• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Canon wireless flash - ratio operation

Doug Kerr

Well-known member
Introduction

In another thread, Ben Rubinstein had contributed descriptions of various features of Canon EOS cameras that were not well known.

As part of one discussion, he commented that, in using the Canon wireless multiple flash system, when setting a ratio between the exposure contributed by the A and B group units, a A:B setting of 1:2, which might seem as it could be expected to produce "one stop" greater exposure contribution by the B units compared to the A units, would actually produce more like 1/2 stop of difference.

Ben suggested that this was a consequence of the ratio settings on the flash units of interest being settable in 1/2-stop steps.

I demurred, pointing out that indeed the available ratios were in 1/2-stop steps, but that should not affect the significance of the ratios.

For example, working only from a ratio of 1:1 upward, the available ratios on a typical unit are:

1:1, 1.4:1, 2:1, 2.8:1, 4:1, 5.6:1, and 8:1

Here, the "full-stop" ratios, shown in bold, are actually marked on the ratio scale, while the intermediate "1/2-stop" steps are not marked, only having their places held by little dots. We would have no reason to believe that a ratio of 1:2 would be intended to produce a "1/2-stop" difference in exposure contribution (any more than a set aperture of f/4.0 would have a different significance whether the camera allowed apertures to be set in 1/3-stop, 1/2-stop, or full stop increments).

Ben agreed that my interpretation made sense, but pointed out that, nevertheless, his consistent experience had been that the apparent affective exposure from the two flash unit groups, for a set ratio of 1:2, seems to be more like 1/2 stop (that is, a ratio of 1:1.4)

We agreed that the are many strange things that go on in "the mind of E-TTL".

I thought that I would do a little testing to look into this matter.

The test setup

This figure shows the test setup.

Flash_Ratio_G02529F.jpg

The target comprised two leaves of white Foamcore board at an angle of 90°, with the dihedral angle toward the camera and the leaves extending away from the camera.

The test camera was a Canon EOS 40D. It carried a Speedlite 580EX II flash unit, set to be the Master in a wireless flash network. It was aimed at the ceiling, and set not to fire for exposure.

On each side was a Canon Speedlite flash unit on a light stand, a 550EX on the left and a 420EX on the right. Each was set to illuminate one of the target leaves head on. The distance from the face of the flash head to the target leaf was 36" in each case.

Both side units were set to the Slave mode, on the same channel.

The premise

The premise of the test program was that the "contributions" of the two flash units to the overall (perhaps "average") illumination on the test scene (and thus on exposure) are physically separated by the test setup so that they can be ascertained, separately, from the exposure results on the two leaves of the target.

Flash metering

The test shots were all taken with E-TTL flash metering in effect, but in the average, rather than evaluative, mode. My thought is that there might be fewer intrusions from the inscrutable "mind of E-TTL" that way.

The test approach

In each series of tests, with a specific operating mode of the wireless flash system, a test shot was taken (when flash ratios were in effect, for various ratio values).

The resulting images were examined in an image editor, and the relative luminance of a small patch in the center of each target leaf was determined. Then, assuming that the sRGB color space model was being rigorously followed, the implied photometric exposure on the two leaves were ascertained. Since the two leaves had the same reflectance, this then told us the relative illuminate-time product on the two leaves. We took this to be effective "exposure ratio" between the two flash units.

Tests and results

Series 1 - Flash units in one group

Here, both flash units were set to group A. The master was set for basic operation (no ratio).

Comment: In this case both flash units receive the same output level command (based on the measurement of the joint pref lash from the two units). Note that this is in absolute terms, not as a fraction of the "full" output of the individual unit (those "full" output levels being different between the 550EX and the 420EX).

Since the two units were at the same distance from their respective target leaves, and were set to illuminate them at the same angle of incidence, we would expect a consistent illuminance-time product on each leaf. With the leaves having consistent reflectance, we would expect equal exposure results on both.

Result: The implied exposures on the two leaves were equal within about 0.1 stop (higher on the left).

Series 2 - Flash units in two groups - no ratio

Here, the left-hand flash unit (550EX) was set to group A, the right-hand unit (420EX) to group B. The master was set for basic operation (no ratio).

In this case, we would expect the metering system to strive to produce the same contribution to overall exposure from the two flash units, sending to each flash unit an output level command that, based on the measured return from on the pref lash with that unit, should produce equivalent exposure contribution.

Result: The implied exposures on the two leaves were equal within about 0.2 stop (higher on the left).

Series 2 - Flash units in two groups - A:B ratio in effect

Here, the left-hand flash unit (550EX) was set to group A, the right-hand unit (420EX) to group B. The master was set for A:B operation (A:B ratio).

Comment: Here we would expect the contributions to overall exposure from the two flash units to be related by the A:B ratio set on the master. With a ratio setting of 1:2, we would expect the implied exposure on the right-hand leaf to be one stop greater than that on the left-hand leaf; with a setting of 1:4, a full stop greater.

Result:

In this table the first column is the A:B ratio setting; the second column shows the "expected" exposure of the right-hand leaf as compared to that of the left-hand leaf (in stops); the third column shows the observed implied relative exposure of the right-hand leaf as compared to the left-hand leaf, in stops.

Code:
4:1  -2.0  -2.2
2:1  -1.0  -1.1
1:1   0.0  -0.2
1:2  +1.0  +1.4
1:4  +2.0  +2.5

Conclusion

In this particular idealized test setup, with a simplistic test protocol, and a single round of tests, we find that the apparent ratio between the exposure contributions of the "A" and "B" flash units was consistently slightly greater than the ratio implied by the A:B ratio setting, by as much (in one case) as 0.5 stop.

We make no effect to explain why this was observed, nor to extrapolate this to actual practice.

Best regards,

Doug
 

Ben Rubinstein

pro member
Lot of work gone into that, I'm willing to bet if it had been two of the same flash unit you would have had different results. Let's face it, those ratio's are only ever going to be approximate which sucks but it's something at least, I'd be interested in seeing how accurate the nikon system is in comparison.
 
Top