Odd Odin Behavior+Odin question+a flash meter question (Phew! Too much?):

[jonathan7007]

Behavior: I finally did a firmware update 9/6/2014 -- issued by Phottix a year ago for TCU and receivers. My Odin kit was 1.0 and those sold now are 1.5, but Phottix implies that the firmware update makes them functionally identical (they use the word "compatible".)

I set up a little test shoot to see the results of the firmware -- and practice some use cases I have not yet done with the Odin -- for clients. I just want to learn more about the Odin kit. Here's the problem: Since the update yesterday there are times I shoot a frame in which no flash occurs and then the second shot will fire as expected -- as will further frames as long as I keep going. This wake-up frame seems to occur when there is perhaps a minute or two of inactivity.

I first noticed this yesterday so I cleared the TCU, took out the batteries and let it sit without power for a day. This had helped some photographers last year at the time of the firmware update, which didn’t go smoothly for everyone. My flashes in this test are one each 580EX and 550EX. They are set to stay on, not to go to sleep. It is not battery power. But tonight I had that blank frame reoccur.

I was shooting tethered to an Android tablet and thought something wonky in its shutter button but the same issue happened with out tether. A pause takes the camera into some form of standby. I will ask Phottix. Any thoughts here?

NEXT,
Nowhere can I find a detailed discussion about when it's best to leave the flash units as TTL using "TTL Compensation" on the TCU to affect a ratio. Non-TTL is called Manual but the flash is STILL set for "ETTL". I have been asking myself how TTL on controller *and* flash head (speedlight) can work when the shoot calls for the flash unit to be facing the inside of a umbrella of softbox. Using TTL in the test setup I did see some change in the flash output when I put a white card behind the mannequin head (my "model"). In TTL with changed aperture the whole rig was "OK" at producing similar histograms. But one of my strobes, the 580, was flashing into a silvery umbrella on every shot, The other light was a 550EX directly at the model.

Changing mode on the TCU to "manual" would seem to avoid confusion from TTL decisions, I guess, so I set full power from the umbrella and 1/8 power from the direct 550. Shoot. This strong ratio does show on the back-of-camera screen as soon as I hit the shutter. Changes in ratio show right away.

I have very little experience with TTL and trust manual (admittedly limiting). I expect to meter the flash as I own a Sekonic L558. Even though there appears to be no preflash in this Phottix version of "manual" my meter is confused. It reads f3.2 for a flash level that is f10 when I get the histogram right. (I know in-camera histograms aren't perfect…) (In preflash for TTL operation my meter successfully waited for and displayed pretty accurate results of a "second" flash stronger than the first.

Why is there a wake-up shot required?
Why doesn't the speedlight inside something or facing an umbrella mess up the Odin TTL system?

 

[Datfish]

Johnathan Hi again,

First up the confusion between 1 and 1.5 is common I think. Version 1 shows the 1/3 stop increments as +0.3,+.06 and version 1.5 shows them as 1/3,2/3...that is the only difference as far as I understand.
Why, you ask , are there two variants?...good question.
It seems the v1 was released but many requests were received to change the way the display showed 1/3 stop increments. The change could not be done simply by firmware, as it needed a hardware change to the LCD, so Phottix released V1.5 with the only change being the display as described.

I am not sure you are up to the latest firmware of 1.24 for the Odin 1 which I have, but I will assume you are?

With respect to "missing the first shot" , I have experienced that on occasions and do recall reading something about the Odin needing one shot to "learn or establish ' the radio ttl control signal or something along those lines when initially set up. My own experiences have been:

a) First up apparent misfire when initially setup but then no issues. However this first up "misfire " is not universally true as sometimes the first shot "fires". I have not been able to verify, as I have not specifically tested BUT I have a THEORY that the "misfire" only occurs when dramatic changes such as "mode" or Hss ect are made in the initial setup. IOW if I simply finish a session with triggers and guns in manual (or TTL)l, dismantle and pack away then setup again but still use guns and triggers in the same manner as the last everything will work fine. Might be worth you testing that!

(Note simple changes of power, or even mode "mid stream in a session" don't cause an initial misfire after change as the radio TTL connection is already established...?)

b) have had occasions where not having the TCU seated firmly enough on the hotshoe causes misfire issues first up.

Regardless I always fire a test shot anyway to verify connectivity and firing.....suggest it might be good practice for you too!

Your NEXT point is a bit confusing and I am not sure I fully understand the question but here goes!

Regardless of where your flash is pointing, and within reason where it is behind an obstacle relative to the camera , the TTL control functions will work as they are via a 2.4 ghz radio signal. I think your question is about how and where does the TTL metering take place in your test set up?... AND more importantly WHAT is being metered?

Can I suggest that you set up a one light test to figure that out. I'm not a user of TTL ratios so I am unfamiliar with it but for your purposes I think you need to test each component one at a time and then combine.

a) Verify accuracy of TTL with single light shooting straight at target....should be fine but check anyway
b) Verify accuracy of TTL when bouncing the same single light off an umbrella...should be fine but check anyway..it might be 1-2 3rds out
c) Combine the two in a two flash setup using manual to determine what you think is "correct" and convert to a ratio.
d) switch to ratio mode and set according to C) calculated above

As for the metering question "I have been asking myself how TTL on controller *and* flash head (speedlight) can work when the shoot calls for the flash unit to be facing the inside of a umbrella of softbox. "....................the metering happens in the camera and is taken from the pre-flash of both. How else could it work?

Remember me from FM etc and the tests of output from the Mitros+? The discussion there which was all about power to setting accuracy, and those test were conducted using the test button to trigger the flash in manual or the hotshoe.

IMPORTANT POINT:
ONLY when using the "test button" do you get sufficient delay ( 2.5 seconds) between the pre-flash and main pulse for a light meter to pick the difference and register the main pulse accurately.
Contact with Sekonic verified the above point.

Your observation of "It reads f3.2 for a flash level that is f10 when I get the histogram right." bears this out. I'm guessing but I think you are trying to measure "live" ttl output from the flash ?

IF that is the case I suspect your f3.2 is the result of the pre flash ( not the main) at 1/32 power which just happens to be the power of the pre-flash on my 580 ex ii. Check yours by metering your flash and pre flash pulse by setting the light up on the Odin RX and the mode on the flash set to TTL, Grp A etc and trigger via a TCU (TEST BUTTON) with Group A set to Manual. Set to full power and you will see the meter first register the pre-flash and 2.5 seconds the main pulse will fire. ( Note: you might need to set the test shot power level to full instead of 1/32 via the Cfn menus before hand.... Cfn-07 = 1 on a 580 exii )

In short I don't think you can use a meter to measure the TTL output when its in "live" situations because in "live" situations the delay between pre and main flash is in milliseconds. The meter registers the pre and can't recover when the pre > main.

"Why doesn't the speedlight inside something or facing an umbrella mess up the Odin TTL system?"

Odin uses radio TTL for controlling the settings, the ttl metering is done in the camera via and from the pre-flash. The results of that metering is then communicated to the flash guns via the Odin radio system as opposed to hard electrical via wire or hotshoe connection.
In either case the GUNS ARE BLIND so it does not matter where they face or where they are positioned ( within reason) behind obstructions.

Also DON'T confuse RADIO TTL with IR/OPTICAL TTL!. The actual TTL metering stays the same (in Camera) but the fundamental difference is the way in which those signals and controls are sent to the guns. IR/Optical systems need to have a sensor "see" the signal where as Radio TTL systems use radio receivers. That difference is indeed the major benefit of Radio TTL control as you are not restricted to line of sight.

One last point and a personal opinion only...I think the use of the simplistic term TTL to describe the entire control system is WRONG or at the very least CONFUSING. TTL is in fact the metering system and the translation and use of that metering result to control the output of a Flash. I might be wrong but it, like many other innovations , was first brought to the table by Olympus and was limited I think initially to on camera/ cable use , because the "transport system" was limited to Hard wire. Then came IR and Optical

Note I said "the translation and use of that metering result" and NOT the transport of it. Systems like Odin, Pocket Wizard etc have brought the innovation of radio signal "transport" of those "results" that's all. TTL has not changed dramatically but the transport systems have.

Your understandable confusion stems from the widespread bundling of all things under the one "TTL" title, and is probably a result of you being familiar with IR/Optical systems limitations to a line of sight environment. Radio blows that limitation out of the water.

I know it wont change but I think "we", as in the photography population, should be a little more precise in the way we describe what we're talking about :

TTL = Through the Lens metering and its translation to Flash settings
E-TTL = Through the lens Metering using evaluative metering and its translation to Flash settings
Radio E-TTL = above with radio transport of the settings to Flash
Optical E-TTL = above except using optical pulse to transport the settings to Flash
etc. etc


Hope this helps Johnathan and apologies if I have missed the point or "explained" something to you that did not need explaining!

 

[jonathan7007]

Datfish,
Good to exchange ideas with you again, yes I remember and have been following your pursuit of Nikon-system consistency for Odin implementation here and on FM forums. In fact, your test descriptions made me dig out my Odin for some "quality time", which I am embarrassed to admit I did not put in when I got the kit. I have a TCU, three Odin rcvrs, and one Strato2 (at the time I planned to use speedlights and a 2000WS Norman pack for interiors.) Now that I have the Einsteins and their dedicated receivers I haven't needed the Phottix triggers very often. The arrival of the Mitros+ got me thinking about adding one of those to my speedlight kit, so --- testing!

OK, back to my rant/questions.

All my testing occurred right after upgrading to the v1.24 firmware.

I had forgotten that "TTL" systems were based on in-body analysis. Duh! Years ago I did a lot of editorial work with those Vivitars and Sunpacks that watched the scene in front of them and STOPPED the flash when the head saw enough light for the chosen setting. I didn't "reset" my head (no pun intended) and realize that speedlights in modern TTL were just emitters -- smart emitters -- but their position in relation to the photo subject was not used directly for info input. Therefore the speedlight could be buried in a softbox and the camera would analyze the light shed -- from all emitters -- on the subject in the preflash.

I read more after my post and got good at pushing test on the TCU and resetting the Sekonic for the "real" flash 2.5 seconds later. I checked with Sekonic and they confirmed that my L558 will reset for the next strobe flash within about a second. THIS IS INTERESTING: the tech told me that even though the meter reads the light "seen" during the pre-flash it will ignore that reading after reset -- without clearing the meter info screen --and give me a clean reading of the full-strength pulse. He felt I did not have to push the Measurement button again.

OK, switch to manual light output settings on some(?) or all of the channels, and there is no longer a 2.5-second gap between preflash and "real" pulse. So the meter cannot be used. No way to get a reading without using the camera sensor, I guess.

Missing flash trigger on first shot... Clients/models get nervous when this happens and I don't like using anything that makes them worry. I don't remember these "wake-up-moments" from shoots a while back before upgrading the firmware. Maybe they happened. Missing frames doesn't square with everyone's eager praise foe 100% perfect Phottix Odins. I haven't heard this described by anybody.

I wrote Phottix some days ago and added to the thread on the Help interface. Nothing after a kind of boilerplate response to my first question: everything off, take out all batteries, replace, try again. Maybe helped but I was looking for 100% firing.

Your reply was helpful, thanks. I'm pretty experienced (had my TTL head screwed in wrong...) but wanted to dig in and really understand the system so everything you've been investigating has moved all of us further along. I see myself using the manual ratio control set for the most part. That way I lose the metering, though...Frustrating and bush compared to the way we took elaborate care to know exactly what light we had before we exposed film, especially view camera work. Even with Polaroids for testing...

BTW the ease of access to the HSS capabilities of the Phottix rig is truly neat.

Again, thanks for a detailed post to help out.

Jonathan

 

[Datfish]

Johnathan,

No worries

"Missing flash trigger on first shot... Clients/models get nervous when this happens and I don't like using anything that makes them worry. I don't remember these "wake-up-moments" from shoots a while back before upgrading the firmware. Maybe they happened. Missing frames doesn't square with everyone's eager praise foe 100% perfect Phottix Odins. I haven't heard this described by anybody."

I have done some more checking of this with my Odins and to be honest I can't seem to get them to miss a first frame!!
That said I am sure that this does happen/has happened to me occasionally I'll just have to take more notice when it does. I have had other misfire issues at times but found them to be related to the flash or TCU not being properly seated, but that is not what I am speaking of here. Like you describe a first up "miss" then no further "misses" for an entire session has happened?.....but I can't seem to replicate it in testing.

I even just now tried a complete battery change and re set on all the odins without a first up misfire, and then tried the same using MF to ensure there was no shutter/button lag waiting for AF....same result with no misfires....BUT I KNOW IT SOMETIMES DOES HAPPEN!

HOWEVER I also took special note AND FOUND
IF the TTL comms are correctly established during setup......WHEN you engage Metering via the halfpress of the shutter button or via another assigned function button, the following happens:

a)The green LED on front of ODIN receivers stops blinking and remains on.
And IF the same button also engages AF
b) the Red LED in the 580 exii and Mitros+ Flash units (used as IR pulse) also lights up and pulse a couple of times

( Note : I could see the red pulses despite them being taped over with gaffer tape ,to block the pulsing. That pulsing annoyed me and I thought it to be superfluous when using radio!?...perhaps its not entirely? ......they might be uncovered, at least partially, now? :)

So "maybe" instead of just taking an initial shot to establish the TTL control comms and expecting it to misfire,( as my vague recollection of a "post" on another forum suggested) we should just be disciplined enough to test as above before we blaze away...just as we would if connecting by cable/IR/Optical now or "back in the day"....we sometimes are unreasonably hard task masters?....its just the way the Odins are setup to operate and not a malfunction/fault!

THAT SAID it helps when you know how the kit works in "detail" , BUT the instruction manuals leave a lot to be desired!!... BUT HEY!..we're working this stuff out through forums like this I suppose!

There is after all, a fair bit that needs to happen with RADIO TTL control as opposed to a simple manual trigger signal like a pc-sync connection ( i.e set mode, power levels & zoom for those groups/guns on the TCU then trigger) so it might be worthwhile just taking that extra step to make sure the "transport" system is working!?

Sure Radio signalling is fast but I could imagine a situation when using MF ( or when AF is already established), where the shutter is released before that initial TTL signal is established and results in a "misfire". Quirky perhaps but feasible! That is perhaps why Phottix put in the annoying IR light pulse to signal "connection established"...( No arguments that they should have better user manuals!).

"OK, switch to manual light output settings on some(?) or all of the channels, and there is no longer a 2.5-second gap between preflash and "real" pulse. So the meter cannot be used. No way to get a reading without using the camera sensor, I guess."

I don't understand the above. Do you mean having the flashes themselves set to manual or the TCU. If the guns are set to TTL and the TCU group is set to manual the pre flash will still pulse if USING THE TEST BUTTON! If the guns themselves are set to manual the odins behave as a simple manual trigger and no pre-flash is emitted.? why would it need to for manual? BUT you can still meter.

The only time you can't meter is when using in TTL mode ( ie TTL on The TCU for the group, and the flash itself) and shooting "live". (ie relying on the camera to trigger and NOT the Test Button of the TCU)

".........He felt I did not have to push the Measurement button again."

Correct!..if the L558 is the same as the L308 in non cord mode (I think it is) when the gap is 2.5 seconds it firstly records the pre-flash and then records the main pulse and overwrites the initial pre flash reading on the LCD...no need to press the "meter" button again.

Andrew