The ColorRight diffuser and flash work

Drew Strickland has posted to the ColorRight promotion portion of the Pro Photo Home forums a piece with two test shots he says illustrates the superiority of the ColorRight diffuser over other diffusers when flash lighting is involved, by virtue if its greater effective though transmission.

You can see the piece here:

http://www.prophotohome.com/forum/c...ight-works-better-flash-than-competition.html

His basic premise is that with a lower transmission, "you are not getting much of the flash light into the reading" (during the WB measurement or WB test frame exposure), thus supposedly leading to an inappropriate overall WB measurement (the issue being the joint effect of the flash and ambient lighting).

It is not clear how a lower-transmission diffuser lets through less of the flash light but (presumably) just as much of the ambient light.

A discussion of metering is also given:

"The meter reading doesn't really know how to account for the amount of flash that will be added back into the shot. Are you bouncing the flash? How high is the ceiling? How far away is the subject, really?"

I'm not sure what all that is supposed to mean. But in any case, with through-the-lens metering, the effect of the transmission of the optical chain should not have any effect on the resulting photometric exposure (assuming metered exposure), nor on the relative proportions of the photometric exposure attributable to the ambient and flash components.

Thus I cann't find any basis for believing that the greater transmission of th ColorRigh was advantageous in this case.

(It does come into play if the light is so feeble that a Nikon camera refuses to make a WB measurement, but that was evidently not a factor in this demonstration.)

What Drew neglected to exploit was a very interesting implication of his two test shots. One used WB correction based on a reference frame ("from the camera position" with an ExpoDisc diffuser, and the other the same but using a ColorRight diffuser. Critical to the story is that the background was white (evidently very neutral) paper, filling the frame.

The correction using the ColorRight reference frame was noticeably "more accurate" than that with the ExpoDisc.

Now what does that mean?

Well, first, we cannot expect to get an accurate WB measurement from the camera position using a "cosine" diffuser (such as the ExpoDisc). And in fact, it is not recommended by its manufacturer for that mode of combat.

But if we handily have a neutral background, we can make a predictably-meaningful WB measurement from the camera position with a diffuser if we use one with a narrow acceptance angle, so that most of the light it gathers is from the neutral background, and little from surrounding items that may not be neutral.

Thus this in fact is a situation in which a narrow-angle diffuser (like the ColorRight) is in fact predictably advantageous. (Recall again that this is contingent on the presence of a neutral background, a "giant gray card".)

So, Drew, you should blow that horn.

Glad to help you out, Drew.

In defense of Drew's story

This is in regard to Drew's assertion that the higher transmission of the ColorRight diffuser is advantageous in a flash situation.

I had said that the metering would produce the same mix of flash and ambient photometric exposure for the test shot and the actual shot regardless of the transmission of the diffuser. Thus there would be no consequence of the diffuser transmission.

Drew points out that in a Canon EOS camera, operating in P mode, the metering system will not invoke a shutter speed slower than 1/60 sec or an aperture smaller than f/4. Thus, in many low ambient situations, the "ambient component" is underexposed.

Any "attenuation" of the light during measurement (from less than 100% transmission of the diffuser) will not change the flash component of the mix (since the ETTL flash metering system will bump the flash output to still produce the standard photometric exposure from the flash), but the ambient photometric exposure will not stay constant (since the shutter speed and aperture are already at their limits under P mode).

Thus the "measurement" exposure will be "flash-rich" compared to the actual exposure, typically leading to an "overly blue" incident light WB chromaticity determination (and thus an "overly yellow" corrected image).

This phenomenon is enlarged by greater attenuation (lower transmission) of the diffuser. Thus, in an ambient plus flash scenario, with a low ambient (as is typical for indoor flash situations), and with the Canon P mode in use, a greater transmission for the diffuser would in fact be advantageous.

The same would pertain if we used M mode for the flash work (as I often do).

Sorry I didn't think of that.

But I've drawn Drew into an actual technical discussion! I'm proud of ya, kid.

Best regards,

Doug

I thought I was wrong but I made a mistake

Just a final note to fill in a gap in my previous message.

Drew Strickland was quite right to point out that the shutter speed (and even aperture) limits imposed by the metering in Canon EOS cameras in P mode can produce an inconsistency of the mix of "flash" and "ambient" light between the WB measurement frame and the actual shot. I explain this in further detail in my previous message. Indeed, the extent of this phenomenon should increase for greater attenuation in the WB measurement diffuser (lower effective total transmission)

The direction of this phenomenon would be that the WB measurement frame would be "overly rich" in the flash component, biasing the WB chromaticity measurement in the direction of the flash illumination at the expense of the ambient illumination (again, compared to what would obtain for the actual shot), more so for lower-transmission diffusers.

But it turns out that this does not support Drew's original assertion, which I challenged:

"When using a flash with the expodisc [sic] you are not getting much of the flash light into the reading."

The implication of that is that, with a higher-attenuation diffuser (i.e., the ExpoDisc diffuser), the WB chromaticity measurement would substantially underrepresent the flash illumination compared to the ambient.

I (again) find no merit to that observation whatsoever.

The P mode phenomenon that Drew later called to my attention would, as I mentioned above, suggest an overrepresentation of the flash illumination in the WB reference exposure, increasingly so with a higher-attenuation diffuser, the exact opposite of his stated premise.

And evidently the ruse almost worked on me! Well, I'm an old guy, and not always too quick on the take in this tricky technical stuff. It didn't occur to me that I had been duped until about 3:00 am this morning.

Best regards,

Doug