Canon flash gurus - I need your input...

Recently I tried the wireless manual flash. Did a lot of reading before the

exercise, including RTFMs etc. The setup: MT-24 on the camera, set to manual

mode. 550EX and 580EX set both as C-slaves, forced into manual mode by holding

"Mode" buttons, both connected to fresh battery packs.

MT-24 set to A=1/64/ B=1/64, C=1/1 ratios, and both A and B heads blinded with

covers. 580EX Custom Function 7/1 set to pop the flash with full output set for

the flash, and 550EX fires at the set output in Manual mode by itself (talking

about test firing slaves with Pilot button here).

 

Camera in Manual mode, HighSpeed sync disabled, shutter set to 1/150 sec

(although this is not really relevant,since I wasn't taking pictures).

 

I have spent half a day trying to make it work - with very dubious results.

 

The first problem: even with slaves at a distance of 6 feet, directly looking at

the camera (and the MT-24) the triggering of slaves is unreliable. Sometimes 550

fired, sometimes 580, sometimes both. Changing MT-24 mode from Manual to E-TTL

mostly fixed the problem - but not 100%. I eliminated any flash recycling

problems by providing fresh battery packs and waiting till red lights on the

back of each slave was lit. It should work (and it does after a fashion) - but

is NOT reliable. I do not understand what is happening.

 

Second problem: metering slaves' output with a lightmeter is impossible despite

the absence of preflash in manual mode: the lightmeter reading are obviously

false and show metering of the secondary preflash (which in manual mode is used

to fire the slaves). So, no manual metering of the slaves.

 

Third problem: since I couldn't meter the slaves - I do not know if changing "C"

ratio on MT24 (master) acts as a multiplier applied to individual slave settings

(for instance: if MT-24 has C-ratio set to 1/2 and slaves to 1/4 and 1/16, would

this translate into 1/8 and 1/32 respectively ?). Or does the MT-24 overrides

the individual slave settings and forces them to fire at 1/2 power ?

 

My conclusion (possibly erroneous, that's why I am posting this) is that Canon

flashes in manual wireless mode are basically useless (although wireless E-TTL

works OK).

Second conclusion is that in order to make the wireless manual setup work - I

need Pocket Wizards on slaves, which can be triggered by radio module in the

flashmeter (Sekonic 558) for metering, plus a Pocket Wizard on the camera for

taking the shot. Initially I thought that maybe I should get ST-E2, but the

experience with using MT-24 as a master (which should essentially be equal to

ST-E2) is not encouraging.

 

 

I would appreciate any comments regarding the three specified problems - and my

conclusions. If I made any errors - I would be happy to hear where I screwed up.

I always considered the Canon flash system as overcomplicated and a triumph of

technology over common sense - but this I could live with - after thorough

familiarisation with the system and its quirks. What I can't stomach is

unreliable and unpredictable triggering in Manual mode (lightmeter problems I

expected).

 

To avoid unnecessary speculations - all contacts were CLEAN (with nothing to

contact...), all batteries FRESH (voltage measured OK) etc. Same results with

camera ON and OFF.

 

Any takers ?

Leszek-- I use Canon speedlights wirelessly all the time. I always use them in manual mode, so I can't comment on wireless ETTL and the ratios, groups, etc. and I've also never used the MT-24 (not even sure what that is...) but triggered with the ST-E2 or 580EX in Master mode using 580EXs and 430EX's as slaves in Manual I have never had any problems at all. The ST-E2's range is fairly limited however, especially so outdoors. If you need longer range or non-line-of-sight capabilities, then I think PocketWizards are the way to go, although some cheaper (not better) alternatives can be found on the big auction site.<br>Feel free to contact me via email if you would like more info.<p>Steve

"both A and B heads blinded with covers". Don't understand. If you do that, how is the MT-24EX supposed to send any signals to the slaves?

"both A and B heads blinded with covers"

 

Like Robin I am a bit baffled by that, mind you I am not flash guru.

 

My understanding is Group C flashes are only intended for illuminating the background rather than the subject so I am not sure what the exposure algorithm will make of that assuming the remotes are being signalled OK.

 

Yes the preflash will stuff a flash meter reading, annoying isn't it. Again my understanding is you still, of necessity, get pre-flashes in manual mode if triggering wirelessly. I am sure not but perhaps if you don't trigger wirelessly but switch to manual that might suppress the preflash, in which case you could use a slave trigger for the remotes.

Have you had any problems using the 580 as a master with only the 550 as a slave?

 

I have tried something similar using the MR-14 as a master (I assume you mean MR-24

not MT24) with two 550's but found it to be really difficult to control, In the end used a

550 as master and the other 550 illuminating the subject the spill bouncing off a reflector

 

I cant help feeling that another flash gun would be more useful than the ST-E2, but that

depends on whether the ST-E2 has more control than a 550/580 or a better layout.

 

Another problem I have is turning the master flash off. Either it wont respond to Flash off

or (more likely) the operator is a banana fingered baboon. I have the 550ex manual, it isnt

much help, Oh and the Lantern 550ex guide really is no help with wireless magic.

 

Good luck and I will be following this thread

 

Cheers G

Graham: yes, I tried 580 as a master with 550 slave - in ETTL mode. It works (no surprise). MR-14EX is a Canon ring-flash, MT24-EX is a Canon macro twin head flash, which can fire group C (with same power output for all flashes in group C). So - I meant MT-24EX, no mistake here. I keep reading funny things though about algorithms used to control group C in ETTL mode and mostly agree that this is somewhat mysterious - as are the controls when ratio A:B:C is used. Ratio A:B is directly controlled and the power is calculated based on preflash, while controlling C is via something akin to exposure compensation (but relative to what if Manual mode is used?). But this is another subject.

 

Lester and Robin: I did not need the preflash from MT-24 twin heads, I only needed the triggering signal for the slaves. So - I set the master to M mode, then both heads A and B to minimum power 1/64 - and simply put covers over the heads. The triggering signal does not come from the heads, so it should work. And it did - but not reliably. Also, I hoped that the near-infrared triggering signal would not affect the lightmeter - but it obviously did. Of course, it is possible to pop both slaves individually with their respective Pilot buttons and use the lightmeter in multi-flash mode (which worked OK, but I hoped that I could pop the slaves from the master directly).

 

Steve: I will give it another go today, and confirm/dispel my observations. At any case - it is possible that ST-E2 works better than MT-24EX as a master. In the end I will get a set of decent studio strobes with Pocket Wizards, but right now I am looking for some intermediate solution. I just tried to manually set the power on each slave and pop them with a signal from MT-24 as a master, with known results. Metering is not really aproblem - as I said above, multi-flash mode on the lightemeter and manual firing each of the slaves works OK (just a lot of walking...). Will send you E-mail if further help required :).

 

On a more general note - the instruction manual should cover ALL the aspects of performance and using the equipment. This far from being the case - as witnessed by numerous posts in this forum and elsewhere.

If you read/have read this: http://super.nova.org/DPR/Canon/MultiCanon/ and http://super.nova.org/DPR/Canon/Manual/ then that is a better start than the manual.

 

Chuck Gardner was extremely helful to me in setting up flashes, e.g. he sent me this answer re setting up channel 'C' (this is all Chuck's work):

 

"The camera meters A:B together, then pre-flashes C. It then sets C to match the level of A:B in the foreground. That's the zero FEC baseline for C, matching the foreground. Of course as with all automatic exposures the camera can't know if the background is a white, gray or black wall, or for that matter how you wish to render it in the photo. So you shoot a test shot, evaluate how the background looks relative to the foreground and then adjust the FEC for group C to make the background lighter or darker. +1 FEC for C simply means you want to override the camera's guess by one stop (2x more flash power).

 

That is all FEC means in any situation when you use ETTL, even with one flash. Setting FEC = 0 is telling the camera -- you decide what the exposure should be. You paid $1,500 for the darn thing because it is so smart so its only fair to let it see how smart it actually is before trying to out think it.

 

So you fire off a test shot at FEC =0 then LOOK AT IT. From the OEW

and histogram you will be able to tell at a glance if it is

overexposed (and where using the OEW ) or under (gap on right in

histogram). If under (usually the case at zero FEC) the size of the

gap will tell you how much FEC needs to be added. If over (rarely

occurs) I dial in - 1 FEC and make a second test shot.

 

Zero FEC is the camera's baseline exposure - its best guess. When you dial in FEC you simply are "second-guessing". + 1 FEC means "I think your guess on exposure was 1 stop too low for that scene"

 

If you find the camera consistently needs +1/3 stop FEC, then you can change your starting baseline for the test shot to FEC = +1/3 instead of zero.

 

I don't do the test shot scenario for every shot. I just keep an eye

on the OEW and histogram in the playback as I shoot and when I see

the highlights getting blown in critical areas or the playback get to

dark and histogram gap too large I adjust as needed.

 

If you compose well for flash, with your center of interest in the

foreground (i.e. nothing in front of it to fool the metering) once

you get the FEC dialed in all the other shots will be the same. ETTL

is very consistent shot-to-shot once you get FEC dialed in. You can

even change ETTL ratios and it will keep the highlights exactly the

same, as shown in my test sequence.

 

Its not perfect or foolproof, but its unrealistic to expect it to

be. As in all things you need to understand its limits and work

within those limits to mitigate risk. The only real risk in shooting

digital in RAW mode is overexposure (which you can't fix) or more

than 1-stop underexposure (which looks lousy when you fix it). So

if you are risk averse you don't try to stay on the bleeding edge of

overexposure in the camera you err on the side of slight under

exposure you can easily fix with a tweak of brightness in RAW"

OK, it appears that I was dumb (a bit). Covering flash heads on MT-24 actually renders its wireless capabilities ineffective, as it communicates with group C via signal from the heads (which I should have known if I thought a bit about what is NOT written in the FM). With heads A and B not covered, the communication in Manual mode appears to be fine. Even lightmeter works when popping slaves from the master - provided that the heads are set to 1/64 and pointed to the BACK (away from subject and the lightmeter). Otherwise the lightmeter picks the preflash. Not reliable, though, and requires trial-and-error placement of the master heads in relation to slaves. So, the way to go is probably Pocket Wizards, and in the meantime using MT-24EX as a master just for firing the slaves, with heads turned back and individual (manual) firing each of slaves for lightmetering purposes.

 

All in all - not a big discovery and something a lot of people know already - I just hoped to get this info without spending too much time on experimentong. But not a waste of time (although a big waste of batteries) :>D

 

Thanks for all inputs (I still think that this flash system is far from simple or user-friendly and has a lot of "undocumented features")

Other Canon master-capable flash units use visible light from their main flash tubes for their command channels. Even the ST-E2 uses a near-IR-pass filter over a standard flash tube. I'm not familiar with the operation of the MR-24 - are you sure it doesn't use the A/B channel tubes for signaling? What results do you get if you don't hood the tubes?

"Covering flash heads on MT-24 actually renders its wireless capabilities ineffective, as it communicates with group C via signal from the heads..."

 

That's what Robin and I were trying to tell you.

 

Sounds like you are trying to use the MT-24EX as a remote control, you might be better off getting a ST-E2 for this so you can have groups A and B off camera.

Yes, well, perhaps we don't need a post-mortem on that one, but it does serve to illustrate that the manuals for what are nowadays very complicated systems could usefully start with a basic explanation of what bits do what - although that gap has been plugged to a significant extent by some of the on-line information sources put together by enthusiasts.

 

Two small points that may be helpful if you do want to use the MR-14EX or MT-24EX as a master. First, unlike the 550EX/580EX/580EXII, there is no master-off capability. Secondly, it really is quite a good idea to understand what CF-5 on the flash actually does.

Lester and Robin - thanks for not being too harsh on me :)

 

Of course my mistake was to assume that MT-24 communicates with near-IR beam (while it does not even have any red transparent window - the way 550, 580 and ST-E2 have). So - OK, assumption is the mother of all f...ups. Lesson learnt (for 100th time, but still).

 

I know what CF-5 does: what I do not know what is the resultant power of each slave. The master controlling groups A and B can specify the output power (1/1 until 1/64) - but the slave in Manual mode also needs the power to be set. So, if for group A the master says 1/2 power, and slave A is set at 1/8 - what is the resulting output ? Is it 1/2 (as the master wants), the 1/8 (as set directly on the slave) or 1/16 (in case the power setting on the master and the slave are multiplied) ?

 

That's why I wanted only one function: to just fire the slaves. I used CF-0 on the master and put both slaves in group C, with manual setting of power output for each slave. I hoped that in the absence of any other input - the slaves would just fire at the specified setting.

 

Whatever the case may be - more is not necessarily better. In most cases it is just the opposite. Programmes and designers of electronic gear have this overwhelming and irresistible desire to put more functions, add one more layer of menu - just because it can be done. The result is a mess.

 

What I am arguing is that the system is just too complicated to be used efficiently.

 

ST-E2 may work well in E-TTL mode, where you just select ratio A:B. What about Manual mode ? Setting the ratio does not make sense here, so I am guessing it just sends the signal to fire the slaves...but, as I said before, assumption is the mother of...you know what.

 

The system is too complicated and poorly documented. No wonder that most of people who shoot for money get a bunch of Metz guns with Pocket Wizards. At least they know EXACTLY what the goddamn thing is going to do when they press the button. And - this is what I am going to do. Have no time for second-guessing and discovering of "undocumented features".

 

Thanks for the input anyway. If anything - it convinced me that while Canon's E-TTL system may be the state-of-the-art, it is poorly documented and not ergonomic to use.

 

Would you like to drive a car which is as complex to put in gear as the 580EX is ?

"Would you like to drive a car which is as complex to put in gear as the 580EX is ?" I used to; it was a Mitsubishi Pajero (Shogun in the UK). It had a conventional 5F1R manual box, a 5-state transfer box also controlling the centre difflock, and a rear difflock. Makes its successor, a LandCruiser, look childishly simple.

 

Back on-topic, my ASSUMPTION would be that a flash set to manual produces the output level set on the flash, and all that the wireless system does is to trigger it. Since I don't use manual flash, I have never needed to verify that. I have always believed, I hope correctly, that the various forms of E-TTL-mode ratio control available with wireless (and, indeed, between the two tubes of the MR-14EX and MT-24EX) have nothing DIRECTLY to do with power ratios anyhow; they surely control illumination ratios between the different groups, which is not the same thing at all, and controlling the power is simply the mechanism through which that is achieved.

"...it does not even have any red transparent window..."

 

Leszek, AFAIK all the flashes communicate via visible light pre-flashes, the exception being the ST-E2 which uses near IR. The red transparent window is just for the focus assist beams.

 

All I can suggest is a careful 'reading between the lines' of the manuals and other on-line sources such as PhotoNotes http://photonotes.org/articles/eos-flash/

 

As Robin points out, the EOS flash system is quite complex for a consumer electronic interface, anything beyond more or less letting it do its own thing requires a lot of research and experimenting to develop a "in head model" of how it works; so I feel your pain.

 

I think what you have been doing, trying to expose using only group C flashes is probably completely alien to the way it is intended to work will cause a lot of problems. If you wanted to expose for the subject using the MT-24 groups A and B and the group C for the background as the system is designed, I expect it would fall into place fairly easily.

 

To be honest, if your aim is to use manual flash most of the time I think there is an argument to step back in time and use some old 283's with the manual power control adaptor, a flash meter and some slave triggers.

 

Not that you can't use Canon flashes in manual, but with the preflash system it just seems to make it more complex and expensive than it needs to be for manual flash.

Actually I don't even mean illumination (shows just how careful you have to be) because different groups could be illuminating parts of the scene that were darker or lighter. I mean something like "what the meter picks up from each group".

 

While I'm back ... Leszek, you mention that the 550EX, etc, have a lot of dark red plastic on the front. Just concentrating on the 550EX, which is the one that I am still using, that's got nothing to do with sending signals, which is done entirely by white light from the main flash, although only the near-IR part of that is used. It's only the ST-E2 that sends an IR signal. What's behind the red plastic is the receiver used when the flash is set as a slave, and - this does not seem to be well-known - not one but two AF beam projectors - the system chooses between them, based presumably on the focal length and possibly the focusing point in use.

Lester - I just wanted to trigger the slaves. No exposure control, no nothing - just trigger the slaves. Which, for all the sophisticated features of the Canon flash system - appears to be not an easy thing to do after all. Not if you want to use lightmeter (and yes, I am aware of how one can use the camera to meter flash in M mode). Is this too much to ask ?

 

No, I am not going to use the flashes in M mode all the time: E-TTl works fine and as advertised, no problem with slave control. Except that I do not always want to take pictures the way Canon thinks they should be taken - and I would want to be able to control the flash directly.

 

As Robin rightfully noticed - the ratios A:B etc. are really not power ratios - but rather illumination - and this with a caveats regarding relative distances and reflection angles. So, no one really knows what E-TTL flash produces - except that the exposure is mostly OK, particularly if you watch the histogram and play with flash exposure compensation. Except...this has very little with learning how to set the lighting properly. You just entirely at the mercy of an unknown algorithm.

 

Regarding "going back in time" - majority of the goodies that modern civilisation has on offer is just pure bull...t. Supposedly more features are being offered - but at the price of great complication of the inteface and operation of the device. Of course - all the useful features and functions are there, just that they are buried within complex menu system and require pressing 30 buttons in a precise order. Well, I was never very good with programming a VCR with just 3 buttons.

 

Anyway, guys, I think this was quite a productive exercise. I know now that for manual flash - there are much better and cheaper solutions than Canon flash system. Have no argument with E-TTl.