fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Lighting & Studio Techniques | Join Upload & Sell
  

Archive 2012 · Speedliting in groups?
  
 
no_surrender
Offline
• • •
Upload & Sell: Off
p.1 #1 · Speedliting in groups?


I'm confused and could use some help. I keep reading about using Speedlites in groups (A,B,C) but still not sure what the purpose is. Why can't I just set my flashes to manual mode and go from there? Do I still need to use groups? My Canon 30D can't control flash exposure from the camera menu anyways, do the groups still apply to me? Please enlighten me!

Kevin


Apr 07, 2012 at 12:05 AM
dmward
Offline
• • • •
Upload & Sell: On
p.1 #2 · Speedliting in groups?


Groups permit ratios relative to the on-camera master speedlite. This can be useful in manual as well as TTL.

The manual that came with your flash is a good source of information as well as several sites that should be found with a google or bing search.
Apr 07, 2012 at 08:29 AM
hugowolf
Offline
• • •
Upload & Sell: Off
p.1 #3 · Speedliting in groups?


The Canon three group system is quite limited, when used for ETTL. It was design for a ratio between two groups (A:B) and with the C group outputting at the combined output of A+B. The only way to adjust group C, in ETTL, is by using FEC.

In remote manual mode, you have three groups to work with and can set the output of each group, A , B, and C. Except that the master flash is always part of Group A – you can however, set the master flash not to fire.

With an ST-E2 as the master, then you are back to A:B and C only as a background light.

Brian A
Apr 07, 2012 at 11:02 AM
no_surrender
Offline
• • •
Upload & Sell: Off
p.1 #4 · Speedliting in groups?


Thanks you two. I don't normally shoot with my Speedlite(s) in ETTL anyway. I guess I'm just going to need to experiment with this and try and learn from it. I would imagine grouping the flashguns would better suite someone shooting with a newer model body and/or controlling the slaves from the Master unit.
Apr 07, 2012 at 01:26 PM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.1 #5 · Speedliting in groups?


no_surrender wrote:
...Why can't I just set my flashes to manual mode and go from there? Do I still need to use groups?

The advantage is that you can vary the power of the remote flashes from the master flash; that's handy if the remotes are up on light stands or otherwise inconvenient to reach.

For example, on a recent shoot I had a 580EX II key light in a 20" beauty dish (Group A), a 580EX fill flash in a 28" soft box (Group B), and a 430EX II hair light in a 12" gridded dish up on a boom stand (Group C). I could independently control the power levels on all of them from the camera menu on my 7D, using the pop-up flash as the Master. On my 20D, which lacks built-in Master capability, I could have done the same by using another Speedlite as the Master, or by connecting the key light to the camera with a long ETTL to use it as the Master.
Apr 07, 2012 at 08:32 PM
no_surrender
Offline
• • •
Upload & Sell: Off
p.1 #6 · Speedliting in groups?


Thanks BrianO, that would have been one advantage to buying another Canon Speedlite a few months ago as opposed to the LumoPro. I love the light from the SB's, but it sure is a PITA opening up the diffusion panel every 5 seconds to make adjustments! Just one more reason though to eventually buy another Speedlite or upgrade bodies....key word, eventually.

Kevin
Apr 08, 2012 at 12:47 AM
BrianO
Offline
• • • • •
Upload & Sell: Off
p.1 #7 · Speedliting in groups?


no_surrender wrote:
...I love the light from the SB's, but it sure is a PITA opening up the diffusion panel every 5 seconds to make adjustments! Just one more reason though to eventually buy another Speedlite... ...key word, eventually.

"Eventually" is right. For the time being I'm not buying any more Speedlites; I'm saving my money until I have enough to get two or three of the new 600EX-RTs with built-in radio sync.

Until then I'll make do with what I have, and maybe add some studio lights.
Apr 08, 2012 at 03:07 AM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.1 #8 · Speedliting in groups?


The Canon system isn't too difficult to grasp once it is actually used and the role of key, fill and background lights are understood.

The role of fill is to lift the entire scene the camera sees to the point the sensor records details in the black objects of the scene where you want detail (a subjective decision). A 580ex in the hot shoe or on the OC-E3 cord over the camera on a hot shoe over the lens performs the role of fill quite effectively. In the Canon system it also works as the Master controlling the off camera flashes. It defaults to Group A, so in the Canon ratios with Master as Fill A= Fill.

The role of the off camera "key" flash is to create highlights on the raised surfaces of 3D objects. The resulting 2D contrast pattern of highlight over the Master-filled shadows is what creates the illusion of 3D in the photo. How real that illusion look is for the most part a function of key light angle relative to the object. For example faces are rendered very realistically if the key light is placed 45° higher than the eye line and 45° to the side of the subject's nose — wherever it is pointing. The photo below is typlcal of that configuration:

http://super.nova.org/MP/BBBwide.jpg
http://super.nova.org/MP/BBBhs.jpg

For wide shots of action, still life subjects and animals it is often more effective to more the off camera light around behind the the subject to "rim" light it with the flash on the camera providing the frontal lighting on the shaded side. Below are typical examples where a Group B slave was placed behnd the subject as "rim" lighting.

http://super.nova.org/MP/10mm_ActionB.jpg
http://super.nova.org/MP/StillLife1.jpg
http://super.nova.org/MP/DogBackLeft.jpg

In either situation, slave in front as "key" light or behind and "rim" light it's intensity can be controlled two ways remotely from the 580ex Master connected to the hot shoe. In ETTL mode the relative intensity is controlled via a bar diagram on the back of the flash showing the A:B ratio. The range is 8:1 (8x more A than B) to 1:8 (8x more B than A). The numbers represent the INCIDENT strength of the two groups at the subject, not the conventional lighting ratio which is a REFLECTED ratio inferred from assuming the two overlap—key overlapping fill:

So when the Group B slave is used in front overlapping the fill this is what happens if a 1:1 ratio is selected:

H:S
1:1 Group A (Master) being centered illuminates everything with 1 unit of light.
1:0 Group B (Slave) being off center only hits the highlights with 1 unit of light.
===
2:1 But the REFLECTED ratio wind up being 2:1 because KEY OVERLAPS FILL IN THE HIGHLIGHTS.

That's not the case if the slave is moved behind as "rim" light. Then it is just a matter of balancing the lights visually. In most cases the rim light would be pegged slightly brighter than the frontal lighting. In backlighting situations it is also a common practive to bounce the light off the ceiling or use a diffuser which splits the light up / forward into "key" (modeling) and "fill" (shadow control) components.

With respect to exposure control M or ETTL can be used. In M mode I use the following workflow:

1) Adjust the Group A fill power until I see detail in the black objects in the foreground

2) Adjust the Group B power (as "key" in front or "rim in back) until the white objects in the foregroud are just below clipping (I like to retain highlight detail at capture — YMMV).

Both are done visually because the pre-flashes used to command the remote Group B slave will affect the meter reading and the flashes cannot be conveniently fired separately to deterimine a metered ratio. Manual control works with direct or bounced flash — if the lights and subjects stay at the same distance. It is the better choice or studio style portraits because once set the exposures will be consistent frame-to-frame.

In ETTL mode the workflow is different. For starters one needs to know from experience what each of the A:B ratios typically look like. That can vary depending on lighitng strategy (direct vs. bounce) and how much "spill fill" in bouncing off the walls and ceiling as seen in my examples above.

I find from experience testing and using my pair of 580ex flashes for seven years that in most situations setting an A:B = 1:2 ratio will allow my cameras (20D and 50D) to record a full range of detail from black suit to white shirt in a typical scene. That numberical ratio, on the Canon flash means that 2x more light from the slave (key) is hitting the subject, vs the fill on the bracket above my camera, which due to the overlap of key ove fill creates a 3:1 reflected ratio per the convention used for portrait ratios:

H:S
1:1 Group A Fill hits everything
2:0 Grouo B "Key" overlaps creating highlights on top of filled shadows
===
3:1 Ratio

My goal ratio-wise isn't numerical, it is recording a full range of detail. In any given situation only one A:B ratio of fill/key will match scene-to-sensor where the flash. A:B = 1:2 does not always match scene-to-sensor it is simply my starting baseline.

My workflow using ETTL ratios is as follows:

1) Set ratio to 1:2 as starting baseline
2) Adjust FEC until higlights on white scene content in foreground is below clipping.
3) Evaluate shadow detail / tone and adjust ratio as needed.

In ETTL the primary control is FEC, which is adjusted based on how the highlights look. The shadows and the A:B ratio can't objectively be evaluated until the FEC is adjusted for the highlights, however the photographer chooses to expose them.

The advantage of ETTL vs. M is that the camera metering and flash control logic sort out how much power is needed on the A Master and B slave to keep the B flash 2x stonger (incident) regardless of the distance, whether light is direct or bounced, and other variables affecting exposure. It is able to do that because it fires separate metering flashes for the A and B groups. So it knows after seeing the Group A pre-flash and comparing it with Group B's how much of total capacity each needs to output so that B winds up being 2X more at the subject than B. What FEC does it move the point of correct exposure at that ratio forwards (- FEC) or backwards (+FEC ) from the camera (assuming they are being used directly).

Because there are so many variables affecting scene reflectance and because the metering zones on the viewfinder are relatively larger and imprecise ETTL exposures will vary slightly from frame-to-frame on a fixed scene. In situations where the photographer or action is moving each new shot becomes a new problem to solve. When shooting with ETTL flash I consantly review results with the clipping warning and histogram visible in the playblack. Seeing the clipping warning tells me FEC needs to be reduced. Seeing no clipping and a big gap on the right side of the histogram tells me the last shot was likely underexposed and I will raise FEC for the next shot. The changes in FEC needed are on the order of 2-3 click of the dial on the back of the Master (1/3 - 2/3 stop). FEC can also be adjusted via the camera, but I find doing it on my Master 580ex over the camera more convenient — YMMV.

Group C is handed differently depending on whether M or ETTL is used. With M mode there is no metering so lights can be placed anywhere in any role, but the one connected to the hotshoe as Master will always default to group A. So shooting in M group A would still be Fill for me (YMMV) with B used as "key" (frontal modeling) and C for either background or rim lighting, as in this three flash shot:

http://super.nova.org/MP/ThreeFlashesGPS.jpg

In that situation I set the f/stop needed for DOF, then adjusted the power of each flash visually until the Group A (fill) was revealing the desired shadow detail and Groups B and C were creating the desired highlight exposure. It's not rocket science: fill for the shadows, adjust key and rim based on clipping in the playback. It simply takes a few test shots.


In ETTL mode the metering doesn't consider the light emiitted from the C goup when controlling the A:B output. Canon warns in the manual that in ETTL if the group C light hits the foreground being controlled with the A:B ratio the results will be unpredictable. So in A:B C mode the C group is expected by the metering to be lighting the background as in this example:

http://super.nova.org/MP/LightingDiagram.jpg

There I was moving around shot-to-shot with Group B slave on a rolling stand and Group C super clamped to whatever was convenient, adjusting A:B (frontal / rim light) and C, visually based on the playback.

That's with the optical signalling. The new radio system changes the paradigm. It allows up to five groups and simulateous use of some with ETTL and others in M mode.

Most of the time I shoot with two flashes in a key over bracketed fill configuration with the slave on a single stand I roll around in front as "key" or behnd as "rim" as the situation dictates. I control how the background and overall space is illumiinated by opting to bounce the key and fill or use it directly, so I rarely find it necessary to use a third flash as either background or rim lighitng (with key and fill in front). I also have a set of studio lights and use them when I want a 3-4 light portrait senario.

Because ETTL does vary shot to shot the more flashes you add into a ETTL strategy the more likely the results will vary. So in most situations requring more than 2 flashes M mode is usually more predictable and consistent and for that reason the better choice.

But before running out and buying five 600EX-R speelights @ $630 each stop and consider the logistics of setting up 3,4 or five lights on stands isn't much different for studio mono-lights and inverters vs speedlights. The more complex and static the lighting situation the fewer advantages there are to using speedlights vs. monolights. The mono-lights would offer more power, more modification options, and depending on brand selected lower cost than buying a 3,4, or 5 580ex or 600EX-R speedlights. So while it is technically possible to create and control 5 group scenarios with the 600EX-R there may be little if any logistical advantages except in situations where the light power must be adjusted remotely. Even then options like the Buff Cybercommander system might be more ideal solution. Just food for thought.
Apr 08, 2012 at 06:58 AM
ukphotographer
Offline
• • •
Upload & Sell: Off
p.1 #9 · Speedliting in groups?


BrianO wrote:
I'm saving my money until I have enough to get two or three of the new 600EX-RTs with built-in radio sync.

The 600EX-RT's seem like a great solution with a host of added benefits.
Apr 08, 2012 at 08:13 AM
no_surrender
Offline
• • •
Upload & Sell: Off
p.1 #10 · Speedliting in groups?


cgardner wrote:
The Canon system isn't too difficult to grasp once it is actually used and the role of key, fill and background lights are understood.

The role of fill is to lift the entire scene the camera sees to the point the sensor records details in the black objects of the scene where you want detail (a subjective decision). A 580ex in the hot shoe or on the OC-E3 cord over the camera on a hot shoe over the lens performs the role of fill quite effectively. In the Canon system it also works as the Master controlling the off camera flashes....Show more →

Chuck, once again thanks for the description, how-to, and sample images! Since only 2/3 of the flashguns I have are Canon Speedlites, I'm going to start practicing with just a fill and key light. I'd like to see what the difference is between shooting a scene (whether indoors or outdoors) with only ambient vs one Speedlite mounted in camera hot-shoe vs one Speedlite off-camera vs the two light A:B ratio. I'm going to do some practicing later this afternoon and hope to post some results by this evening. There's really no other way to learn than to just try and experiment, unless I take a formal course I guess. I can't afford to wait and try and learn this during a real shoot, which by the way is this coming Saturday. The shoot will be in an old barn as well as outdoors with my friend and will unfortunately be my last shoot before I head off to Korea. Here's to hoping I can put all of what I've learned from you, BrianO, and others to use and make ya proud! Cheers!!

Kevin
Apr 08, 2012 at 11:33 AM
Jim Quinn
Offline
• •
Upload & Sell: Off
p.1 #11 · Speedliting in groups?


As usual, Chuck Gardner makes a lot of sense, especially in recommending that photographers give careful consideration before buying the latest, greatest gear. I'm sure the new radio-equipped Canon flashes would be great additions to anyone's kit, but manual flash settings (whether you're using Canon's ETTL system in manual mode or studio flashes) produce consistent, reliable light in a "formal" setup - in a portrait situation when you're planning to use key, fill, and accent lighting, for example. In such cases, the advantages of greater power, similar portability (assuming manual strobe gear like Buff Einsteins with mini lithium battery packs), and close-to-equal cost might tip the scales in favor of studio monolight systems. It's hard to avoid being infected with GAS (Gear Acquisition Syndrome) when new equipment hits the market, but it may be wise to take a deep breath before breaking out the credit card.


Apr 08, 2012 at 12:59 PM
cgardner
Offline
• • • • •
Upload & Sell: Off
p.1 #12 · Speedliting in groups?


no_surrender wrote:
Chuck, once again thanks for the description, how-to, and sample images! Since only 2/3 of the flashguns I have are Canon Speedlites, I'm going to start practicing with just a fill and key light. I'd like to see what the difference is between shooting a scene (whether indoors or outdoors) with only ambient vs one Speedlite mounted in camera hot-shoe vs one Speedlite off-camera vs the two light A:B ratio.
Kevin

These are range tests I did in HHS mode but they show ambient only, one flash and two flashes outdoors:

When I test I usually start with a baseline ambient shot for comparison, exposed per the playback clipping warning to keep the highlights below clipping:
http://super.nova.org/MP/HSS/_MG_5034.jpg

A scene indoors or out either matches the 7-8 stop range of detail of the sensor or not. A backlit scene is 10 of more stops so choosing to expose for the highlights results in overall loss of shadow detail. Also everything darker than white gets reproduced darker than "normal" in the photo.

What defines "normal"? For me it is what I PERCEIVE in person in the same lighting. It differs from what the camera records because what it sees at a fixed aperture and shutter speed, whiile the eyes constantly change exposure digging more detail out the the shadows when focusing on them. The brain proceesses what the eyes see to the overall perception of the scene lighting.

When you look at a photo like the one above your reaction is likely that it doesn't look "normal". Ask yourself why. YMMV but for me what is normal looking or not is base on: 1) what I think is most important in the scene, and 2) how I expect it to look. In a photo, especially one assisted with flash the expectation would be that the foreground content, which is usually the most important, is exposed "normally" as seen by eye after the eyes adapted to the light level in the shadows. I could have opted to expose the front of card normally by opening the lens 2-3 stops, but that would have blown the highlights and make the later flash assisted shots seem less normal.

In the next shot I've added a single flash, a 580ex mounted over the camera on my Stroboframe bracket. I used 1/1 power for the test and kept moving closer until as best I could judge from the playback the foreground looked "normally" exposed — a subjective decsion not a technical measurement based on eyedropper readings.

http://super.nova.org/MP/HSS/_MG_5026.jpg

As with the ambient only shot the overall exposure with shutter/aperture was based on keeping the sunny highlights below clipping. What happens when a single flash is added from near the axis is that the exposure on the front side can be lifted separately because of the chose of ambient light angle and the fact the flash doesn't overlap the sunny highlights significantly. So the closer I got with the flash at 1/1 power the more "normal" the foreground became EXPOSURE-WISE. But note that at the same time the flash was correcting the shadow--side exposure it has eliminated the skylight modeling seen in the ambient only shot.

There several important concepts to grasp here with respect to using single flash.

1) The ambient light is not "flat". Natural skylight has direction from above and at other than midday from the side because the sky is brighter overhead that at ground level. Skylight has both a 'key' modeling direction and an omni-directional 'fill' component that work to modell objects seen in open shade and backlight on the shaded side.

2) Near-axis flash is flat, regardless of whether it is used indoors or outdoors.

3) Near-axis flash, added to natural skylight will overpower it and cancel it's natural downward modeling.

So what single flash near-axis does is solve one problem by normalizing exposure on the shaded side, but creates a new one; killing the natural 3D modeling of the skylight. Moving the single flash off axis on a stand would result in normal looking highlights on front where the flash hit, but in the shadows the flash does not hit would remain as dark as in the ambient -only shot. That might work for some situations but not for others, but most significantly the photographer has no INDEPENDENT CONTROL over the highlight:shadow ratio on the face.

Indoors with single flash off axis is no different. To the extent there is any fill in the shadows the flash doesn't hit directly with it's light rays the fill is coming from the ambient room light and "spill flill" bounce off the ceiling and walls. The shadows will be lighter in a small room than a big one because therre is more "spill fill". But there is no independent control of the lighting ratio. That's the advantage of using two flashes.

Here in the third part of my test I used two flashes, both direct 580ex in HSS mode with the off camera flash at a 45° angle to the right and about 45° higher as I might light a full face view of a person from the side.

http://super.nova.org/MP/HSS/_MG_5035.jpg

Here the first goal was to visually set overall exposure per the sunny highlights from the back rim lighting. Next, because the flash on the bracket was now acting as fill, I moved it until I could see visual separation in the shadows in the playback. I was not seeking specific eyedropper values on the target because I have no way to measure them when shooting. When shooting I'm "flying" VFR (visual flight rules) based on what I see in the playback. When I moved in to 10' I saw sufficient separation in the darker tones of the foreground.

You might note that that with the single flash I would up at 7' but here with two flashes I was at 10'. That's because now the flash on the camera is being set based on shadow detail on the front side, not normal looking highlights. If it were a portrait subject like a bride and groom in backlight I'd base the fill power on the details in the dark clothing.

I don't have a shot to illustrate this, but if you were to see a shot of just the background exposed for the sunny highlights, and the near-axis fill adjusted for the shadow detail without the "key" light you'd see "normal" looking shadow detail (i.e. the laces on the grooms shoes and buttons on the black coat) but the front of the white highlights would look flat and gray because the single flash is canceling the modeling of the skylight as in the single flash shot, and not strong enough to expose the highlights.

The "trick" to making portrait lighting and lighting overall seem NATURAL is to understand and duplicate the angles of natural light. That's why my off camera flash in the test was placed at 45° above and 45° to the right of the target, to mimic the natural modeling of the skylight the near-axis flash cancelled. Since I was using the two flashes at 1/1 power in M mode getting "normal" looking highlights on the front side was just a matter of moving the off camera light closer at the 45/45 angle until the balance between the sunny highlights from the back and the flash created highlights on the front looked "right" which is to say how I'd expect to see them by eye — normal balance of rim / key / fill.

Again for me the balance is a subjective decision based on many factors, including how much of the background is seen in the shot. Here's the same shot as above in a tighter crop:

http://super.nova.org/MP/HSS/_MG_5035_Cropped.jpg

To my eye it seems more "normal" than the wide shot. Thinking about why that's the case I concluded that when comparing the wide and tight shots the underexposed background of the wide shot was influencing my perception of the foreground. In the tighter crop there is no background context to influence my brain's decision about what in the photo looks "normal" or not.

Your brain and mileage may vary but that's how I've come to approach the use of flash outdoors.

First I determine whether or not the scene without flash will fit the sensor. On a cloudy or overcast day I wouldn't use flash the same way as this test, or at all, because it wouldn't be necessary.

If scene exceeds sensor the next question is whether or not I can find something interesting to put in the foreground within the range of the flash. If so I will put the sun to the back of the foreground subject using the sun as rim light, and then light the front of the foreground with the flash. If not I'll compose the shot in a way that the content which is in shade and falls beyond the sensor isn't significant or noticed by putting the sun at my back and the scene in light that casts the shadows down behind things.

If shooting full face views I find I can get acceptable results with a single flash on bracket because I shoot with a bracket because it creates natural downward modeling and hides the shadows the flash creates down behind the subject. The shadow on the sides of the face wind up dark, but that works to frame the face and make it look slimmer. That's true of flash on a bracket indoors or out which is why in the 40 years I've been using flash mine has rarely ever been in the hot shoe of the camera.

I'll add a second flash indoors or outdoors when I want to control the lighting ratio on the subject. For example for this outdoor shot...

http://super.nova.org/MP/ButterflyCanonExample.jpg

I put the sun behind the subject and shot from high on a ladder to surround the her with the grass of the front yard where I shot it. In that situation the flash on the bracket above the camera became the frontal key light, with the slave placed lower in front of the ladder I was shooting from about chin level with her...

http://super.nova.org/MP/ButterflyCanon2.jpg

Had I only used the one flash on the bracket I wouldn't have gotten the shadow details seen in the uniform or her shoes, which she wanted to be seen in the shot.

Here's a series of single and dual flash shots I took when friends dropped by with their daughter. I had the camera and bracketed flash handy so I grabbed two quick single flash shots:

http://super.nova.org/MP/AlexHK1.jpg

I put her in the white chair intentionally to create a white background the same distance to the flash, knowing the angle of the light on the bracket would hide most of the shadows. But you can seen the shadows with are visible are quite dark and unflattering. Seeing she was cooperative I ran downstairs and grabbed the slave on the stand, flipped the switch on the 580ex on the bracket to Master, and set ETTL ratio to 1:1 because I wanted lighter than "normal" shadows.

http://super.nova.org/MP/AlexHK4.jpg
http://super.nova.org/MP/AlexHK5.jpg

This one was shot the same as those above, but I tweeked the middle slider in Levels to make the shadows a bit darker and shifted the color balance cooler to reflect her change in attitude:

http://super.nova.org/MP/AlexHK3.jpg
Apr 09, 2012 at 06:56 AM





FM Forums | Lighting & Studio Techniques | Join Upload & Sell
    
 

Welcome back
Log in to your account
Reset passwordNo account? Register