Does the 5D3 has WB problems?

Discussion in 'Canon EOS' started by yakim_peled|1, Sep 24, 2012.

  1. Take a look here and tell me what you think.

    Happy shooting,
    Yakim.
     
  2. Wouldn't read to much into it. CFLs have notoriously variable spectrums. Regardless of what the box (or bulb) says - heck, just consider yourself lucky if you've never found one emitting high UV (they don't warn about that on the label!) for this to be anywhere approaching scientific, one would have to evaluate each bulb continuously during the test...
    ...the author also did not include enough on his methodology. For instance, if he was shooting at ~ 1/80 or faster, the camera could have easily picked up on the low voltage side of the single phase AC curve. @ ISO 400 /f5.6, this could easily be the case. That factor alone would totally skew his results.
    That's not to say that AWB is even remotely always right, but that is a key reason we shoot in RAW in the first place.
     
  3. I've used both the 5D1 and 5D3, and I'd say the 5D3 is probably marginally better. It's easily fooled by special circumstances though. For example, fill the frame with a woman in a blue dress against a black background, AWB will balance for florescent even if the light is tungsten.
    Since I shoot RAW it doesn't matter to me much.
     
  4. Somebody with way too much time on their hands....
    My 5DMkIII seems fine as far as WB. Besides.. .it is so easy to correct in post if needed.
     
  5. Does anyone care? It is so easy to correct that it's barely worth thinking about.
     
  6. The 5D Mark III is as accurate as any camera that I've used in Auto and better than most. Of course, auto white balance
    is an I exact science, like auto exposure. Expect it to get close most of the time, but it's not usually right on the money. Think about it. It's impossible for the algorithm to get the colors right, because the camera doesn't know exactly what colors exist in the frame. Is an object blue, or is it lit by a blue light?
     
  7. If the photo doesn't include any neutral gray, it's hard for any camera to adjust white balance. Its best chance is to "know" what objects like sky or grass "should" look like, and adjust accordingly. But trinkets under odd lighting in a studio? How can any camera get that right?
     
  8. Does anyone care? It is so easy to correct that it's barely worth thinking about.​
    I was thinking along the same line but wondered if any 5D3 owners have encountered that.
    Happy shooting,
    Yakim.
     
  9. My 5D3 shoots a little on the cool side, but I shoot RAW so it doesn't matter.
     
  10. You know, even though WB has never in the slightest way, annoyed me (esp. as I shoot RAW 99% of the time), it obviously annoys a lot of people. A simple solution would be to have an 'assign custom WB' function in the menu (or even programmably accessible w/ a button push). The camera would prompt you to point it at something white (or maybe grey) - or that you perceive as white. and when it is you press the button, it evaluates an 'optimal' WB setting for that white thing, and sets your custom WB to that temp. It would take about 2 seconds to do, and you'd have something reasonably reliable to work w/ until the light changed. Seems a pretty straightforward solution to me (and all software - so implementation could be done w/ a firmware update).
     
  11. Of all the auto features on a camera, I would think WB is possibly the one factor for which the photographer should never need any help. Focus assist, sure. Auto exposure, yes, sometimes, when the light is changing frequently/rapidly. But AWB? Why?
     
  12. "Does anyone care? It is so easy to correct that it's barely worth thinking about."
    That's if you don't have thousands of pictures to edit.
     
  13. Harry, when you're editing a bunch of photos that are substantially alike, just tweak one of them the way you want in DPP, copy the recipe, and then paste the recipe to all the other photos as a group. It's very fast to do it that way.
     
  14. Harry, when you're editing a bunch of photos that are substantially alike, just tweak one of them the way you want in DPP, copy the recipe, and then paste the recipe to all the other photos as a group.​
    Or just select a batch of images, and tweak them all at once.
     
  15. "Or just select a batch of images, and tweak them all at once."
    Excuses, Excuses...
     
  16. Who uses AWB on any manufacturer's camera?
    That is the "Good Luck" setting on cameras.
     
  17. Who uses AWB on any manufacturer's camera?
    That is the "Good Luck" setting on cameras.​
    Plenty of people who shoot Raw and shoot in rapidly changing light - many of us don't have the luxury of resetting the camera WB before each frame...
     
  18. AWB is never going to be right on. I find it to be a bit blue/greyish for daylight and have had similar results, a bit too warm, as those I saw in the posted examples with tungsten (1dsmkIII & 5dII).
    Some will claim that the WB setting can affect the histogram and such on the camera and give false information regarding exposure. I think this is true, but I don't know that the magnitude of this is so great to actually cause real issue. Unless one is shooting JPEGs, I don't think the setting really matters much between AWB and the setting that matches the light you are shooting in.
    Although I have shot in AWB for years, I do find that I use a selected mode from the camera settings most of the time these days. I only use a custom WB when I am in studio and wanting a neutral result. I don't think using (or not) any of these things diminishes one's status as a photographer, that is kind of ridiculous--like the arguments between which camera is better. The most important issue is to find the method that works for you and the way you work. WB is not fixed like it was in film days by the film, it is totally flexible and non destructive to the file (actually, there might be some argument here but it is not related to any settings we are talking about here). Figure out what works for you and use it.
     
  19. AWB is never going to be right on.​
    Yep, true - but it's going to be closer than "Daylight" when, just as the bird I've been trying to photograph for the last hour appears from the undergrowth, a big black cloud completely obscures the Sun and kills the light...
    And it's another reason why shooting Raw matters.
     
  20. Of all the auto features on a camera, I would think WB is possibly the one factor for which the photographer should never need any help. Focus assist, sure. Auto exposure, yes, sometimes, when the light is changing frequently/rapidly. But AWB? Why?​
    Well, simply put, there's a focus indicator to tell you when a manually focused lens is in focus, a meter bar to show you an exposure.... All those features a 'real' photographer should never need help with! (j/k of course), but last I check a live spectrometer wasn't a feature... So a photog only has the highly subjective, and continously adapting eyeball to 'judge' an absolute color temp... In a JPEG, exposure is far easier to correct than WB.
    As I said, I don't have a horse in this race (since I shoot RAWS and regularly manually correct WB to suite the mood of the pic), but having a 'calibrated custom WB' feature would be kind of nice.
     
  21. Keith, I agree and I think with my editing of my entry here, I left out some key points. One is that regardless of what you use for WB, most will end up making some tweak to it in post--and the WB sliders aren't the only way it can be affected. I don't know that I ever leave the WB on my own images as they were shot. It isn't always that we want "perfect" WB--obviously it could defeat the purpose of shooting in the "golden hour". But even in more routine shots, little nuances in color can convey different effects. Perfect WB can be useful in some cases but it can also sterilize a scene. The goal should be to use, for the initial WB, whatever works best for us so that we can then interpret the scene as we intended most effectively.
     
  22. I always shoot AWB. It's one less thing to worry about and I can correct it in post. Exposure (to a degree), focus, dof, and
    sharpness can't be corrected if your settings are wrong, but WB can. The less I have to worry about during the process
    the more I can focus on the photo.
     
  23. Exactly, John - it's just a starting point, and better than others in many respects, especially if you shoot away from easy, comfortable, controlled situations, and where time is an issue.
    What Nathan says, really...
    And it's certainly not something simply to dismiss in the condescending way that John T did...
     
  24. "Does anyone care? It is so easy to correct that it's barely worth thinking about."
    That's if you don't have thousands of pictures to edit.​
    In Lightroom you can edit a thousand photos almost as easily as editing one. Get one looking right and apply the same white balance to the rest. If that's not acceptable, use a gray card or the amazing Expodisc to set the white balance prior to capturing the scene. Or set a Kelvin temperature while reviewing the 5D3's gorgeous LCD screen.
     
  25. The other great thing about auto white balance is that even when it's wrong, it gives you an interpretation of the scene that you probably wouldn't have come up with on your own.
    Sometimes a slightly incorrect white balance (too warm, too cool, not exactly the right tint) is creatively more interesting than a clinically perfect setting. (No one shoots correctly balanced sunsets, for example. You'd lose all of the color!) If you shoot raw you can always overwrite the results of the AWB, but I like to have it as a creative option.
    I got the idea to use AWB most of the time when John Shaw said that's how he sets his cameras. I've learned over the years to trust Mr. Shaw's advice.
     
  26. What if you have thousands of photos to edit not all from the same shoot Dan. Some people work out in the field. Not everyone works in a studio and takes thousands of photos of the same thing. You can't copy paste develop settings on different scenes like that.
     
  27. Steve, what you describe is what people did for years with film. Most only carried daylight film and then maybe a few filters that they "guessed" seemed appropriate for the scene. You could adjust the color on each one, as desired, when the print was made.
     
  28. Correct me if I'm wrong, but can't you just click the AWB setting (with a RAW image) in DPP if that's what you want? So let's say your camera was in fluorescent mode as you shot pics of a daylight scene. No problem! Just hit the AWB, and the image is autocorrected, just as it would have been in the camera. Done. If there are a thousand pics, each shot at a different place, with a different type of lighting, just select them all and click the AWB setting. Done. It might not be a one-click operation for such a large number of photos, but perhaps a few clicks will get you there -- far fewer than the number of clicks and taps undertaken by any of the contributors to this thread.
    This all begs the question of whether a particular camera's AWB can really be messed up, compared to its peers. All the camera does is to apply an algorithm to what comes off the sensor -- probably substantially the same algoritm that every other camera uses and that DPP also uses. When I see someone complaining about color balance as it relates to fluorescent lighting and other spectrally weird light sources, I'm always rather skeptical.
     
  29. John I'm saying that the idea you can simply copy-paste wb in Lightroom across a range of images does not hold up if they are of different scenes. When I shoot in tropical rainforests for several months, every image must be evaluated separately for white balance. There are rarely more than two or three shots from each setup, and hundreds of setups.

    Sarah, in Lightroom at least, "as shot" (which could likely have been the auto setting on the camera) is different than "auto" in the software. If you shoot the scene in fluorescent for example, then click auto in Lightroom, the result is not the same as if you had chosen auto wb on the camera at the time of shooting, and then left it on "as shot" in Lightroom. That's why you can use auto all day in camera, then if you choose "auto" in Lightroom (as opposed to leaving it on "as shot"), it changes. In other words auto in the camera and auto in Lightroom are not the same thing.

    I agree with others, AWB is a no-brainer. I always use it and it always gets me 90% of the way, in widely mixed light. Messing with wb is a great way to miss meaningful opportunities, at least with documentary and photojournalistic work.
     
  30. But Steve, that's in Adobe's Lightroom. Adobe and Canon would probably not share algorithms. If you select AWB in Canon's bundled Digital Photo Professional (DPP), don't you end up with the same thing as AWB in the Canon camera, again assuming the image is in RAW format? If so, then you can apply the AWB setting to zillions of photos with just a few clicks and be 90% of the way there, as you put it.
     
  31. Sarah, but I do think what Steve is saying is that even that would be an extra step if you just set the AWB in camera. I haven't tried DPP's algorithm, but I know that ACR's AWB does not generate something I would expect from the camera setting or something I would ever use, at least that is my experience. The other issue is that even if DPP worked as well as the camera setting does, I don't know many who use it except in rare cases.
    But, except when one shoots certain types of things for a client, you don't need to correct WB on every image anyway, you correct the one the client picks. In fact, if you are shooting for yourself, you only need to correct it for those that are worthy of working on.
     
  32. I use RAW+AWB as well and am fine with it (1D, 40D, 7D). If necessary I change it in DPP if needed but most of the time it works just fine. The question is, does 5D3 users have any particular problems with this procedure?
    Happy shooting,
    Yakim.
     
  33. Sarah, I don't use DPP so I don't know about that, but that's not how Lightroom works. In any case it would be easy to find out. Just shoot a scene on AWB, then the same exact scene on fluorescent or whatever, bring both into DPP, change the non-AWB capture to auto in software, and if the resulting wb is identical to the other capture, then DPP works that way.
     
  34. You're singing to the choir, John. I don't really worry about WB until post, and AWB is simply something I've never found reason to use. If I do ever want the AWB interpretation then, I can get it in post. In fact when I click the eyedropper in DPP to correct the WB for gray points, the application automatically applies the AWB setting as a starting point, until I then select gray points within the image. In DPP I can also choose any WB camera setting I want, including auto. I would have a hard time believing that this "auto" is different from the camera's "auto." As far as I'm aware, there's no difference in any of my cameras, but perhaps the 5DIII is different. (I doubt it.)
     
  35. Sarah, it might be worth the experiment as Steve suggested. The problem with interpolating the results using the specific light settings and AWB in the RAW processor is that the latter is evaluative based on the image whereas all of the others are fixed, specific numbers. I would have to assume that those specific light settings in DPP do match Canon camera settings and thus the results would be the same--I have experimented with this in ACR, setting the camera to different specific light settings than normalizing them all in ACR--they all look the same. (That said, ACR's specific settings do not match the in-camera settings (at least for Canon) in this regard--generally neither color temp nor the green-magenta slider.)
    While I was bracketing a shot in AWB in a desert landscape that had nearly equal cool shadows and warm sunlight, I noticed a massive shift in color balance when I exposed for a darker version of the scene--noticeable on the back of the camera let alone on the computer. The image went very blue. When I saw this, I went back and forth between more open and darker exposures to see if it was an anomaly or not--it wasn't. So, I don't know how well DPP would do with a file in auto compared to the camera setting when shot. ACR is generally pretty far off in my experience.
     
  36. Actually, LR is particularly helpful in one respect, hitting the 'W' brings up the 'white balance selector', just click on what should be white, and it sets your WB to an appropriate value.
    If you are shooting one scene/color light, you can batch apply that value, but if you need to do it to individual pictures (such as in a dynamic lighting environment - aka everything BUT the studio), it takes about a half second per picture, W then click.
    But then, unlike DPP, LR is designed from the ground up with us in mind.
     

Share This Page