Took this last fall indoors looking out a screen door.
Use two 580 EXII using ETTL.
Trying to get 2/1 H/S ratio
Thanks for any thoughts on how to make it a better picture
Stephen
1) Master is always Group A. That's not clear from the manual which for some strange reason doesn't illustrate the simplest and most logical way to use A:B ratios: one flash on camera as Group A Master Fill and the second off camera as Group B Slave Key light.
2) The ratios are backward from studio lighting H:S convention. The H:S ratio convention assumes even fill on both sides of the face (why S = 1 all the time) with overlapping key light. The fact the two lights overlap is what alters the contrast. A conventional 2:1 ratio is created by overlapping two lights with equal incident intensity, with the math working like this:
H:S
1:1 even fill on face (i.e. from camera axis)
1:0 same strength light overlaps creating highlights over the lifted shadows
===
2:1 highlights reflect 2x more light than shadows on face, hence a 2:1 H:S ratio.
Canon A:B ratio numbers represent relative output of the individual flashes. The metering sequence works like this:
Each Group is pre-flashed at X% of total output and metered separately, then each group is told to fire at Y% of its total output to achieve correct exposure at the selected ratio. Since Master is always A, and A is fill the ratios are reversed from the convention. Only A:B = 1:1 to 1:8 show any modeling. With A:B = 8:1 to 2:1 the fill would overpower the key light..
The image above is a ratio test I did when I got my pair of 580ex. The Master A fill on the camera is lighting both sides of the white binder, the Slave B key just the the right side. I suggest doing a test like that to visualize what the A:B ratio combinations produce. Don't get hung up on lighting by numbers. Instead pay attention to the mood the lightness and darkness of shadows create.
But to evaluate the ratio its first necessary to get the exposure correct. What I do is set the ratio to 1: 3 or 1:4 so its easy to see the modeling, then adjust FEC until white highlights are 1/3 stop below clipping. Its easy to tell when that point is reached if the subject holds a white towel...
Raise FEC until the towel begins to clip in the playback overexposure warning, then back down one-click (1/3 stop). Bingo! perfectly exposed highlights in clothing, eyes, teeth and skin. The shot about way actually done in M mode not ETTL but the procedure is the same. But in ETTL take care not to make the towel too large because it might skew the exposure (crumple it up). It takes about +1-1/3 FEC to get correct exposure with the diffusers I use. The fact FEC isn't zero doesn't mean the exposure is "wrong" it simply means the camera metering made incorrect baseline assumptions. Like the ratios, don't get hung up on the numbers, use your eyes and the playback to tell you when the highlights are correctly exposed.
Once the flash highlights are exposed correctly, then its possible to accurately evaluate the contrast on the face. The neat and impressive thing about ETTL is once you do get the FEC dialed in for the highlights you can change the ratio and the the metering will keep the highlights correctly exposed as the shadows get lighter and darker.
When shooting outdoors its important to first correctly expose the background so everything is below clipping. Evaluative metering compares ambient and pre-flash and if the background is over-exposed it can cause the metering to output less flash than is necessary for correct exposure.
That may be what happened with your shot. We've come to accept blown out skies as the price paid to correctly expose the midtones with ambient only, but when using flash in ETTL mode you need to first get all the ambient below clipping ....
Looks very underexposed perceptually per the midtones, because the sensor range is much shorter than the scene. But see what happens when I turn on the flash set to FEC =0... http://super.nova.org/TP/DR_FlashFill.jpg
The fact the ambient was exposed correctly in the ambient sense of not blowing the highlights causes the camera/flash to balance foreground and background. The histogram of just the foreground selection show it matches the camera sensor range perfectly which means detail in the entire tonal scale from brightest highlight to darkest shadow:
Here's your shot corrected in Photoshop with screen (foreground) and multiply (background) to simulate how it would have looked out of camera if you had reduced the ambient to the point where the sky was just below clipping and then had raised FEC to the point where the foreground highlights were correctly exposed 1/3 stop below clipping:
Your shot looks about 1 stop under exposed. Its also broad lit which makes the right side bright and wide looking and hides the left side in shadow. Its not a clean oblique view either - the ear is sticking out. The lighting would be more flattering if the slave flash was on the left side, about 45 degrees from the nose, short lighting the face with the camera shooting into the shadow side.
So to recap:
1) First get the ambient background level below clipping first. If shooting in Av in will require about - 2 EC
2) Increase FEC until white highlight target near face starts to clip, then reduce 1/3 stop (one click)
3) Adjust A:B ratio to taste for lighter/ darker shadows using A:B = 1:1 ... 1:8
4) Use short lighting - key light 45 from the nose (not the camera) - shooting into the shadow side
Thanks so much for all the infomation.
I set both light equal distance from me, with the off camera light at about 45 deg. off my nose. I was trying to use what i read on your web page.
On camera was set to A group and off camera was set to B group.
Not in this picture because it older but now i try to use your white towell for exposure
I am still working on this.
I have gotten more detail information from reading your post.
It will take some time to take all this information you provided thanks again.
DUDE!!!!!!!!!!!!!! We shall forever call this technique the zroin
i.e. the ear shadow on the mesh screen - subject right...
Think about this - millions of ringflash pictures against brightly colored walls, doors and crumbling concrete... Now superimpose this same shadow effect over a sunset or sunlit treeline such as yours or mountain range or...
IF the "scrim" is parallel to the sensor, your personal vigenette is going to effect the BG -EV of anything behind the subject...
IF the "scrim" is not parallel to the sensor, then the effect could control the shadow proportionally NOT in relation to the angle or distance of the BG subject...
Hmmm, think about this order - ringflash, (_8_ feet) , subject, (_3_ feet) , "scrim", (_2500_ feet) , Mt. Rushmore.
Result, Subject with ringlight drop shadow overlaying Mr.Rushmore during high-noon conditions
