White Balance Info from Camera

[Slobodan Blagojevic]

I have a Canon 60D. I was shooting an event the other day and used a flash wb all the time. However, upon transferring the files to LR3, two things were strange:

- the wb in LR was not labeled as flash, but "as shot" (ok, that might not be the strange part, perhaps that is how it always was, just I did not pay attention)

- but the real strange part is that "as shot" is not 5500 K, as the flash wb is

So, the question is:

- Is LR disregarding the info from the camera
- Or, perhaps, my 60D is not "listening" to my instructions (to use flash wb) and is nevertheless using auto wb?

[eliedinur]

The color of the light from a flash varies according to the strength and duration of the flash firing (which in turn varies according to information sent from the camera regarding distance to focus point, reflectivity of the main subject and aperture set). Canon speedlites feed back to the camera data about the color of the flash and if the WB is set to AWB or Flash the WB will be calculated accordingly.

If a non-Canon flash is used, Canon's default value for the Flash WB setting is 6147 K. However, it is not unlikely that LR, when reverse calculating the temperature from the WB multipliers for the "As Shot" reading, will display a different number.

LR's Flash WB preset is no more than a rough guess, "one size fits all", approximation.

[Slobodan Blagojevic]

Thanks elied.

Your knowledge of how Canon flash works is surely better than mine. I was aware that Canon's eTTL can measure color temperature and adjust it, but I thought it is to take into account, say, colored-wall reflections, and bring it back to the "standard" flash WB of 5500K. Hence my expectation that when I set camera's WB to "Flash", I shall see it in LR both as "Flash" and 5500K. However, most of my files open at around 6300K, +-200-300K.

There was one extra factor in play I did not mention: I was using Garry Fong's  Lightsphere diffuser (the opaque white one) with the white dome attached and positioned directly (i.e., not bounced). I guess that might have "fooled" the flash to consider it non-Canon, or otherwise throw the WB off?

Another question, not related to LR though: when I use the diffuser on my Canon 580EX and 580EX II, it seems that I need to adjust flash exposure on camera between +1.5 and +2 f/stops, to account for the loss of light due to the diffuser. What puzzles me though it that those flashes are supposed to be TTL (i.e., measure light falling on sensor through-the-lens)... if so, shouldn't the camera/flash combination determine the proper exposure even if bounced and diffused, without me having to adjust it so drastically manually?

[Ellis Vener]

- Is LR disregarding the info from the camera

No, it is telling you the WB values that were encoded by the camera, and how the camera interprets light which don't conform to what Adobe's values for a hypothetical (and perfect world) Flash WB are: 5500˚K and no green/magenta tint. I've worked with flashes from the ridiculously cheap and small to the ridiculously expensive and powerful and whether with or without a light modifier I've never seen one that perfectly matched  that theoreticall standards. The closest have been the PAul C. Buff, Inc. Einstein 640 monolight and the Broncolor Scoro and Grafit pack and head systems. Even with high end pack and head systems you can see some slight variation from head to head.

And of course as I said earlier, cameras (i.e. sensor and processors and programming) make a lot of difference.


- Or, perhaps, my 60D is not "listening" to my instructions (to use flash wb) and is nevertheless using auto wb?

That is unlikely.

Try shooting an Xrite 24 patch color checker and building a custom profile for your camera, applying that profile and then white balancing. And then you can start contemplating the whole "Accurate Color " vs. "Pleasing Color" conundrum. My thinking is that camera makers tend to bias more towards what they think is "pleasing color. After all their goal is a large base of happy customers.

[Nigel Johnson]

Another question, not related to LR though: when I use the diffuser on my Canon 580EX and 580EX II, it seems that I need to adjust flash exposure on camera between +1.5 and +2 f/stops, to account for the loss of light due to the diffuser. What puzzles me though it that those flashes are supposed to be TTL (i.e., measure light falling on sensor through-the-lens)... if so, shouldn't the camera/flash combination determine the proper exposure even if bounced and diffused, without me having to adjust it so drastically manually?

Slobodan

Are you sure that the diffuser is not reducing the light so much that the flash is firing at maximum power and is still not producing enough light for the aperture you have set? Other than that I can't immediately think of an answer to your query.

The way E-TTL II (as used by your camera) works is quite complex, here are a few useful links to sites that describe how Canon flashes works, perhaps you will find a solution:

http://photonotes.org/articles/eos-flash/
http://super.nova.org/DPR/Canon/
http://www.learn.usa.canon.com/resources/articles/2011/speedlite_tip_pt1_article.shtml?categoryId=121
http://www.learn.usa.canon.com/resources/articles/2011/speedlite_tip_pt2_article.shtml?categoryId=121
http://www.learn.usa.canon.com/resources/articles/2011/speedlite_tip_pt3_article.shtml?categoryId=121
http://web.canon.jp/imaging/flashwork/index.html

Regards
Nigel

[Rhossydd]

I was using Garry Fong's  Lightsphere diffuser.... I guess that might have "fooled" the flash to consider it non-Canon, or otherwise throw the WB off?

The camera couldn't know that.

[digitaldog]

The default White Balance in ACR and Lightroom is "As Shot" which means, whatever your camera thought the WB should be. Regardless of whether your camera’s menus were set to "Auto WB" or or Daylight, etc, they all come into ACR/LR as "As Shot".

[Ellis Vener]

"Your knowledge of how Canon flash works is surely better than mine. I was aware that Canon's eTTL can measure color temperature and adjust it..."

I have been looking all day - on the web, in camera manuals, in Speedlite manuals, in Syl Arena's "Speedlighter's Handbook" -to find something to corroborate that claim.

The closest thing i have found is that  In AWB white balance mode, connecting a Canon Speedlite directly to the hotshoe will automatically switch the WB to Canon's Flash WB settings and absolutely nothing suggesting that E-TTL II does anything to shift the WB of a photo to compensate for different flash durations, or gelling/filtering the light or to compensate for the color  of a surface you might bounce the light off of. On page 17 in the english language manual of the 430 EX II there is an instruction that contradicts this claim: "The wall or ceiling should be a plain white color for high reflectance. If the bounce surface is not white, a color cast may result in the picture."

On page 23 of the same manual "When the flash fires, the color temperature information is transmitted to the EOS DIGITAL camera. This feature optimize's the flash picture's white balance. When the camera's white balance is set to <AWB> or <(lightning bolt symbol)> it will work automatically."

I'd love to be proved wrong so if you point me to an authoritative source I'd really appreciate it. Thanks in advance.

[Slobodan Blagojevic]

Ellis, as i said, I do not claim I know much about Speedlites. What I wrote was my remembrance of something I read some time ago, and of course I might have read it wrong or remembered it wrong.

However, reading one of the links in the Nigel post above (thanks for the links Nigel), "flashwork", I did find something that might explain why I remembered it that way (under E-TTL II Technology):

Color temperature compensation ensures faithful color reproduction

When an EX Series Speedlite is mounted on an EOS DIGITAL camera, it automatically informs the camera of the color temperature of its illumination. This enables the camera to calculate white balance with extreme accuracy.

* White balance must be set to "automatic" or "flash use" mode.

[Ellis Vener]

 I asked Chuck Westfall of Canon last night about how Canon handles Speedlite color and he was kind enough to respond:

"Most shoe-mount flash units are capable of controlling automatic flash exposure very precisely by
varying the duration of the flash burst. This method has the disadvantage that color temperature levels
vary slightly according to the duration. Battery power levels also have an effect, with higher power
levels resulting in slightly cooler color temperatures than lower power levels. For these reasons
among others, installing colored gels or panels over a flash head can only control color temperature
to a limited extent. Canon recognized this issue several years ago and resolved it by adding a feature
called automatic color temperature compensation to its EX-series Speedlites starting with the 580EX.
With this feature, a circuit in the flash reads the battery power level and the flash duration for every
shot, and then applies a color temperature compensation factor to the white balance data in the
resulting image files. This has the effect of equalizing the color temperature for all flash photos taken
with the 580EX or newer Canon Speedlites with EOS digital SLRs. "

Based on this authoritative information I removed my last two posts from this thread. It just goes to show you can  learn something everyday.