White Balance Tools Test
I decided to actually run a test to see what kind of results we got from using various kinds of translucent "white" plastic (Pringles lids, Big Gulp lids, Tupperware, yadda, yadda) instead of investing in an Expodisc or some other (pricier) tool. My reason for doing so was really curiosity, since I use an Expodisc, and a Photovision Digital Calibration Target, depending on the application. I just wanted to convince myself that I hadn't wasted all that money. My suspicion was that I had not, that there was just too much variation in these uncalibrated cheap "solutions" to be reliable.
Test methodology: Nikon D200. Bright, sunny day. Test conducted under solid overhang on the north side of the house. Very consistant light. All shot in NEF (raw). For each of the five "tools" tested, I first created a custom WB then took an exposure of the tool on a (reasonably) white piece of cardboard. I then imported the files into Lightroom and compared the WB temperature. The glasses are there just to give me something to focus on.
As I expected, there was a considerable difference between the WB settings obtained with the several materials. The Expodisc gave 8100k, and the others ranged from 6150k to 7050k. Now, I do not purport to say my test proves that the Expodisc is correct, though my real life experience tells me it is. I will say however, that my test proves without any doubt, that there's an unacceptable variance in range of color temperature when you use these other materials that were never intended to provide a neutral for photographic purposes. Particularly noteworthy to me was the 150k difference between two lids (Tupperware) from the same product line. YMMV but I'm sticking with calibrated tools designed for the task at hand. Hmmm . . . wonder if my Expodisc would keep me from spilling my Big Gulp???
I've exported the files as small jpegs to a Smugmug gallery, and have imbedded them here even smaller:
Expodisc - 8100k
Nalgene Lid - 6150k
Kitchen cutting sheet - 6900k
Tupperware #1 - 7050k
Tupperware #2 - 6900k
Test methodology: Nikon D200. Bright, sunny day. Test conducted under solid overhang on the north side of the house. Very consistant light. All shot in NEF (raw). For each of the five "tools" tested, I first created a custom WB then took an exposure of the tool on a (reasonably) white piece of cardboard. I then imported the files into Lightroom and compared the WB temperature. The glasses are there just to give me something to focus on.
As I expected, there was a considerable difference between the WB settings obtained with the several materials. The Expodisc gave 8100k, and the others ranged from 6150k to 7050k. Now, I do not purport to say my test proves that the Expodisc is correct, though my real life experience tells me it is. I will say however, that my test proves without any doubt, that there's an unacceptable variance in range of color temperature when you use these other materials that were never intended to provide a neutral for photographic purposes. Particularly noteworthy to me was the 150k difference between two lids (Tupperware) from the same product line. YMMV but I'm sticking with calibrated tools designed for the task at hand. Hmmm . . . wonder if my Expodisc would keep me from spilling my Big Gulp???
I've exported the files as small jpegs to a Smugmug gallery, and have imbedded them here even smaller:
Expodisc - 8100k
Nalgene Lid - 6150k
Kitchen cutting sheet - 6900k
Tupperware #1 - 7050k
Tupperware #2 - 6900k
John :
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
0
Comments
This is a great idea, and I want to thank you for this effort. Very interesting and informative. Who wouldn't like to save a few bucks. But.......
When I look at each of these images, the backgrounds are not white, but a medium gray. When I read the pixels, I cannot find anyone of them that is brighter than maybe 180,180,180. To my mind, I would expect a white background to read near 245,245, 245 + or - 5 or 10 points. Wouldn't you too??
Either these images were under exposed, or not processed completely in RAW would be my guess. And....when shots are significantly under exposed, it is rec'd to correct the Exposure slider BEFORE trying to adjust the color temperature. At least that is my understanding of the use of the RAW converter. Best would be to reshoot. Can you post the exif data for these shots so we can compare the exposure to that rec'd by Sunny 16? It is my understanding that these images were shot in full shade, thus the 8500 K temperature with the Expodisk?
Please do not take these comments as criticism. The reason I am posting is because I think this topic is worthy of the investigation you have done, and want to see that it is completed and as accurate as possible.
Moderator of the Technique Forum and Finishing School on Dgrin
Since everything was done the same, and the light didn't change from shot to shot, I think the results are still valid (even if the photos themselves are crap. If you (oh wise one) think I should reshoot to have valid results, or at least more credibility, then I'll do it. But! Instead of shooting the stupid plastic, I'll shoot my Digital Calibration Target after doing the Expodisc WB shot. Then I'll have a perfect exposure of a perfectly neutral subject. I just thought it would be "cute" to have the example photos be of the material through which the WB setting was obtained.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
As I said, my goal was not to criticize your work or your intentions. I think it is an interesting topic. I have a gray reflector from SmartShooters for setting WB. If I get time I'll compare it to an Expo disc also - may take a few days or weeks though.
I just know that exposure seems to play a significant role in the accuracy of white balance settings ( at least I think I do:D )
Moderator of the Technique Forum and Finishing School on Dgrin
http://www.youtube.com/user/NYCFilmmakersGroup
http://www.meetup.com/NYC-Filmmakers-and-Actors-Meetup-Group/
Oh, won't that start a war if you come up with different WB settings!
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
Absolutely correct. The first thing the Expodisc instructions say, is to make sure your WB calibration shots are properly exposed. Mine were. What were NOT properly exposed were the shots of the "tools". I think that's where the confunction arises. Anyway, bowing to the weight of public opinion (well two of you anyway) I'll reshoot, but use the Photovision Digital Calibration Target as a subject for the "example."
"Honest, Honey, this is important. People are counting on me," he said hopefully.
"Yeah, right, you time-wasting buffoon" she retorted.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
I'm very content with my ExpoDisk and have posted about it's proper use many, many times.
How did you set your WB with the ExpoDisk?
Did you go to the subject's position, then point the camera/ExpoDisk toward the shooting position?
Or, did you simply point the camera/Expodisk from your shooting position toward your subject, therefore getting a reflective shot?
Both work, but I find that being at the subject position pointing the camera/ExpoDisk at the light source, or if that's not practical, pointing it toward the shooting position works best for me. This method allows the ExpoDisk to pickup all the various hues/tones that are falling on the subject itself.
Just curious, because that info will let me wrap my head around your results.
Appreciate your time and effort
The only time Expodisc suggests shooting toward the subject is if you're shooting a landscape under varying light conditions.
Thanks for making me clarify. Sometimes brevity is not our friend.
Edit: Actually, I see now that I only mentioned Expodisc instructions in Post #3. Sorry - John
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
Summer noon sunlight is typically rated 5500K.
Indirect sunlight (open shade) usually is a higher temperature, typically around 6500-7000K up to a mid-summer max of around 8000K.
The sky itself may be from 12,000-20,000K
http://www.lightsearch.com/resources/lightguides/colormetrics.html
http://www.sizes.com/units/color_temperature.htm
I think 8100K is going to cause undesirable color shifts for human subjects, unless I misunderstand the conditions of the test.
Moderator of the Cameras and Accessories forums
Moderator of the Technique Forum and Finishing School on Dgrin
To be honest, I've had such good results using the ExpoDisk that I've not paid attention to the Kelvin. Why would I? If it's right, its right...
But, by its virtue, the ExpoDisk is taking EVERYTHING thats hitting the subject (if of course the ExpoDisk shot is taken from the subjects position) into account via averaging the light and all the color cast/tones that are present there also. The ExpoDisk is an extensively tested/calibrated product for correcting WB. Are the Pringles lids?
(BTW, this is what sets the ExpoDisk apart from the other methods/products -- Averaging the WB shot as opposed to shooting a white card, then correcting in post --- With an ExpoDisk, you set it and shoot. No need to set a correct WB later in post. That's one less step in my process folks)
I've stated this before: Ever wonder why the product is named "Expo"Disk?
That's right, you can set your "exposure" with it also.
Can a coffee filter do this? How about a Styrofoam cup?
Then, the camera can use this averaged scene to set a proper WB. I've not had reason to do the actual testing about this, but I would certainly believe that dependent on the many factors involved, light source Kelvin, color of decking floor reflected onto subject, etc., the Kelvin will show up at quite different levels.
Remember: The ACR values for shade, sunlight, etc. do not, and cannot account for any other color cast/shades imposed on a given subject/scene.
So, I don't think it's odd that the Kelvin for a given scene will differ (standard ACR Kelvin setting -vs- CWB from ExpoDisk) from a lab canned Kelvin setting per "X" light source with no interaction from other reflected colors/tones onto the subject, to what the Kelvin will be after an ExpoDisk WB shot was taken and set as the CWB in camera.
Don't get me wrong...
For those that want to use Pringles lids and coffee filters and styrofoam cups, more power to ya'... but please, let's just understand that were talking apples and oranges here folks.
Oh, the resultant color temps are lower than yesterdays, I think primarily because I oriented the target vertically (camera held horizontal) insted of lying on the floor like yesterday. So your 8100 kelvin RED HERRING is a non-factor. So there. :deadhorse
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
New test. This time using a Photovision Digital Calibration Target for the images. I did not use the target to set the WB. The histograms of the images indicate they're all properly exposed.
Test parameters and methodology: Nikon D200, ISO 400, Aperture priority f6.3 shutter varies between 1/125-1/160. All "tools" used per Expodisc instructions. That is: When taking the custom WB setting shot, the camera was at the subject opsition, and pointed at the position the camera would be in for taking the image exposure. (OK Randy?) The target was hung vertically from a light stand in full shade, on the north side of the house. There's a large McMansion about 100 feet behind my house, with grey siding, and the morning sun was reflecting off that house., which I'm sure warmed up the light.
Custom WB was set using each "tool", then a shot was immediately taken of the target, allowing the camera to properly expose the subject. All seven sets were shot within 15 minutes, and I'm pretty confident the light stayed acceptably constant. There was a high, thin overcast.
Here's a link to the Smugmug gallery of all the shots with exif stuff for those of you so inclined. I put more comments in the gallery than on the images here.
#1 - Expodisc - 6450K - Best result when compared visually to the Target
#2 White plastic cup - 4950K - Much bluer than target
#3 - Genyoowine Tupperware lid - 5650K - Still a bit blue
#4 - Nalgene bottle lid - 5650K - Same as the Tupperware sample
#5 - Nikon SW-12 diffuser sheet - 6000K - Getting pretty close but still a bit blue
#6 - Random plastic packaging foam - 6100K - closest of the lot save the Expodisc.
#7 - (My personal FAV!) Plastic end of a Smugmug shipping tube - 3150K Damn thing sure looked white. Guess not!
What's it all prove??? Maybe nothing to you, but it sure proves to me that a calibrated and purpose-made WB tool, whether it be a shoot-thru like the Expodisc or similar product, or a known neutral target, like the Photovision or any number of others, is well worth the expense.
Consistancy is important to me in the kind of shooting I do most (interior architecture) and I can't take the chance on the WB that would result from picking up some piece of styrofoam out of some fast food joint's garbage. I didn't test a coffee filter, 'cause all mine are unbleached (brown) but somehow I doubt they'd fare much better. Sure, after much trial and error you might find a particular piece of free plastic that gives you acceptible results, but I know my Expodisc works, and the other things I tested (today and yesterday) didn't.
Oh, and before you say, "Well your #1 (Expodisc) image looks (magenta, brown, cream, pink, whatever) on my monitor," know this: The Photovision Digital Calibration Target white panel is not brilliant white. The flip side of the target is a white white. The target side does indeed have a "creamy" look to it. The Expodisc WB without question resulted in the image that most closely matched the actual color of the target.
My fingers are tired.
John
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
Well Ziggy, I looked at your links. Not arguing, but the term they use is "average" summer shade. When you get high up on the kelvin scale, a 100k diff ain't much. Not like the difference between 3000 and 3100. When I move my Lightroom temperature slider (on my #1 image from yesterday) between 8000 and 8100, I can't see one iota of change. "I mean no disrespect, Godfather."
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
When I read the pixels in the Expodisk sample, the white reads 241, 233, 233 - slightly warm white like you described it, but definitely more white than in all the other samples shown. The gray reads 140, 140, 140 which is a perfect mid gray. The variation from 128,128,128 may be due to my monitor being brighter or darker, but it is definitely a mid gray. The black patch reads 42,42,42 with some slight variation. A dark gray, that I will accept as black.
I agree with Randy that the ExpoDisk is a clear winner. I will use mine more in the future than I have in the past, based on this series of tests. Thank you for your effort, Icebear. I owe you a Molsen's sometime.
Moderator of the Technique Forum and Finishing School on Dgrin
What I was trying to convey, admittedly not very well, is that measuring white balance is tricky and may require more than one measure to determine the particular setting appropriate to the particular shot.
The "Expodisk" seems to be using a prismatic filter and may be sampling a "very" wide field. It does not appear to work in quite the same fashion as most other white balance tools, including laboratory grade tools, so I'm not sure it's fair to compare with simple diffusion devices because you may be sampling different things. I am not claiming that any particular system is absolutely accurate or preferable. One system might work better in particular circumstances than another.
Moderator of the Cameras and Accessories forums
And you said a mouthful! The Expodisc takes the average color temperature of the light hitting the subject (when used correctly) but that's not always what you want. Often, in a situation where you have different kinds of light illuminating different areas of a subject, you must decide whether you want to average them , and maybe have everything be a little off (but in different directions) or pick the most important part of the subject and correct your WB for that area. I find that shooting a calibration target works better in that kind of situation than using the Expodisc. That's why I have, and use, both.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
First, let me give kudos to IceBear for a job well done!
I'm fascinated by this thread, because last summer, I did much the same thing. I told our local camera club (COBA) that I would do an investigation of do-it-yourself white balance filters, and summarize it in a presentation.
I did hundreds of tests on well over ten different kinds of material and found that they tended to fail for one of four reasons:
1. They looked white, but actually were colored.
2. They were white, but had particles (for diffusion) that also introduced a reddening of the light as it went through due to scattering -- just like a sunset.
3. They didn't obscure detail enough to provide a uniform white-balance image.
4. They generated different color shifts for off-axis light sources -- typically due to having a textured surface that introduces more scattering for off-axis sources.
In the end, I concluded there is value in an engineered white-balance filter, and engineered one (the Phoxle SpectraSnap). It's not that you can't come up with DIY solutions that are close, but you'll never know for sure until you've measured it with a spectrophotometer, and have done a lot of on-axis, off-axis, specular and diffuse source testing. Even then, you're limited to the materials available instead of being able to tune the response with custom filters until everything is just right. That said, it's a lot of fun to try :-)
Chris
Phoxle Founder
Thanks, I think. But SPEAK ENGLISH DAMMIT! I realize there are lots of non-native English speakers who participate in this forum, whose English is as good as mine, but WTF language are you speaking????
What I found as telling as anything about my testing was that two different Tupperware lids tested 150 degrees kelvin different. Admittedly, that was around 7000K where 150 isn't really that noticeable, but still . . .
Spend the money. Buy the consistancy.
Plus you don't look so stoopid to your client as you do putting a coffee cup over your lens. Credibility matters when your client is watching. Lots of photographers buy big honking battery packs for their D200s (for instance) so they look more like D2s just so their clients will be impressed by the size of their cameras. This is not a trivial thing. The last thing in the world I want is for a client or her associate to say, "Oh, my uncle Bert has that same camera." Don't look like a putz. But that's for another thread.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
I couldn't agree more about spending the money to buy consistency.
Thank you for your time and effort.
Icebear - let me add my Thank You to the mix for a job well done (or it appears so with my limited background in this subject area).
My Photos
Thoughts on photographing a wedding, How to post a picture, AF Microadjustments?, Light Scoop
Equipment List - Check my profile
Two things:
1. Where's your comparison. And while you're at it, test against the results for a Photovision target too.
2. Do I have to come to Indiana to collect that Molson, or will your bike make it to West Virginia (compromise)?:D
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
Yikes!! Its only been two days!! It is tax time if you haven't noticed. That's my priority right now - I don't want to go back to the slammer!!:D
I said give me a couple weeks - It was overcast and raining all day here also - I would like to have a nice dry sunny afternoon to do this. I will include a MacBeth color checker as well when I get to it.
The last time I rode through West Virginia was September 2004. Are you going to Moab, perhaps?
Moderator of the Technique Forum and Finishing School on Dgrin
:bigbs
No, no, no . . Tax time started 1/1/08. One word Compadre - TURBOTAX!!! I should talk. I've done my kids' returns, my inlaws' return, and mine, but still haven't filed mine, and I have a credit, so who's a dumass?
Hey, anybody can do a test on a sunny day. But the Gretag-MacBeth color checker is a cool idea. You get extra credit for that.
Seriously. I have a nice cabin in the mountains in WVa. Awesome paved twisties. The local breweries suck, but the local grocery has a decent micro-brewery selection, and there's plenty of likker in the cabinets anyway. Forget Molson. Bring your GS and we'll act stoopid (after the side stand's down).
You know what disappoints me a bit about this thread . ? . ? All the responders already know all this stuff. Hopefully all the other viewers got something out of it. All I got was crap from SWMBO.
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.
That's because some of the rest of us (moi, anyway), don't already know this stuff, and feel a little intimidated and inadequate to comment and admit our ignorance. I did get something out it...an interest to learn more (which I also got from another related thread about sticky-gel filters...fascinating stuff, if you ask me).
Still a bit confused....still learning.
Any one try the WhiBal?
Works great.
WWW.WhiBal.com
Also there is great educational stuff on topics from Lightroom to Raw shooting
Mention my name "Roger Galburt" and you might get free shipping or something.
This guy is v e r y knowledgeable about technical stuff having to do with anything Digital
Have fun
Natural selection is responsible for every living thing that exists.
D3s, D500, D5300, and way more glass than the wife knows about.