Please explain how multi-group ettl preflash works?

[Marsu42]

I know what ettl2 is and what the preflash does in general, what I'd like to understand is how multiple groups work.

* Is it that every group does successive pre-flashes on their own so that the camera can determine what ettl ambient/flash ratio to set?

* Is this also the reason why pre-2012 don't work in group mode as they are only able to meter two pre-flashes (a:b and c) and rt requires up to five (a-e)? However, even if so Canon could at least have made it possible to use groups a+b in ettl and d-e in m flash mode (not to bash Canon, just for my understanding of the system)?

 

[privatebydesign]

I know what ettl2 is and what the preflash does in general, what I'd like to understand is how multiple groups work.

* Is it that every group does successive pre-flashes on their own so that the camera can determine what ettl ambient/flash ratio to set?

* Is this also the reason why pre-2012 don't work in group mode as they are only able to meter two pre-flashes (a:b and c) and rt requires up to five (a-e)? However, even if so Canon could at least have made it possible to use groups a+b in ettl and d-e in m flash mode (not to bash Canon, just for my understanding of the system)?

* Yes.

* Probably to all of it. It is also the reason pre 2012 bodies can't do HSS and ETTL at the same time with the RT system and the reason the flash sync speed is lowered by one stop when using groups (not Groups) via the RT system.

 

[alexanderferdinand]

Thanks for the information.
Thought about that too.

 

[Marsu42]

It is also the reason pre 2012 bodies can't do HSS and ETTL at the same time with the RT system and the reason the flash sync speed is lowered by one stop when using groups (not Groups) via the RT system.

Thanks for the explanation concerning the preflash - but my pre-2012 60d *can* do hss and ettl at the same time via rt (it's just limited to the 2 group traditional a:b-c system) and I now can also say that the crop body works @full sync speed via rt, with my 2 rt flashes I never saw the black bar you experience with your 1d (see also http://cpn.canon-europe.com/content/education/technical/syl_arena_new_speedlite_system.do?page=4)

 

[privatebydesign]

Another logical conclusion to draw shoots down a criticism of the new system, that it can't do both optical and RT at the same time, but think about it, you'd have eight groups that all needed a pre flash, there just isn't enough time between pressing the shutter button and the exposure starting to send all those signals, read and process the results and then send all the instructions out. A 1DX shutter lag as standard is 55ms, it takes a good amount of that time to stop the lens down so all test pulses and readings have to be done before that time, which is closer to 36ms.

 

[Lawliet]

Every group gets its own preflash, for a:b both are metered seperately to find out how much power each has to provide to get the desired lighting ratio. At the same time those seperate preflashes (and the low res metering sensor) make the "don't cross the streams"-rule neccessary, the camera can't tell whether to responses will light two areas of your subject or add due to overlap.

For pre2012 its hard to tell how much of the timings and such are hardwired in the metering subsystem. There just might be some weirdness gremlins be lurking in there.???

 

[privatebydesign]

It is also the reason pre 2012 bodies can't do HSS and ETTL at the same time with the RT system and the reason the flash sync speed is lowered by one stop when using groups (not Groups) via the RT system.

Thanks for the explanation concerning the preflash - but my pre-2012 60d *can* do hss and ettl at the same time via rt (it's just limited to the 2 group traditional a:b-c system) and I now can also say that the crop body works @full sync speed via rt, with my 2 rt flashes I never saw the black bar you experience with your 1d (see also http://cpn.canon-europe.com/content/education/technical/syl_arena_new_speedlite_system.do?page=4)

You can select HSS and ETTL in pre 2012 bodies, it just is nowhere near as consistent as non HSS ETTL, I get regular +/- 2/3 exposure variation across HSS + ETTL use where as for same framed images ETTL by itself is very consistent.

As for the bar/shadow in HSS, I still haven't pinned down exactly when I get it, again I can select 1/250 (my 1Ds MkIII sync speed) and get the warning icon, but it will still fire and work fine, sometimes I have a feint bar, sometimes nothing, and sometimes I do lose the full stop to get rid of every trace of shutter shadow. I can well imagine the crop cameras being much more forgiving as the shutter has less distance to travel. I do think Canon were being very conservative in their warning of losing one stop sync for pre 2012 bodies, I am just very glad they left us the ability to actually select our standard sync speed with the RT system on the pre 2012 bodies.

 

[Marsu42]

for a:b both are metered seperately to find out how much power each has to provide to get the desired lighting ratio.

Interesting, in that case it's even more strange Canon doesn't allow three independent a-b-c groups with independent fec for the optical legacy system if they do 3 preflashes anyway...

You can select HSS and ETTL in pre 2012 bodies, it just is nowhere near as consistent as non HSS ETTL, I get regular +/- 2/3 exposure variation across HSS + ETTL use where as for same framed images ETTL by itself is very consistent.

That's very valuable information, I might be experiencing the same thing with the 60d - though it's hard to pin down where the problem is, but I'll keep watching for it if I do optical (I still have a 430ex2) vs. rt.

I am just very glad they left us the ability to actually select our standard sync speed with the RT system on the pre 2012 bodies.

Yes, somehow doesn't sound like them

There just might be some weirdness gremlins be lurking in there.???

Like marketing gremlins :-> ? ... but ok, there really might be hardware contraints due to the amount of preflashes, but even if not Canon almost never backports such features anyway. I still think optical fallback with a 2 or 3 group constraint would have been possible to keep on using optical flashes while rt are linked via radio at the same time.