 for each series, the pictures turned out the same from either PC connector. The pictures I displayed here are the set from the D700's sync connector.

I think it is safe to say that when the system is in FP mode the sync output from both the D700 as well as the SB-800 are on an FP-sync basis.
Certainly seems that way. And not at all surprising. (Note than on the Canon EOS cameras, with the "system" in FP mode, the output from the PC connector on the body is still X-sync. That's not surprising either.)

I think you are both right in that we are seeing the declining tail of the flash duration. Bart I do have a couple of very long sync cables, I will see if I can find the right adapters to set them up and see if anything changes.
It will be interesting to see what happens. I cannot imagine any phenomenon by which the timing would be affected (in any way we could see) by change in the sync cable length. But until ya know, ya don't know!

For what it's worth I tried the same test with my Elinchrom Ranger RX speed AS pack with the A head and the results were not the same, very minimal light in the exposure at any setting if I remember correctly. The Ranger has a very short flash duration (see below) so I suspect this is why.

And a short tail may come with that.

Thanks again for your work on this.

Best regards,

Doug
 

Doug Kerr

Well-known member
I suspect that a long burst tail may be fairly common with "studio-type" flash units.

Modern on-camera flash units normally have a gate turn-off thyristor or equivalent that completely shuts off the current flow to the flash tube when the desired (or needed) pulse width has occurred.

But, although some studio units may in fact have such, I know that many don't. They of course only have manual control of luminous output, which is typically done by:

• Varying the voltage to which the energy storage capacitor is charged, thus varying the energy stored, or

• Switching capacitors into or out of the storage capacitor "bank", thus changing the capacitance and thus the amount of energy stored for any given charge voltage,

or both.

We can assume that voltage variation is involved with the Travelite 750 since the manual reminds us, when reducing the power setting with the unit lit, to manually fire the system (thus discharging from the storage capacitor its no-longer-appropriate higher voltage change, so the next use will have appropriate output).

The duration of the tail depends on various factors, including the shape of the E-I curve of the flash tube and the resistance of the discharge path.

I know that in some cases resistance is intentionality insinuated into the discharge path, reducing the peak output but affording the potential for a dramatically-lengthened tail.

Best regards,

Doug
 

charlie chipman

New member
P.S. The new Pocketwizard series of wireless flash sync units (MiniTT1 and FlexTT5) allows to precisely adjust the (pre-)trigger moment of the attached flash unit. On many camera bodies (Canon, and Nikon is being implemented) it allows to add another (or even two) regular shutter speed step(s) before the strobed Hi speed sync kicks in (thus adding more flash reach in broad daylight), called HyperSync. I suppose it could be used to delay sync as well, have yet to try that.


I knew they had a high speed sync function but was unaware of how and even if it actually worked. Have you had any success with that feature using studio lights as opposed to speed lights.

Will the pocketwizards allow the camera to go into FP sync mode if the speedlight is attached to the pocketwizard's hotshoe?
 

Doug Kerr

Well-known member
The normal "X-sync limit" of most cameras is not only determined by the length of time it takes the two FP shutter curtains to travel across the frame but also has some built-in margin to deal with the fact that the flash may not trigger exactly when desired (owing to variable delays both in the camera and in the flash circuity).

For example, if the shutter curtains traverse the frame in 0.0025 sec, then at a shutter speed of 1/400 sec, the frame will be open all at the same time (but just barely). Imagine a flash unit with an extremely short burst. If the (normal) flash burst peak occurs just as the frame is fully open, all is OK. And we could conceptually operate with a shutter speed up to 1/400 sec.

But if the flash burst peak is late by, say, 0.0005 sec, the second curtain will have traveled about 1/5 of the way across the frame, and that bottom 1/5 of the image will not have been illuminated.

To compensate for that, we would need to impose a speed limit of about 1/333 sec.

The problem is exacerbated if the flash burst has a significant duration - suppose its duration is 0.001 s (imagine a "rectangular" burst, with no rise or fall time), but its timing is perfect. For all of the frame to be "fully illuminated", the frame will have to be fully open for 0.001 s. With the curtain transit time we postulated above, that corresponds to a shutter speed of about 1/286 second.

If we again allow for the presumed 0.0005 sec trigger uncertainty, that corresponds to a shutter speed limit of 1/250 sec.

Thus, the manufacturer may recommend a more conservative (slower) shutter speed as the "X-sync" limit, and enforce it by the logic in the camera when a dedicated flash is aboard. In fact, the Canon EOS camera manuals often urge the user to observe an even more conservative speed limit in the case of non-Canon flash units (where trigger time uncertainty, burst width, and exposure to loathsome diseases may be greater): 1/250 second for compact flash units, 1/60 for studio-style flash units.

**********

The "HyperSync" feature offered by the Pocket Wizard people seems to be intended to "tune" the flash trigger instant so as to:

• For flash units with bursts shorter than the curtain travel time, make best use of the fully-open period of the shutter, thus attaining fastest best possible "X-sync limit".

• For flash units with bursts at least as long as the curtain travel time, allow operation at up to essentially any shutter speed.


Best regards,

Doug
 
I knew they had a high speed sync function but was unaware of how and even if it actually worked. Have you had any success with that feature using studio lights as opposed to speed lights.

The Pocketwizard's 'Hypersync' works fine with Canon's 'High Speed Sync control'. All that's necessary is for the entire image plane to be uncovered when the flash pops. Whether that is at the indicated maximum X-sync time or at a bit faster shutterspeed depends on the camera. On my 1Ds3 the official max X-sync speed is 1/250 sec., but with the FlexTT5 I can expose at 1/320th without problems, and at 1/400th only a narrow strip at the edge will be covered. The flash timing can be tweaked for that edge to be a first, second, or both curtain shadow. On cropped cameras one can sometimes gain a bit more from the max X-sync speed because the slit between the curtains can be smaller, and the flash can be timed to that exactly.

For studio flash units, things depend very much on the flash discharge pattern, but 1/320 sec. is no problem for my Multiblitz units, and the camera could possibly get away with 1/400th, but for studio flash I can't (and won't have to) tolerate any shading by the shutter. Of course the studio flash can't do the 'High Speed Sync Control' strobing.

For flashes with a very specific discharge pattern one may be able and time the units for their maximum power during the shutter time, and thus gain a bit of output power. In theory (I haven't tried it) the timing could be set for the 'tail' of the discharge pattern.

Will the pocketwizards allow the camera to go into FP sync mode if the speedlight is attached to the pocketwizard's hotshoe?

Yes, I can confirm from personal experience that on the FlexTT5 that will work, the MiniTT1 is probably no different in that respect. One can set the shutterspeed at which HSSC kicks-in in the TT5's firmware (in my case I set it to kick in on 1/400, or 1/500th should I need that).

Cheers,
Bart
 

Doug Kerr

Well-known member
Hi, Bart,

Primitive tests here suggest that on my EOS 40D, the time required for the shutter curtains to traverse the frame is about 2.47 ms (about 1/400 sec).

Using the 40D PC sync terminal to fire a Vivitar 285HV, the fastest shutter speed I can set and not get a black bar is 1/320 sec. (The X-sync limit enforced by the 40D when a Speedlite is in place and active is 1/250 sec.)

That test data is consistent with the notion that the X-sync closure occurs when the first curtain is a little bit beyond "clearing" the frame. And we could get that margin time back with the "HyperSync" time-diddling of the PocketWizard system, as you discuss.

But to get higher shutter speeds would require a flash with a burst that was substantially flat (lets say was between the T 0.5 points) over the curtain travel time.

All very interesting.

Best regards,

Doug
 

Doug Kerr

Well-known member
Hi, Bobby,

Well, as you can see, your "hint" about outdoor flash fill has stimulated quite a discussion of the surrounding technical issues.

My take on the situation right now is:

• With your "system" (D300 + SB-800) actually operating in the "FP" mode (with the shutter speed above the "X-sync" limit you have set), the sync terminal on the SB-800 fires an external flash unit with "FP sync". That means that the trigger occurs just before the shutter first curtain starts to open.

• Your Travelite 750 does not emit any kind of supernatural burst under this situation.

• The flash fill you are getting is from a "long tail" of the Travelite luminous burst.

When you get a chance to do some more test shots, I would suggest you include the following sequence. This would be done indoors, where there would be no substantial ambient illumination. Shoot a fairly-uniform colored wall.

• With the shutter speed at 1/250, set the Travelite flash to full output. Adjust the ISO sensitivity and aperture until a non-blown exposure is obtained across the wall.

• Set the shutter speed to 1/500 or thereabouts. Shoot again.

• If the exposure is "too dim" to see any effect, increase the ISO sensitivity or open the aperture until it is.

• Set the shutter speed to 1/2000 or thereabouts. Take a shot and, if necessary, increase the ISO sensitivity or open the aperture.

The interest, of course, is to get some idea of the illuminance on the wall during the exposure (and how uniform it is over the time of shutter curtain travel).

Best regards,

Doug
 

charlie chipman

New member
I think you are both right in that we are seeing the declining tail of the flash duration. Bart I do have a couple of very long sync cables, I will see if I can find the right adapters to set them up and see if anything changes.

Instead of using very long cables to attempt a slight delay in the timing of the flash I used the pocket wizard multi max's which have a trigger delay function.

The mere act of using a wireless trigger as opposed to a cable seems to delay the flash ever so briefly as I could see a slightly brighter exposure. When I started to apply the delay (.0001 - .0005sec if I remember correctly) the exposure was getting brighter, meaning I was catching the declining tail of the flash a bit earlier. Then moving the delay to .0010-.0020sec the exposure was brighter again but the shutter curtain had also started to appear in the frame.

I don't think a delay will be useful in any real world applications.
 

Doug Kerr

Well-known member
Hi, Charlie,

Instead of using very long cables to attempt a slight delay in the timing of the flash I used the pocket wizard multi max's which have a trigger delay function.

The mere act of using a wireless trigger as opposed to a cable seems to delay the flash ever so briefly as I could see a slightly brighter exposure. When I started to apply the delay (.0001 - .0005sec if I remember correctly) the exposure was getting brighter, meaning I was catching the declining tail of the flash a bit earlier. Then moving the delay to .0010-.0020sec the exposure was brighter again but the shutter curtain had also started to appear in the frame.
Nice report, and I think consistent with what we seem to believe is going on here.

Thanks.

Best regards,

Doug
 

Bobby Deal

New member
Sorry for my absence, things have been more then a bit hectic these past few days. I have a studio full again today but as soon as I get a chance I will take the control shots. While I have not taken the shots you suggest I have had a chance to play with the system a little in the studio. While I presume you are correct in how and why the system is working the important part for me is that it is indeed working.

Interesting observation though it definitely appears that it is not a true 1/8000 sync. I did some test shots of rotating fan blades which should be frozen solid at 1/8000 and still the 1/8000 exposure shows a touch of motion blur on the edges of the fan blades.

In the end though for a photographer like myself it is not really a matter of needing to understand why it works only that it does indeed work. The ability to shoot mid day with mono lights at f/2.8 is a beautiful thing and something that just does not work out the same with a reflector.
 
Top