474 Full Power Pops! No kidding!

Thanks, guys.

Leo, Dave, et al;
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.

dmward wrote in post #11666669
They asked me to configure the Flex and Mini to act as basic triggers to confirm.

Didn't know you could do that but that's a useful test.
You also might want to copy that entire post onto the AB techforum just so they know about this quirk and the workaround...

It's on the AB tech forum.

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.

More from PW. There is a problem and they are addressing it.

From PW TS:

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.

dmward wrote in post #11668589
More from PW. There is a problem and they are addressing it.

From PW TS:

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.

Give yourself a pat on the back, David. Well done!

Dave F.

dmward wrote in post #11668589
More from PW. There is a problem and they are addressing it.

From PW TS:

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.

Good job, David, finding the "discrepancy" between the hotshoe sync and the P2 port sync. I'm assuming it's just with certain flashes that this phenomena can be demonstrated? Because I couldn't replicate this behavior with any of my hotshoe flashes although you did with your Yongnuo.

Thanks also for the updates from PW.

Leo,
I was a bit surprised when the YN exhibited the exact same characteristics when plugged into the Flex via the PC connection.

I did it more to confirm that there was a discrepancy with the Einstein.

I just tried a 580EXII just for grins. Same shadow problem at 1/200 when being fired via a cable on P2.
Naturally, the flash on the Flex hot shoe was playing with all the TTL controls, including automatically switching to HSS when I went over 1/200 so those examples aren't really valid for this testing sequence.

dmward wrote in post #11671968
Leo,
I was a bit surprised when the YN exhibited the exact same characteristics when plugged into the Flex via the PC connection.

I did it more to confirm that there was a discrepancy with the Einstein.

I just tried a 580EXII just for grins. Same shadow problem at 1/200 when being fired via a cable on P2.
Naturally, the flash on the Flex hot shoe was playing with all the TTL controls, including automatically switching to HSS when I went over 1/200 so those examples aren't really valid for this testing sequence.

Hmmm, that's interesting. You could replicate the problem with the 580EX II as well? And I couldn't. What firmware version are you on? And what's you FlexTT2 revision level?

Also, were you triggering the flash using a remote FlexTT5 or the one mounted on the camera (or doesn't it matter)? I just tried it with the flashes hooked up to the P2 port on the FlexTT5 on the camera. Maybe the issue is with the P2 port on the remote FlexTT5.

Leo,
All these tests have been with a Mini on the camera and the FlexTT5 connected to the various flash units.
The software is latest production version, I think that's 5.2. Not sure about hardware revision level. I've used a couple of FlexTT5s during the testing.

dmward wrote in post #11674418
Leo,
All these tests have been with a Mini on the camera and the FlexTT5 connected to the various flash units.
The software is latest production version, I think that's 5.2. Not sure about hardware revision level. I've used a couple of FlexTT5s during the testing.

Thanks, David. I was able to replicate the problem by connecting a hotshoe flash to the P2 port of the remote FlexTT5 unit. So this is not a problem unique to the Einstein but also to all hotshoe flashes, and other strobes that work like the Einstein, connected through P2 of the remote FlexTT2.

All IGBT controlled flashes. Which would tend to be all hot shoe mountable flashes. Not sure about Quantum.

And it does not appear to be a problem with Voltage regulated monolights ala AB or Genesis. (The two that I have.) Primarily because they have such long tails, especially at low power settings.

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.

dmward wrote in post #11712239
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.

You forgot to include "when plugged into port P2". 'Cuz it works fine when connected to the hotshoe via a hotshoe adapter, right?

Thanks for the amplification.
You're right. They did work if using the hot shoe.