CLS groups question

Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
I'm re-reading the Simon Stafford AF speedlight flash system book and can't seem to answer a couple of question. Actually I have a lot more than two questions, but I'll stick to these two for now.

It states that Nikon advises that you not use more than three remotes on any one group, why?

Secondly, nowhere can I find if the master, in my case the onboard flash of my D90, has any knowledge of how many remotes are in play. If I understand it correctly, the master produces a preflash, analyzes the reflected light and through magic makes all the calculations. Then it sends an IR flash telling the remote(s) how to produce the needed flash. It is said that the communication is one way only, master to remote. So, the master has no idea of how many remotes are going to respond, or how much light is going to be produced.

I'll ask you to please overlook any technical flaws in that chain of events and focus on the end results.
 
Joined
Nov 17, 2008
Messages
440
Location
Toronto, Ontario, Canada
If I understand it correctly, the master produces a preflash, analyzes the reflected light and through magic makes all the calculations.

The crucial part here is that the configured remotes also send a pre-flash during this moment. By analyzing these pre-flashes and how they interact with your scene, your camera figures out how much power all your flashes should output during the actual exposure. This is how your master technically knows how many groups are present. If this didn't happen, your camera would have no idea what to send to the remotes.

You can test this for yourself. Set your camera for a very long shutter length (1 second +), use rear curtain sync (so your actual flash is at the end of the exposure) and use mirror lock-up or exposure delay if you have it (not sure about the D90). This will delay the steps taken in an exposure and you can see when and how many pre-flashes there are.

The sequence is basically like this:

1) You press the shutter release button: just before the mirror goes up your camera sends a series of pre-flashes that tell all your remotes to fire. The remotes (the ones you've configured in the menu to be active and also the pop-up flash if you have it set to contribute to the exposure) send a pre-flash. These flashes are analyzed and your camera determines what power your remotes should be set to.

2) The mirror goes up and just before the shutter curtains open, your camera sends another series of pre-flashes. These tell all your remotes how powerful they need to fire later on.

3) The shutter opens and your exposure starts. At the end of the exposure just before the shutter closes again your camera sends out a final series of pre-flashes that signal the remotes to actually fire (at the end since it's rear curtain sync, if it's not then this will happen at the start and you won't notice it as being separate from the other pre-flashes). The remotes (the groups to be more specific) fire and you've got your picture.

This is from my own experimentation and will be different depending your settings (for example if your remotes are not set to TTL, then you won't see the first set of pre-flashes, etc.)

Hope that helps. Don't know the answer to the first question.
 
Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
I see where you're going and I do appreciate you taking your time to respond, you knew there was a but coming (didn't you?)

Recently I had asked a question on this site about going with a double mount for my key, used with an umbrella. I was told that it would add one stop to the key light. Unless I didn't understand what you wrote (always possible), the master would just turn down the two key flashes and I wouldn't get any increase. It seems to be at odds with what you are saying.
 
Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
I should of also said that the remote firing a preflash now makes sense. Thank you for clearing that up for me. I now remember reading that but hadn't retained it, so that I could connect the dots.
 
Joined
Apr 15, 2008
Messages
911
Location
Orange, CA
There is no IR flash. Infrared is present in the spectrum of light all flashes create.

However it is ONLY the IR spectrum that the sensors use to trigger/and analyze data from.

This is the reason if you placed a IR filter ontop of say your built in flash your remotes will continue to work without any visiable light coming from on camera.
 
Joined
Nov 17, 2008
Messages
440
Location
Toronto, Ontario, Canada
Recently I had asked a question on this site about going with a double mount for my key, used with an umbrella. I was told that it would add one stop to the key light. Unless I didn't understand what you wrote (always possible), the master would just turn down the two key flashes and I wouldn't get any increase. It seems to be at odds with what you are saying.

No, you are on the right path. Adding another flash would increase the total possible power output of a group, but that doesn't mean you need more power to start with because flash exposure is done through TTL.

So for example, if with one flash, for a particular scene your camera told it that it needs to fire at full power. Then if you add another flash, theoretically they are now told to fire at only half power (combined, they still output enough light to equal the full power of one flash).

This also means that your camera does indeed know how many flashes there are in a group.

Hope this answers your question.
 
Joined
May 24, 2007
Messages
208
Location
McKinney, TX
Camera flash awareness...

...This also means that your camera does indeed know how many flashes there are in a group...

Not sure I agree with that statement...

I'm pretty sure that communication between the commander unit and the remotes is one way:

Commander -->Remote.

The commander tells each active flash group when to fire and at what power to fire for the preflash. The camera evaluates exposure through iTTL, evaluates an adjustment, and then tells each group when to fire and at what power to fire for the actual exposure.

At no time does the operator "tell" the system how many flashes are in any one group. Nor is there any kind of "initialization" system for the system to query the remotes to find out how many are there or establish two-way communications.

I'm not sure how it does the calculations - and now that I think about it I can't explain multiple flashes being adjusted to compensate for multiple members in a group. I do think that all of the members of a single group are treated as one "contributor" to the exposure.

There are web sites that have documented the whole CLS/AWL communication sequence - and I am pretty sure that the communication is one-way.

Someone please chime in if I am completely wrong here...

Or several of you could send me 3 - 5 SB units of your choice so I can do a complete battery of tests to figure it out. Shouldn't take more than a year or two to report back.


EUREKA!
Figured it out - it doesn't matter how many flashes are in a group (post-edit here).

To restate:
"The commander tells each active flash group when to fire and at what power to fire for the preflash. The camera evaluates exposure through iTTL, evaluates an adjustment, and then tells each group when to fire and at what power to fire for the actual exposure."

It doesn't matter to the camera how many are in a group! The "evaluation" sequence just needs to know "how much to adjust" and tells all of the flashes to adjust accordingly. Essentially something like this: Commander to remotes: "I don't care what your power setting was during the pre-flashes! But all of you in Group A need to adjust your power to 1/4 of of what it was during the preflash sequence to yield the exposure I think you need to give."

I will think about it and see if I can find a better way to communicate this "theory".

Good luck
 
Last edited by a moderator:
Joined
May 24, 2007
Messages
208
Location
McKinney, TX
Joined
Jan 8, 2009
Messages
538
Location
PA
What flash units can be in these "groups"? Can they be any SB flashes or just the SB-600 & SB-800?

I also have a D90 and pretty much avoid using flash but when I do, it's mostly limited to the on-camera flash. I also have a recent non-SB flash and an old Vivitar-283 that I can slave off the on-camera flash but anything other than the on-camera flash is totally manual.
 
Joined
Nov 17, 2008
Messages
440
Location
Toronto, Ontario, Canada
At no time does the operator "tell" the system how many flashes are in any one group. Nor is there any kind of "initialization" system for the system to query the remotes to find out how many are there or establish two-way communications.

Do the remote flashes send pre-flashes? If yes, then there is two way communication.

It's a simple receive-respond type of communication between the master (camera and commander flash) and the remotes. It might be a single burst of pre-flashes from the remote flashes, but it's there and by my definition, two way.

If there was no response back from the remote flashes, then the camera would never have been able to determine the needed exposure in the first place through TTL (assuming that the pop-up flash was set to not contribute to the exposure in the commander menu).

This might be a case of semantics, but I do believe it's important to point this out in light of the original poster's questions

I'm not sure how it does the calculations - and now that I think about it I can't explain multiple flashes being adjusted to compensate for multiple members in a group. I do think that all of the members of a single group are treated as one "contributor" to the exposure.

I think you are right.

After I thought about it some more, the system used by CLS could be much simpler than I thought. The key is that when you have more than one flash per group then the overall power of the pre-flash from that group would be stronger, therefor the TTL system will read a stronger reflectance from the scene and the camera will dial down the flash power needed to expose the scene.

For example:

If you have one flash, it sends a pre-flash, the TTL system measures it and determines that the remote flash needs to fire at full power to get a "proper" exposure.

Now add a second flash to the group. Now the combined pre-flash power is twice as strong (assuming that the pre-flashes actually used for exposure metering are all the same power - the actual power doesn't matter, but it's definitely not full power of the flash as that would be a waste of power), the TTL system sees that there is twice as much return from the reflectance in the scene so it determines that you now need half the power from the flash. So now it sends a signal to the group to fire at half power. The message is received by both flashes in the group and they fire at half power. Half + half = full power so you are back at the same exposure!
 
Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
So, you're saying that if I'm using a double mount (2 flashes) on my key light, they're told to half the ouput and I'm getting the same results as if I had one flash on that postion? Hurrrumph, that's not what I was wanting the hear.

My skill level doesn't permit me to argue the point, but that doesn't sound right. In that case a SB900 with its higher guide number might not output any more light than a SB600, it would just be lowered (output) while the SB600 might be set to full output. Not sure that was a coherent thought, but I can't find a different way to put it. That removes any control from the operator. I've seen triple flash mounts, if the CLS is going to just lower the ouput by a third for each, what are you accomplishing.

Confusing.
 
Joined
May 24, 2007
Messages
208
Location
McKinney, TX
Two way...

...Do the remote flashes send pre-flashes? If yes, then there is two way communication.

It's a simple receive-respond type of communication between the master (camera and commander flash) and the remotes. It might be a single burst of pre-flashes from the remote flashes, but it's there and by my definition, two way.

If there was no response back from the remote flashes, then the camera would never have been able to determine the needed exposure in the first place through TTL (assuming that the pop-up flash was set to not contribute to the exposure in the commander menu)...

Then we agree... I think.

In my way of thinking - "two way" communication is not just simply following a command. "Two Way" implies communicating something that the commander cannot logically conclude without additional information, something "additional". If my model (theory) is correct, the commander is not really receiving information that it cannot ascertain independently. The remotes are merely responding to information from the commander and reacting. The commander/camera combo is doing all of the logic.

I was thinking of "two way" communication as something that the remote "flashes back" to the commander via the very quick pre-flash signals. I can see with your assessment of the actual pre-flash used to estimate exposure as a form of communication. It is not the way I have mentally modeled the system - but an open mind is a virtue and I see your point.

Thanks...
 
Last edited by a moderator:
Joined
May 24, 2007
Messages
208
Location
McKinney, TX
It is a Simple test...

So, you're saying that if I'm using a double mount (2 flashes) on my key light, they're told to half the ouput and I'm getting the same results as if I had one flash on that postion? Hurrrumph, that's not what I was wanting the hear.

My skill level doesn't permit me to argue the point, but that doesn't sound right. In that case a SB900 with its higher guide number might not output any more light than a SB600, it would just be lowered (output) while the SB600 might be set to full output. Not sure that was a coherent thought, but I can't find a different way to put it. That removes any control from the operator. I've seen triple flash mounts, if the CLS is going to just lower the ouput by a third for each, what are you accomplishing.

Confusing.

Try it - If you have 2 CLS/AWL capable remotes.

Set one up to fire at an inanimate object. A gray card would be perfect, since that is exactly what the metering system is attempting to expose perfectly. Setting it up on the included stand would be perfect. I would suggest that the flash's focal length be set to something that completely covers the gray card.

Now set another flash right next to the first flash, programmed in the same group, with identical settings for focal length. (I am referring to the zoom or focus feature on the flash, not of the lens mounted on the camera.)

If my theory is correct, the second exposure should be the same as the first. The flash system would communicate to the flashes to fire at 1/4 power to get the exposure correct.

If you are correct, the second exposure should be twice as bright as the first. The flash system would communicate to the flashes to fire at 1/2 power and overexpose the subject.

I know where I'd place my money...

Sources for potential variation:
The two flashes are not aimed precisely at the same location.
 
Joined
May 24, 2007
Messages
208
Location
McKinney, TX
What kind of SB flashes?

What flash units can be in these "groups"? Can they be any SB flashes or just the SB-600 & SB-800?

I also have a D90 and pretty much avoid using flash but when I do, it's mostly limited to the on-camera flash. I also have a recent non-SB flash and an old Vivitar-283 that I can slave off the on-camera flash but anything other than the on-camera flash is totally manual.

By definition, the flash MUST be iTTL and CLS/AWL (Nikon language = Creative Lighting System/Advanced Wireless Lighting) capable.
Four flashes meet those requirements:
SB-900
SB-800
SB-600
And the little ??-200 flashes that come with the macro flash kit.

All previous flashes will not work. The SB-400 is (in my way of thinking) a "neutered" flash. It cannot do anything without being attached to a CLS-capable body (I believe it is all digital SLR bodies released after the D70).

There are a couple of ways to mix non-CLS flashes into a CLS system - but the theory is complex, requires additional hardware, and an advanced understanding of CLS.
I have never tried it myself, but I know an "expert" with the system and he has successfully done it. To me, if I need non-CLS lights, then I would just revert back to full manual flash operation and traditional triggering systems (radio, wires, dumb optical triggers, etc.)

Good luck.
 
Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
Well, if I could ever get off the merry-go-round of my D90 going to Califorina and back, I would. Presently it's getting old in El Segundo.
 
Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
From what I've read you're supposed to try to balance your key and fill light at a 1:2.1:4 ratio (off the cuff, so if I'm off on this, let it slide.) If that is the case and CLS is doing a overall balancing act with your flash compensation values, wouldn't it alter the contrast that you're trying to achieve. It in effect would be working against you. Again, I wish I was better with the written word.
 
Joined
Feb 25, 2009
Messages
324
Location
Independence, Mo.
Possible semi intelligent thought. By seperating the flashes in to groups you prevent Nikon from averaging out your outout ,from overiding your creative plans??? Yes/no, I'm a loser and won't ever get this? Key in one group and fill in another, seperation of control????
 
Joined
May 24, 2007
Messages
208
Location
McKinney, TX
Now you're catching on...

Possible semi intelligent thought. By seperating the flashes in to groups you prevent Nikon from averaging out your outout ,from overiding your creative plans??? Yes/no, I'm a loser and won't ever get this? Key in one group and fill in another, seperation of control????

Yes. Exactly.

If you were to (for the sake of discussion) put 2 flashes each on two stands:

Scenario 1: The two flashes on each stand are "mixed up" so that one flash on each stand was in a different group, you should get flat lighting (little conrast) or unpredictable results at worst.
[Stand 1 - Channel A, Channel B]
[Stand 2 - Channel A, Channel B]

Scenario 2: The two flashes on each stand are "paired" and set to the same channel. They should behave in a rational way - in effect acting as a single light.
[Stand 1 - Channel A, Channel A]
[Stand 2 - Channel B, Channel B]

In effect, the second scenario can be used to provide more power when needed or can be used to reduce cycle time while shooting. Each flash does half as much work, so the recycle time drops considerably.

I wish I could give a more definitive answer about exactly how the EV values relate to fstops for light ratios. I can tell you that I generally shoot with the fill at -1.3 to -1.7 EV from the main, however.

Good luck.
 
Joined
Sep 17, 2006
Messages
12,748
Location
Ashburton , New Zealand
So, you're saying that if I'm using a double mount (2 flashes) on my key light, they're told to half the ouput and I'm getting the same results as if I had one flash on that postion? Hurrrumph, that's not what I was wanting the hear.

That's what the exposure compensation settings are for .
If you have three different channels and and the key light needs to be brighter simply dial in a '+' compensation on that channel - you can't override the cameras decision by adding more flashes , that would create havoc .
Wireless flash meters the same as TTL/BL flash , I did a blog on TTL/BL and it includes a bit on wireless flash .
Imagine if you had 4 speedlights on each channel and were getting 4X the exposure what a nightmare that would be !? All the main flash does is tell the remotes to fire a preflash [ I saw the thread on nikonians ] , each speedlight on the specific channel fires its preflash and the camera works out how much power each 'channel' needs to fire on . It doesn't know if there were 4 flashes at a distance firing in one area or one flash close up so it gives the same exposure either by , for example , telling one flash to fire full power or 4 to each fire 1/4 power as mentioned . Use the exposure compensation settings to adjust each channel . Imagine if you had only one flash and it was too bright for your taste - you can't take any flashes away , you have to dial in negative compensation on that channel .
 
Joined
May 11, 2006
Messages
1,916
Location
Northern Italy, Piemonte
...... snip........ "Two Way" implies communicating something that the commander cannot logically conclude without additional information, something "additional". ...... snip........
That's exactly what is happening. The commander asks the remotes to fire group by group and measures the reflected light from each group in sequence via the TTL flash metering system. The reflected light returning to the camera's CLS/AWL computer makes up the second leg of the two way communications system and it contains information that the commander could not logically conclude without this "additional" information. The return leg of the two way system is made up of both the fixed response of the remote flashes AND the reflectivity of the subject. Without the return leg, the system won't work. So, for me at least, it is clearly "two way".

It's not unlike the kid's pool game, Marco Polo. The kids responding "Polo" give the person who is "it" good information about bearing rough information about range, something the person who is "it" couldn't conclude by himself. If nobody ever says "Polo", then hollering "Marco" won't do much good.
 
Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Nikon Cafe is a fan site and not associated with Nikon Corporation.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Forum GIFs powered by GIPHY: https://giphy.com/
Copyright © Amin Forums, LLC
Top Bottom