White Balance Calibration

I have a dilemma, I am attempting to calibrate my Sekonic light meter and that is pretty easy but I am having a problem with WB between 3 pieces of software.
1. ACR 6.5
2. DxO Optics 6.6
3. Canon DPP

I have set all color spaces to sRGB including the camera for uniformity)
Here is what I have done ...

1. I took a WB reading to set my Canon 1D MK IV to basically a WB in the ball park.
2. I took a Lastolite grey target and lit it with a 50" Octobox to get balanced light across the grey target.
3. I metered with my light meter and set the camera to the light meter readings
4. Set camera to meter readings (1/60th f/8 iso 100)
5. Took the shot.

When I loaded the image into DPP the RGB settings were 129, 127,125 (pretty much perfect) [See image of RGB numbers] and everything was zeroed out.

When I loaded the image in DxO Optics 6.6 the RGB numbers were 130, 130, 130 (again everything has been zeroed out and those are the raw numbers with the software doing NO adjustments.)

When I load the image into ACR 6.5 however it is a completely different story.

ACR 6.5 gives me the following numbers 67, 67, 66 and wants to up the blacks from 0 to 5, wants to move brightness from 0 to 50% and wants to change the contrast as well.

Can someone give me an idea of what is going on here, if I use ACR all my images are going to be blown out (and have been which is why I am going to all of this effort to get it correct.

I am adding the 2 screen shots of DxO Optics 6.6 and DPP in this post and then in the next post I will add the screenshot of the ACR histogram and the actual image.

I do pretty much all of my raw processing in ACR and then I go in Photoshop CS5, so this is an issue for me.

Hopefully someone can assist me here,

Many thanks

Gavin

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

The other 2 files as promised (ACR 6.5 and the original file)

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

Do you have the "lighting/hdr" module of DXO enabled? If so, take a look at the values that are set there - that is what may be boosting your histogram. Same with ACR - there is a whole lot going on that can confound the expected results. Sometimes it is best to start by setting all sliders to ZERO and all curves to straight, linear. If there is a choice for rendering color profile, or picture style, set it to neutral. All three applications you are using have these controls. Figure out what reference you are striving to achieve and try to make sure that you know how to make it happen in each piece of software you are using.

Good luck, don't drive yourself too crazy. I know this may sound like a bad idea, but you also may want to consider using dcraw, where you can get a little less behind the scenes monkey business going on when conveying the raw data to RGB. If you use mostly ACR, then maybe you want to use it as your reference. Irwin M Fletcher, you choose.

Kirk

Thanks for your reply, and no I do not have the lighting / HDR module enabled, in fact I disabled all modules so that I could establish a baseline. I looked at the settings for DPP and they are all disabled as well, and I zeroed out all of the settings in ACR 6.5 which is what gave me those low numbers.

The reflective meter in my camera says that everything is correctly exposed as did the light meter when I used it's indecent meter.

I am concerned that when I work on my images I am going to over expose them as I have been trusting what ACR and PS CS5 have been telling me.
I did open a thread at Adobe and will see what that brings, and will post any pertinent info here.

Lastly if you look at the image that I am using it is uncorrected and it is certainly not 50% underexposed (which is what ACR is telling me)

Again thanks for the help.

ACR/Lightroom uses different multipliers from DPP so the resulting numbers are different but they are consistent. Its consistency in the RGB channels that you should look for, not an exact match with DPP.

ACR's default settings of Brightness=50, Contrast=25 should not affect how WB is determined. To test this crop your WB target in ACR to eliminate the black border; I'd wager that the RGB reading is the same.

PixelMagic, it is a little better, the ACR numbers post crop are 103, 103, 102 with Brightness=50, Contrast=25 and everything else at Zero.

Now if I am in CS5 I can create a new layer and fill it with 50% gray (edit, fill, 50% gray) and the numbers on that fill are all 128.

I cropped the original image, and then created a 50% gray image and copied the crop onto the 50% Gray to see if it is just numbers that are messing me around, but the crop of the original image is definitely darker by about 1/2 stop with the crop.

Getting there, but not just yet.

There has to be a simple explanation for the discrepancy, ACR and Adobe are far to good products to be this far wrong.

So it would appear that you are shooting with flash - at your camera's sync speed? Are you metering for the flash output? You want to make sure that you are shooting at sync speed to eliminate any ambient light, illuminating the target with the flash. Put the meter on a light stand or tripod, point it at the camera, and position it next to the gray target so it receives the same illumination as the target and does not cast a shadow on the target. Meter the flash and set your camera to that aperture at the chosen ISO (the shutter speed should be fixed to sync).

In an ideal world, the camera, set to the metered aperture, should produce the proper exposure of the target. Otherwise, you will have to bump up/down the exposure (in 1/3 increments, most likely) until the gray of the target, as produced in your raw converter with the settings you like, gives you the results you want. The difference between the metered reading and the corresponding camera setting is the "compensation" that you need to add to/subtract from the meter to get the exposure nailed. Some meters let you enter this shift into the meter calibration so that you don't have to remember to alter the metered reading to compensate when setting the camera exposure. I usually have to overexpose about 2/3 stop with my 5DmkII for the raw converters I use (DXO, RPP and Raw Developer).

Is this how your are shooting each image? Just trying to understand the input to the system.

kirk

I am shooting below sync speed (1/60th) in a dark / near black room with flash.
I have my light meter calibrated to my camera / lens combo [I am assuming as I used DxO 's numbers to calibrate it] and I can reduce or increase my light in 10th's of a stop. My light meter is saying 1/60 at f8.0 at iso 100 which is what I shot this at. My light meter has the ability to be calibrated which is what I was doing when I ran into this issue.WB has been set using a X-rite color checker passport so I have to assume that is pretty accurate then I am using a huge octobox for even light. The lights in the studio get turned off after the timer has been triggered on the camera so when the flash is deployed, everything is pretty much dark and there is no ambient light to mess with this.

Thanks again for your assistance and insight

Gavin

Three points:
1. Using Lightroom - same as ACR - I have to raise Brightness to 64 in order for it to show 18% grey at the proper level when the camera is set to 0 EC. Not a big deal since except for skin-tones I set midrange tones by eye.
2. 18% grey is not 128 (50% of 255). In sRGB it is 119.
3. Every converter has its own TRC. If you are calibrating your meter against DxO, you cannot expect perfect agreement from ACR.

BTW, what does this have to do with white balance?

Thanks Tzalman, it has nothing to do with WB, I outlined my process including WB so that all bases were covered.

Sure each piece of software has it's own TRC however when calibrating a light meter one would expect that the numbers would be close to the same as yuo have to select a range of numbers to get the exposure correct (be it 128 or 119) and based on what you are saying that at 128 in sRGB my light are too hot thus my light meter needs to be adjusted

And you hit on a key point, skin tones, I do portraiture among other things so I really would like a it to be accurate and not blown out, when using either strobes or speedlites, and sure I can adjust in post, but why should I correct base or consistent over exposure in post, when I am putting in the effort to use a light meter and get it correct in camera.

I will however re-calibrate my light meter to 119 in ACR, and take a look at using 64% brightness.

Thanks

Also:

http://en.wikipedia.or​g/wiki/Light_meter

And specifically read the section regarding Calibrated Reflectance and Exposure Determination with a Neutral Test Card.

Consider that in your image of the gray card, the camera is not seeing the full reflected illumination of the target because the soft box appears to be placed camera right (based on the shadow behind the target) and the target is angled away from the camera. The illumination of the target and making sure that the full reflected light from the target is reaching the camera is critically important.

kirk

Very interesting Kirk, I am going to have to put those formulas to test. Basically it appears that I am attempting to compare apples to Oranges and attempting to make sense of it, which is not going to happen
Thanks.

Also - when you finally work out what difference, if any, there is between your meter and your camera, you, hopefully, are not changing the meter's calibration, but the compensation, right? I'm guessing you do not want to mess with the calibration.

What you are trying to do is understand how to measure incident light and reflected light and how those measurements correspond to the settings you need to dial in on your camera. This is a good exercise you are going through. Once you get middle gray worked out, put a black velvet object up there as your target and then a white diffuse surface up there.... Then something like a white terry cloth towel that has texture. As long as your lighting set up has not changed, your Sekonic will always read the same, but your camera will not. That's when the fun really begins.

kirk

Absolutely correct, you change the compensation, though it is discussed as calibrating the meter.