5D3 Second Curtain Sync - Design Flaw?

[CharlieB]

I don't think this is a design flaw.

Either by way of direct mounting, or coupled external flash (cable, wireless), there needs to be an TTL flash "known" by the camera.

The reason, I believe, is due to the nature of the 2nd curtain synch.  Although the curtain travel time is very fast, at some flash output settings - which are tube firing time derived - you will probably get some unwanted consequences. 

Put another way, 2nd curtain synch actually occurs BEFORE the 2nd curtain begins to move.  But how much before?  It depends on how long the flash tube is going fire in order to achieve what the "camera/flash" system believe are correct exposure.  This may be splitting hairs, but on larger units, the flash tube firing time can be just under 1/1000 second, and no doubt Canon has accounted for this timing factor by limiting the way 2nd curtain synch is implemented.

Moreover... I would not be surprised that if you first set the camera to 2nd curtain, and then went with a non-TTL remote setup, that the camera might "say" its on 2nd curtain, but actually has reverted to first curtain.  I can see Canon doing that.

[Meh]

@CharlieB I believe he has said he verified that second curtain sync is maintained after he removed the Canon flash.   However, I tend to agree with you that it is reasonable that the flash system would have to be known/recognized/compatible before any advanced functions/settings beyond simple triggering would become available.  Not that it must be that way but more because from a design perspective Canon would want to test and verify performance of the system before making certain settings available.   If they made every setting available for any attached flash system and something didn't work correctly image the screaming they'd be dealing with.  Perhaps the software designers have more insight into the potential issues in supporting third party devices that haven't been fully tested and may not comply with the system design.  Let's not forget that third party flashes are reverse engineered (I believe so anyway) to be "compatible" and therefore are often not fully compatible.

I'm not saying that second curtain sync is difficult to implement or that it should be restricted when attaching third party flashes/triggers only that Canon decided to turn off the flash control menu for all flash functions if a non-Canon (or non-compatible, non-recognized, non-tested) is attached.  Could be for good reasons or just business reasons to encourage sales of Canon flashes.

[TrumpetPower!]

Either by way of direct mounting, or coupled external flash (cable, wireless), there needs to be an TTL flash "known" by the camera.

No, there doesn't. Second-curtain sync is much, much older than any form of TTL; it's as old as electronic flashes, period.

The problem with high-speed sync is due to the speed the shutter curtains travel, not due to the flash duration. All flash durations are much shorter than the fastest of shutter curtains. When you shoot at a speed faster than your flash sync, what happens is that the flash pulse has shot its wad while one of the pokey shutter curtains still hasn't cleared the sensor.

With a slow shutter speed, the one curtain completely moves out of the way of the sensor, the flash signal is sent, the curtains remain stationary for the duration of the exposure, and then the second curtain covers the sensor. With second-curtain sync, all that happens is that the flash signal is sent just before the second curtain starts to close rather than jut after the first curtain has finished opening.

With a fast shutter speed, faster than the rated sync speed, the shutter speed is shorter than the time it takes the curtain to get out of the way. So, even before the first curtain has fully cleared the sensor, while only a partial slit of the sensor is exposed, the second curtain already starts to close and chase the first sensor. Again, the flash is fired as soon as the first curtain has fully cleared the sensor...but much of the rest of the sensor is already obscured by the second curtain that's hot on the heels of the first.

High-speed sync gets around that problem with some sort of precision-timed jiujitsu that fires the flash multiple times while the curtains are only exposing part of the scene. How, exactly, the engineers figured out how to do that in a way that doesn't look weird is beyond me...but they did manage to pull it off.

And TTL metering is completely unrelated to curtain travel. These days, I think it's mostly accomplished by some sort of high-speed preflash, but there've been so many other sorts of methods used that I wouldn't quote me on that.

Cheers,

b&

[CharlieB]

@CharlieB I believe he has said he verified that second curtain sync is maintained after he removed the Canon flash. 

I took that to mean its maintained in the menu system... wasn't sure if it was actually doing 2nd curtain.

Well aware of all complexities with flash synch... you simply cannot have the start of 2nd curtain travel coincide with the synch.  Synch has to be slightly before 2nd curtain travel begins, and it needs to be enough before that, so that the flash has finished firing before the curtain has covered any of the film gate (sensor gate?).

Not talking HHS here, but good ol' X synch.   Doing a bit of research.... curtain travel time, complete traverse at X synch speed of 1/250 sec, is about 3.5milliseconds, or... 3.5/1000ths of a second.  Clearly, if a longish flash duration of 1/1000 sec is called for, we're gonna have a curtain that is about 1/3 of the way across the sensor when the flash stops firing - IF - the synch is at the beginning of 2nd curtain travel.   Therefore, actual 2nd curtain synch is really "near 2nd curtain" synch, whereas, the synch must occur with at least as much time before 2nd curtain travel, as it takes for the flash itself to fire, and to allow for any systemic delays.

So... I'd say, someplace at least around 1/1500th of a second before actual 2nd curtain movement, you've got 2nd curtain synch, which insures that a long duration (not HSS) flash is fully extinguished prior to the shutter actually starting to close. 

How Canon handles that... speculative, but... its not unreasonable to think they take it into account, and vary the actual synch in relation to the 2nd curtain movement, based on the predicted flash firing time (not HSS).
As we know, with conventional flash photography, the flash duration, not impulse power, determines the effective flash exposure.  The flash is made to fire for shorter times as less "power" is needed.

As far as I know, there is no 2nd curtain synch for HSS, they're mutually implementable modes of operation. 

[Mt Spokane Photography]

I've been using a variety of flash guns at (Nikon, Canon, Metz, Sunpak) Strobist shoots on my 5D2's for 3 years and have always been able to access the flash menu to change curtain sync between 1st and 2nd. A variety of triggers have been used as well.

On my 5D MK II and all my other Canon cameras, if I do not have a Canon COMPATIBLE flash attached, I do not get the speedlight functions, just a error message.
However, I do not need a Canon branded flash or trigger, just one that is compatible.
Are you really able to access the speedlight functions with a nikon Flash attached?
 
 

[gbchriste]

@CharlieB I believe he has said he verified that second curtain sync is maintained after he removed the Canon flash.   However, I tend to agree with you that it is reasonable that the flash system would have to be known/recognized/compatible before any advanced functions/settings beyond simple triggering would become available.  Not that it must be that way but more because from a design perspective Canon would want to test and verify performance of the system before making certain settings available.   If they made every setting available for any attached flash system and something didn't work correctly image the screaming they'd be dealing with.  Perhaps the software designers have more insight into the potential issues in supporting third party devices that haven't been fully tested and may not comply with the system design.  Let's not forget that third party flashes are reverse engineered (I believe so anyway) to be "compatible" and therefore are often not fully compatible.

I'm not saying that second curtain sync is difficult to implement or that it should be restricted when attaching third party flashes/triggers only that Canon decided to turn off the flash control menu for all flash functions if a non-Canon (or non-compatible, non-recognized, non-tested) is attached.  Could be for good reasons or just business reasons to encourage sales of Canon flashes.

I have verified that the second curtain sync setting is maintained and applied after removing the flash from the camera. I went as far as powering everything down, removing the flash from the camera, attaching the flash to a "dumb" radio trigger (fire signal only), then powering up the camera. I set the shutter at 6 seconds so I could clearly detect whether the flash was firing on first or second curtain. It was clearly firing on the second curtain, just tiny but perceptible instant before the shutter closed.

My intent was never to have any of the flash intelligence maintained here for ETTL, power, zoom, etc. I just want a plain and simple "fire the flash now" signal sent to the hot shot.  From there my Cyber Sync trigger fires the remote flash.

I have all that working exactly the way I need and want it. I just don't get why the only path there is via having a Canon flash attached to the camera to start the set up.

[Meh]

@CharlieB I believe he has said he verified that second curtain sync is maintained after he removed the Canon flash.   However, I tend to agree with you that it is reasonable that the flash system would have to be known/recognized/compatible before any advanced functions/settings beyond simple triggering would become available.  Not that it must be that way but more because from a design perspective Canon would want to test and verify performance of the system before making certain settings available.   If they made every setting available for any attached flash system and something didn't work correctly image the screaming they'd be dealing with.  Perhaps the software designers have more insight into the potential issues in supporting third party devices that haven't been fully tested and may not comply with the system design.  Let's not forget that third party flashes are reverse engineered (I believe so anyway) to be "compatible" and therefore are often not fully compatible.

I'm not saying that second curtain sync is difficult to implement or that it should be restricted when attaching third party flashes/triggers only that Canon decided to turn off the flash control menu for all flash functions if a non-Canon (or non-compatible, non-recognized, non-tested) is attached.  Could be for good reasons or just business reasons to encourage sales of Canon flashes.

I have verified that the second curtain sync setting is maintained and applied after removing the flash from the camera. I went as far as powering everything down, removing the flash from the camera, attaching the flash to a "dumb" radio trigger (fire signal only), then powering up the camera. I set the shutter at 6 seconds so I could clearly detect whether the flash was firing on first or second curtain. It was clearly firing on the second curtain, just tiny but perceptible instant before the shutter closed.

My intent was never to have any of the flash intelligence maintained here for ETTL, power, zoom, etc. I just want a plain and simple "fire the flash now" signal sent to the hot shot.  From there my Cyber Sync trigger fires the remote flash.

I have all that working exactly the way I need and want it. I just don't get why the only path there is via having a Canon flash attached to the camera to start the set up.

All any of us can do is speculate and the speculation that has been given is that your 5D3 does not recognize your trigger as "compatible" and therefore the flash settings menu is not available.  Right or wrong, good reasons or bad, like it or not, that is the most likely explanation.

[And-Rew]

I've been using a variety of flash guns at (Nikon, Canon, Metz, Sunpak) Strobist shoots on my 5D2's for 3 years and have always been able to access the flash menu to change curtain sync between 1st and 2nd. A variety of triggers have been used as well.

On my 5D MK II and all my other Canon cameras, if I do not have a Canon COMPATIBLE flash attached, I do not get the speedlight functions, just a error message.
However, I do not need a Canon branded flash or trigger, just one that is compatible.
Are you really able to access the speedlight functions with a nikon Flash attached?

The shoot scenarios were TFP based shoots with Flickr based Strobist groups using:

'Nikon' based RF602's for firing multiple flashes of varying makes in manual mode
'Canon' based RF602's/3's for firing multiple flashes of varying modes in manual mode
Pocket Wizards of varying kinds etc...
Some cheap PW variety of triggers.

In all cases, I just went into the menu and found the curtain synch mode and changed from 1st to 2nd, and 2nd to 1st. In the heat of a TFP strobist shoot where my allocated time to work with a particular model in a particular location is limited to about 5 minutes - I did not make notes - but I do remember doing everything extremely quickly and on only a few occasion did I attach an actual flash to my 5D2 body.

As has been said - why is a curtain synch menu option a Speedlite dedicated feature? I was using both options on my Canon A1 and Sunpak flash back in '87, so can not see the need of a Speedlite sub menu unless working with a dedicated Speedlite flash?

[Studio1930]

I agree that the option should be available even when a flash is not mounted since you may be using a remote flash (with a basic trigger).  Older Canon cameras allowed this and for good reason (lots of photographers used it without a flash or canon trigger mounted).

Cannon needs to put that option back in the menu system full time.  Now I need to go check my 1DX and see if it is missing the same option.   

[Quackator]

Reading through that tutorial, the shot calls for second curtain sync. 

Not for me. You want the flash to fire when they are in their final position.
What is wrong with starting 1st sync in that final position (and thus be sure
about the exact posing and point in time when triggering) and then double use
the flash as signal to the couple to start light painting?

With 2nd sync you have to hope that the couple is in a perfect position after
exactly six seconds (or whatever your exposure is). With 1st sync you *know*
that they are.

[TrumpetPower!]

How Canon handles that... speculative, but... its not unreasonable to think they take it into account, and vary the actual synch in relation to the 2nd curtain movement, based on the predicted flash firing time (not HSS).

If that's what Canon is doing, then it is, indeed, unreasonable.

Second-curtain flash has worked fine for decades without trying to cut it as close to the wire as possible. Presumably, the timing is based off some sort of standard flash duration. If the curtain happens to still be open for a fraction of a microsecond after the flash extinguishes, that's never been a problem in real-world shooting. Similarly, I've never heard of a real-world problem with the second curtain closing too soon -- and, if that did happen because you were using a really, really slow flash, you'd presumably know enough about your exotic equipment to cut the power a bit.

Whatever the reason for this bug, it is a bug, and inexcusable.

Cheers,

b&