• 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!

New review of white balance tools

Doug Kerr

Well-known member
A new article by Sara Frances, M.Photog.CR., "An Investment in White Balance that Appreciates Right Now", has recently been posted on the Pro Photo Reviews and Articles section of the Pro Photo Home forum. In the article, Frances discusses the need for attention to white balance matters, gives a checklist to guide those taking steps in this direction, and in a chart, lists 10 white balance tools and gives them a "color neutrality rating" on a scale of 1-5.

The article can be viewed here:

http://www.prophotohome.com/forum/p...ment-white-balance-appreciates-right-now.html

Before I comment further, let me mention two terms used in the descriptions of the tools' applicability, "incident WB" and "reflectant WB".

We often see the terms "incident" and "reflective" used (inappropriately) for the distinction between making a white balance measurement, using a "diffuser" on the camera lens, at the subject location vs at the camera location for actual shooting.

But that cannot always be be the distinction intended by Frances, since she also applies it in the case of such things as white balance targets. In those cases, she likely means measurements made from a reflective target. In other cases (for diffuser tools), it apparently means "reflective" in the sense described above.

That having been said, let's look at the comparison chart. The tools are divided into three categories: Disks, Targets, and Cards. The "disks" listed include the Colorright [sic], CBL (Color Balance Lens), ClearWhite, and ExpoDisk.

The common element of the "Disks" is that they are round. It may seem trite to mention that, but I must in order to emphasize that this is the only common factor - they are not of a common functionality. Three of them are used in front of the camera lens (as a "measurement diffuser"). But one of them, the CBL, is, despite its name, a target disk, a fairly high reflectance "white" on one side and a "neutral" gray on the other side.

Incidentally, it is hard to find this out this basic fact from the extensive pseudo-scientific material on the CBL manufacturer's Web site - nowhere does it give the slightest hint as to how to use the device. (Frances quite accurately says that "the booklet and website [are] extensive but difficult to understand".)

In any case, Frances must have figured out how to use it, since she includes an brief assessment of its performance in her table: ". . . excellent result placed anywhere, any angle in incident light, consistent in many circumstances."

For the Colorright [sic] disk, she says: "optimal use incident, fair to good reflectant - best of all devices under green leaves". Here, she apparently uses "reflectant" to refer to the so-called "reflective" technique for measurement with a diffuser, in which the diffuser-equipped camera is placed and aimed as it will be for the actual shot.

Overall, she gives the Colorright, CBL, and ClearWhite (a simple milk-white acrylic diffuser) color neutrality ratings of 5, while the ExpoDisk (also a diffuser) gets a 4.

We have no hint as to how the ExpoDisk fell short. It is described as "medium consistent in varying situations". Perhaps this is a reference to the fact that it cannot be expected to perform well in the "from the camera position" ("reflective") mode, for which it is not designed, and whose use is not recommended by its manufacturer.

In fact, nowhere in the article is there the slightest hint of how the testing was done (the author reports it as "per my 6-month on the job testing"), nor what criteria were used in assessing the performance of the different tools and the assignment of color neutrality ratings.

I won't go into any detail on the "Targets" and "Cards" sections of the chart (the distinction between the two seeming to be pretty arbitrary). All of these items received an color neutrality rating of 4 (the apparent scale of 1-5) except for a random piece of Styrofoam packing (reported in the "Cards" section), which was given a color neutrality rating of 3.

Frances comments that this is "better than coffee filters or plastic caps", and lists its disadvantages as "crumbles easily, non-traveling".

Evidently based on the ColorRight tool receiving a color neutrality rating of 5 on the Frances scale (along with the CBL thing and the ClearWhite diffuser), the manufacturer of the ColorRight announced, "COLORRIGHT Ranks Tops out of 10 White Balance Products in New Pro Review".

I conclude with this equation from the autobiography of the author that appears at the end of the article: "I’ve spent my whole career working very hard to make what I do appear effortless." Sara, I think you have succeeded.

**************

Since it figures into this story, let me take a moment to return to the CBL (Color Balance Lens), a truly curious item. Its front face is (apparently) opaque white plastic, into which are molded a number of short straight "ridges" of various cross sections. These are said to give the device its superior properties with regard to acceptance of light. It may well be that the purpose is to attain a nearly Lambertian response to light arriving from differetn angles.

In any case, the entire device is manifestly opaque (since the rear face is opaque gray plastic).

The gray rear face used to be thought of as just the back of the device, in which a product "description" is molded, but they later realized that it was a neutral gray target.

It is interesting that on the CBL manufacturer's Web site, despite the presence of a profusely-illustrated 18 page "Product Training Manual", full of pseudo-scientific jargon, nowhere is there the slightest mention of how one uses the device. I have assumed that, since it is opaque, the intent is for it to be used as a reflective target.

**************

Well, that's the color science report for today.
 
Last edited:

Michael Tapes

OPF Administrator/Moderator
Since every WhiBal is essentially perfectly neutral (we measure every one to stringent specs), I am wondering how I can raise my neutrality score from 4 to 5. This will keep me up long nights!
 

Doug Kerr

Well-known member
Hi, Michael,

Since every WhiBal is essentially perfectly neutral (we measure every one to stringent specs), I am wondering how I can raise my neutrality score from 4 to 5.
Well, I think to get your neutrality score up, you will have to give up this foolish belief in colorimetry and instruments and place your immortal soul in the hands of a higher power (I think that means something in the range of 1.8 to 2.2).

Now as to getting the neutrality score of the WhiBal target up . . .

Do note that one can see Russia from Georgia (especially from Ossetia).

I am so grateful something like this happens every week or so. It's such a blessed step toward rationality from all the election campaign stuff!

My Color Parrot is working well, although Carla has to remind me to use it from time to time. It seems that when I don't, I get coffee stains all over my desk.

Now if you haven't read the CBL manifesto, you really should. It is here:

http://www.internationalsupplies.com/cbl/CBL_Product_Training_Manual_0307.pdf

Best regards,

Doug
 

Jerry Levin

New member
Hi Doug, Michael -

I own two balancing acts: an Expodisc, and a solid grey reference card (solid - plastic all the way through).

However...

We shoot a lot of interiors. Lots of bizarre, mixed ambient light sources, with some of our own lighting thrown in to open up shadows, etc. Lots of very large contrasty, spaces, lots of point-source light, lots of reflections off of nice shiny walls. You get the picture. Think boat interiors with teak wood + tinted windows + wall sconces in lovely yellow, flo fixtures in the galley/kitchen, and so on.


For our grey card, simply placing it in the area is not very helpful of course. There rarely is a "right" place to put the grey card that avoids all the glare, is in the right zone (i.e. the actual brightness I'm exposing for in the camera).

We've not been real happy with the Expodisk (realizing it should be at the location the camera is aimed at, not where the camera is shooting from), but knowing where to place a grey card has been even worse.

I looked at the WhiBal site, but don't see how it will fare any better than the grey card we already own (and hope you will correct me). Seems that if it's not in exactly the right spot (if there is such a place), it'll be subject to the same problems I just described.

Also, FWIW, a number of links on the WhiBal site are throwing 404 errors - might want to check that (e.g. trying to click on the Purchase button, or on the combo photo on the right).

Hoping for some education...

Thanks!
 

Michael Tapes

OPF Administrator/Moderator
Hi Jerry,

You have it pretty much nailed. There is no silver bullet. So let me address a few of your issues.

Is WhiBal better than the gray card you are using?.....I do not know. WhiBal is perfectly neutral to a very tight tolerance. I do not know about your gray card. And the WhiBal surface is a very matte, but along with a glare detector in the Black part of the label, so you can see the glare that will affect your "reading, even thought you can not necessarily see it on the gray area.

In that situation, I will take several WhiBal shots at various areas of interest, and of course shoot Raw. I would also take several "real" shots bracketing the exposure. I would then composite the files in Photoshop after having converted the various exposures and various White balance settings. Smart Objects is used well here too.

In this way you can combine the various developments such that accurate color or artistic license can guide you to the final mix and you can get the rendering of the scene that *you* want. It is either this, or doing fancy lighting and capturing it in camera, which is much more risky for that type of work IMO.

I do not believe that an ExpoDisc type of product has any advantage in this situation. Anything that you can do with an ExpoDisc type of product you can do with WhiBal, but the reverse is more difficult.

Hope this helps.

Regarding 404 errors...Thanks. I have a new site at www.rawworkflow.com. It would appear that the WhiBal.com reference below is not being directed there. You are seeing the old site which in deed is very broken based on moving our host. I will correct these things. Thanks for the heads up.
 

Doug Kerr

Well-known member
Hi, Michael,

And the WhiBal surface is a very matte, but along with a glare detector in the Black part of the label, so you can see the glare that will affect your "reading, even thought you can not necessarily see it on the gray area.
What is "glare"? And how does it "affect your reading"?

Thanks.

Best regards,

Doug
 

Michael Tapes

OPF Administrator/Moderator
For the most part the matte surface of the WhiBal eliminates glare as a factor. However there is a point at which the angle of the camera/WhiBal/Light Source will be source that the light source is reflected directly into the camera point of view. At this point the reflection of the light source affects the color of the WB reading because we are reading the reflection of the light, rather than the WhiBal. Let me emphasize that this is a very minimal effect with the WhiBal, but to completely eliminate it we have a "glare detector" (Black surface with very high gloss) that will indicate glare at such an angle that the WhiBal reading "might" be in danger.

My definition of glare is the when the reflection of a light source (generally a large source) overpowers the diffuse reflection of light from an object, thereby rendering the light source rather than the object (or partially so) from a specific point of view.

Putting aside that I will not get a job with Websters, is that clear?
 

Doug Kerr

Well-known member
Hi, Michael,

For the most part the matte surface of the WhiBal eliminates glare as a factor. However there is a point at which the angle of the camera/WhiBal/Light Source will be source that the light source is reflected directly into the camera point of view. At this point the reflection of the light source affects the color of the WB reading because we are reading the reflection of the light, rather than the WhiBal. Let me emphasize that this is a very minimal effect with the WhiBal, but to completely eliminate it we have a "glare detector" (Black surface with very high gloss) that will indicate glare at such an angle that the WhiBal reading "might" be in danger.

I understand. The concern is with the possibility that the target is not perfectly Lambertian but has some (hopefully negligible) specular reflection component. If a light source was positioned so that the light from it followed this specular path to the measuring instrument (the camera, in this case), that source's contribution to the "measured" chromaticity of the incident illumination might be disproportionate.

The specular black area on the WhiBal target exaggerates this effect and lets us see if the geometry of the lighting even possibly raises such a hazard.

My definition of glare is the when the reflection of a light source (generally a large source) overpowers the diffuse reflection of light from an object, thereby rendering the light source rather than the object (or partially so) from a specific point of view.
A good definition in this context (or it would be if all its words were present - it seems to have lost a couple).

Thanks.

Best regards,

Doug
 
We shoot a lot of interiors. Lots of bizarre, mixed ambient light sources, with some of our own lighting thrown in to open up shadows, etc. Lots of very large contrasty, spaces, lots of point-source light, lots of reflections off of nice shiny walls. You get the picture. Think boat interiors with teak wood + tinted windows + wall sconces in lovely yellow, flo fixtures in the galley/kitchen, and so on.

Hi Jerry,

Mixed ambient light will lead to mixed ambiance (pun intended, but with a serious undertone). You'll get the best results by measuring the lightsources themselves, and let the ambient relfections take care of themselves. You'll never get a reliable White Balance reading close to colored surfaces.

For our grey card, simply placing it in the area is not very helpful of course. There rarely is a "right" place to put the grey card that avoids all the glare, is in the right zone (i.e. the actual brightness I'm exposing for in the camera).

To state the obvious for some, a grey card is intended for exposure determination. Not all grey cards are as neutral as needed (especially when it's been used for a while) for determining the Color/White Balance. It needs to be angled correctly between the camera/subject and subject/lightsource directions, because it has a semi-specular reflecting surface.

We've not been real happy with the Expodisk (realizing it should be at the location the camera is aimed at, not where the camera is shooting from), but knowing where to place a grey card has been even worse.

Expodisk was originally also designed as an exposure tool. It should be used at the subject position pointing to the camera position (thus integrating ambient and incident light falling on the subject). It's not as neutral as really specific WB tools like WhiBal, or the BabelColor White target.

The latter two have a respectively very decent and good diffusely reflecting surface which allows to integrate the ambient and incident light falling on the subject. They are also rather spectrally neutral. The WhiBal is quite neutral and the BabelColor target is even close to perfect across the visible spectrum.

Bart
 

Jerry Levin

New member
Hi Bart, Michael, Doug...

Appreciating your replies, but also trying to boil this down to a best practice here.

I'm in the same designer boat's interior space I described earlier. I have direct or indirect sunlight streaming in through tinted glass windows. I have tinted light fixtures on the walls. I have teak floors, teak walls, and deep shadows, etc. Florescent fixtures in the kitchen. A veritable stew of light colors and sources. The color of the teak, let alone everything else is critical for the client's marketing images I'm responsible for producing.

I have 4 different boats to shoot in one day, each with 6 different camera positions throughout a given vessel. A very long day indeed, just to find the right sweet spots, and proper exposures. That part we can do, and reasonably efficiently.

I want to get do my best to ensure accurate color fidelity for each shot.

What's my best practice (and best tools) for doing so, please, and still keeping this to a single day's shoot?

Many thanks for your input!
 

Michael Tapes

OPF Administrator/Moderator
The ability to properly (meaning best for you and/or the client) WB the picture at capture time is really quite easy IMO. Assuming you have everything else down (which you seem to have).....

You are ready to take your "final shot".
--The camera is on a tripod and set for Raw capture.
--You use a remote control (assistant or Pocket Wizard for example)
--Walk around the interior with a Studio or reference size WhiBal
--Go to each point of critical color, and take a shot of the WhiBal without changing the camera position or settings. The "angle" of the WhiBal should be appx parallel to the point of interest facing back to the camera. This will become a pretty quick operation, once you have experience with it.
--The real exposure(s) are made bracketing the exposure over a very wide range as required.

That gives you all the data you need to create your vision in post. It does not add much to the time involved during capture, and it is clear by each WhiBal shot the area of interest that was recorded (not possible with an ED type device).

I am perhaps over simplifying the process, and it you take multiple exposures with some lights on and off, it might require some additional WhiBal shots. But this technique refined over time can work well for you. It is not the only way, but it is the way I would deal with the assignment until and if experience drove me to a refined or different method.

Hope this helps..
 
I want to get do my best to ensure accurate color fidelity for each shot.

What's my best practice (and best tools) for doing so, please, and still keeping this to a single day's shoot?


Hi Jerry,

When it's not practical to filter/correct the lightsources to a common color temperature, then the approach as outlined by Michael is what I'd do. Depending on whether it's offered by the camera, you could save some storage by using an sRaw format for the WB shots, then switch to full Raw for the actual exposure(s).

The benefits of Raw are assumed to be understood for a scenario you're facing. They also allow to do multiple conversions of a single Raw, with adjusted WB settings for each target area. Then mask and blend the various renderings into a final single all-over White balanced image. Don't take out all color temperature differences, it might spoil the atmosphere.

The approach with a WhiBal type of neutral object provides a great future reference should you want to redo the image at a later time. You could also build and instead keep a record of the multiple color temperature readings with a given Raw converter, but that might not work as well with future Raw converters. I'd suggest keeping all shots.

Bart
 

Michael Tapes

OPF Administrator/Moderator
Bart,

We agree on all points except sraw. At least until I confirm some things. But the Raw file is a one channel bayered file, and the sraw is an RGB de-bayered file. Depending on the converter used this may or may not affect the WB transfer from one file to the other. I would make some controlled test before I wondered into this area. Also, IMO, it is too dangerous in that the camera might be accidentally left on sraw for the real shot.
 

Doug Kerr

Well-known member
Hi, Michael,

. . .and the sraw is an RGB de-bayered file.
I had no idea. Can you tell me a little more about what this means. Do you mean it is demosaiced (just as is done on the way to a JPEG file), but is perhaps left uncompressed?

Thanks.

Best regards,

Doug
 

Michael Tapes

OPF Administrator/Moderator
Doug,

Yes, a full de-Bayer is done on the file, and then it is down sampled. So it is shrunk in terms of file size by virtue of the smaller pixel dimensions, but it is enlarged because it is in fact an RGB image,rather than a raw which is a 1 channel image.

You can think of it roughly in the same sense as the Linear file stored in a DNG. It can still be manipulated in the "raw" converter, but the de-Bayer is "locked-in".

Back to sRaw. What the net affect is is that we are stuck (for the good or bad of it) with the Canon debayer that is in the camera. You can switch to the latest and greatest RC with that new and super sharp debayer algorithm, but it would have no effect on an sRAW because it is already de-bayered. And as you know what goes on inside the de-Bayer is....
--blurring
--sharpening (of sorts)
all in the guise of inventing the "correct" colors and edges, which are only partially defined in the raw file. So yes, in effect you get the image part of the way toward the JPEG.

Regarding compression, I would guess that the sraw, like its big brother the raw, is compressed, but losslessly, so that is a good thing.

Hope that clears it up for you..
 

Doug Kerr

Well-known member
Hi, Michael,

So yes, in effect you get the image part of the way toward the JPEG.

Hope that clears it up for you.
Yes, very nicely. I had no idea - I thought is was some sort of bona fide raw output!

It's sort of depressing. The "name" is really misleading. It is in no way raw-like. It is apparently reversibly ("losslessly") compressed, but that is hardly the hallmark of raw-ness.

Thanks again.

Best regards,

Doug
 

Michael Tapes

OPF Administrator/Moderator
On the positive side....

Keep in mid that if the sraw *was* in fact a 1 channel Bayered file, it would of course be smaller and have less data. So even if we have a better de-Bayer algorithm to throw at it, it ,might yield a less accurate rendering, than the "poorer" debayer in the camera. When shooting in sraw there is only 1 shot at the debayer of the full raw sized image, and thankfully Canon did the debayer on the full raw before the down sample. SO I am not saying that Canon did anything *wrong*. I just want people to know what an sraw is so that they can use it (or not) from an informed place. Kind of Raw without FUD, if you know what I mean. <g>
 
Top