Paul C. Buff, Inc. Technical Forum

I changed the title of the thread because I found the same behavior with a manual shoe mounted flash that I expect is also IGBT controlled.

This was brought up on a forum and I did some testing here are the results I got;
All the testing was accomplished with a Canon 5D2. ISO and aperture were adjusted to make proper exposure depending on strobe power.
Einstein at full power 1/200 shutter full frame illuminated.
Einstein at minimum power 1/200 shutter speed TOP of frame has a shutter curtain shadow.
Advance shutter beyond 1/200 to 1/320 and both TOP and Bottom of frame has shutter curtain shadow.
Reduce shutter speed to 1/100 (PW rear curtain sync default) and full frame is illuminated.
To verify that PW rear curtain sync speed setting was involved I set it to lower value (1/60) there was shutter curtain shadow between 1/200 and it went away when shutter was set to 1/60.

Did same set of tests with Canon 580EX in manual mode, YN560, AB800, Genesis 300B, and Savage FT-150. All these lights delivered full frame illumination at full power and minimum power.

Using other triggers (Cyber Sync and RF-602) the Einstein provides full frame illumination at minimum power as well as full power.

I'm posting this here since PCB, Inc. is working with PW to develop a controller for the Einstein that will work in the Control TL system.

If appropriate, I can cross post at PW support but it may be easier for your engineers to validate the situation and notify PW as part of the development program.

The 5D+ E640 combo results are not surprising. With the 580 and 560 both having TTL capabilities, more information would be needed for those results to be as informative a possible. So for the moment, we will need to reject these results. I also need to assume you understand how standard x-sync works.

What is happening is the PW TT transmitter is tapping into the TTL protocol for the 5D and can "sense" when the flash is aboput to fire. Using this information, it sends a fire signal a tad earlier than x-sync. As most flashes fire and "slowly" burn off, the camera opens and closes the shutter before the flash burns off completely. This is called "hypersyncing" by Pocket Wizard. At full power, the Einstein acts like most other studio flashes: fast peak pulse and "slow" trail off toward the end. As the unit is dialed down, the IGBT cuts this tail off, making the flash durations shorter, much shorter. If you fire an Einstein too early, it will flash and die off before the shutter can make its full travel.

With the exception of the two speedlites listed, none of the other flashes use IGBT control, and all have a longer tail, which gets longer as you dial power down. All of the other remotes are not capable of tapping into the TTL communications of the camera, and therefore cannot fire too early.

My suspicion is the speedlites, even set to manual, may be going into a High Speed Sync mode, and firing at a sustained output for the entire exposure time. This is a different kind of syncronization no studio light can do at this time.

Pocket Wizard is aware of this, as are we. It is a matter of timing, and Pocket Wizard will currently be the best source of answers as to what settings are the best.

I agree that Pocket Wizard has to address the issue.
I think what you are missing is that the Einstein is causing a front curtain shadow to be seen until the rear curtain sync is activated at the slower shutter speed.

The YN560 is a manual shoe mounted flash. Since the shutter speed was never above the nominal sync speed, Hyper Sync and High Speed Sync were not active on the 580EX. Or at least should not have been. That is very likely the problem. By the way, one test was to set the Hyper Sync offset to Zero. This had no impact. There still was a front curtain shadow.

Since, in this instance the shadow is from the front curtain, the extremely short Einstein exposure at low power setting should not have an impact. If anything they would cause the bottom rather than the top of the frame to be dark. Similar to rear curtain shadow when too high a shutter speed is used.

I think you are accurate in suggesting that the PW firmware is pre-firing the Einstein and because of the short flash duration, there is not enough tail light to fill in the shadow caused by the front curtain.

My reason for posting here is so that both PCB and LPA are aware of the flaw in the Control TL software. If not addressed, the Einstein will not be useful as part of a lighting scheme using Control TL for triggering.

I could be wrong, but my impression was that Hypersync (pre-firing) was active at all shutter speeds by default.

One can adjust the Hypersync Delay on the Flex-TT's, so perhaps that could bring it into alignment, or perhaps one could turn off the Hypersync completely.... it's not essential to the E-TTL function, just another feature they've added.

Perhaps the PowerMC2 receivers will address the issue once they're released.

You could be right.
When I adjusted the off-set to 0 it had no affect. There was till a front curtain shadow until rear curtain sync was invoked.

This may be a case were PW added a feature that works with voltage adjusted monolights and fails with more sophisticated lights like the Einstein.

I got a reply from PW. They say its the "faster than wire" pre-fire signal along with a "slower" 5D shutter, combined with the Einstein light quenching for lower power settings is the likely cause.
They asked me to configure the Flex and Mini to act as basic triggers to confirm.
I did and as expected no leading curtain shadow. There was a trailing curtain shadow at 1/200 that went away at 1/160. This has been my experience with PWs and my 5D before the 5D2.

So, with my curiosity peaked, I tried a couple of other tests.
First, I placed a Cyber Sync Transmitter on a Flex set to factory defaults. I put a CSXCV on the Einstein. All was well. Full image at 1/200. Trailing curtain shadow at 1/250 and black frame at 1/320.

And, just because I had them, I also put a Yongnuo RF-602 trigger on the Flex and plugged a receiver into the Einstein. Same thing. Just the slightest hint of a trailing curtain shadow at 1/200, more at 1/250, and at 1/320 almost all the frame was blocked by the leading curtain but there was a thin sliver of exposed image at the bottom. This phenomenon continued all the way to 1/8000.

It has always been my impression, and these tests tend to verify, that the hot shoe on the Flex is controlled by the PW fimrware and never a direct link via radio unless the Control TL functions are turned off via the check box.

So, when the Flex is plugged into the Einstein sync port via a cable it behaves one way. When the Flex is triggering the Eisntein via a radio link then it is behaving differently.

Unfortunately, when a cable is used the behavior is unacceptable.
It also means that PW still has some serious bugs in its software.

The good news it that things seem to work as one would expect when the Cyber Sync trigger is used with the Flex. That means I can control the Einsteins via a Cyber Commander and still have them participating in a lighting scheme that uses Control TL on the camera with Flexes for Speedlites.

That might even be a better approach than waiting for the PW Einstein Modules.

I just did another test using a Yongnuo YN560 manual flash unit.

I plugged a Flex TT5 into the YN560 via the sync port on its side. It exhibits the same behavior as the Einstein. That is front curtain shadow a minimum power at 1/200. No shadow when slowed to rear curtain sync speed. No shadow when at full power.

And just like the Einstein, the YN performed as expected when it was sitting in the hotshoe of the Flex. That is, no shadow at 1/200 at minimum power.

This suggests that there is something in the Flex circuitry or firmware that is treating the sync jack differently than the sync pin interface on the hotshoe.

PW as been heard from again:


Hi again David,

Thanks very much for all the time you've put into testing. I've been trying to reproduce the issue here and have seen some similar results. It does appear like the P2 port is triggering the flash earlier than the hot shoe. I've shared my findings with our engineers and they've already tracked down the cause. We hope to have it implemented in a new firmware release soon, though I don't have a specific time frame for when it will be ready.

UPDATE:
Pocket Wizard Support sent me a firmware for both the Mini and the Flex and asked me to try it with the three lights that had exhibited a problem.

I loaded the firmware and tested the YN560, Canon 580EXII and Einstein. All plugged into the P2 port.
All delivered clean frames at minimum power at 1/200 second. Also at 1/250 second. At 1/320 second the frame was black. This seems reasonable to me. At minimum power the flash duration is so short there is no light available to expose the sensor.

So, when the next maintenance release firmware is available Einsteins (and other IGBT controlled flashes) will expose the frame properly at minimum power.

Did they mention when this firmware will be released? I'm having the same issue, and it is very frustrating. I've contacted pocket wizard, and they haven't been any help.