X-Rite ColorChecker Passport and PWP5

[keithrj]

Tom,

I have a WhiBal from http://www.rawworkflow.com/ which comes in a number of sizes and formats. The one I have has three cards (white grey and black) and works really well and is very durable. You do have to be careful to use it appropriately as Jonathan mentioned.

[doubtingtom]

Thank-you Keith. I have reviewed WhiBal several times, but had no basis for choosing their solution versus some respected others. Perhaps their Studio size will still work comfortably for my hiking needs and offer more surface area to use in PP (if variations_reflections_etc. are present on the card image. Using 24 and 28mm lenses causes me to be more concerned than necessary about card angles (in front of the lens) and distance from the lens.
So much emphasis on Gray with all of the White/Gray products and now my focus is more on adding White to the process.

Regards,
Tom B

[den]

Just curious... what kind Sigma conversion image do you get when you use the camera set to its auto-white balance and 'evaluative metering' and the manufacture's conversion software that came with the camera?

I have found that often until there is a better understanding of a new camera's characteristics that staying close to the manufacturer's suggested default settings and using the supplied software for conversion is sometimes better than than trying to use one's own judgements even with reference cards and/or third party RAW converters.

[doubtingtom]

Hi Den,

Tried Sigma's SPP 3.5 Auto Raw conversion mode and sky color definitely moved more toward the Ricoh and Nikon. Mid-range boost is apparent in all of Sigma's files, but I have factored that into my posted concerns. I must do some more tests at sunrise and sunset since sun's angle and use of gray card is so different from my normal hiking situations (nearly always near mid-day) simply due to time-to-destination factors. I'm feeling now that sun shining directly on gray card was a mistake and perhaps shooting it more reflecting the distant scene's lighting may have been appropriate.


That said, I hear your message clearly re. deference to the manufacturer's analysis and recommended settings both in camera and conversion software.

I also need to take more specific notice of key scene colors and make notes to help recall at later processing. This is never easy, but in this case, I had little remembrance of the sky details among others.

Regards,
Tom B

[den]

Tom...

Your latest image version is improved, at least to my eye. The magenta color cast of the middle distance is lessen, although a cyan cast tends to develop at the low saturation skyline.

In the case of landscapes, do not under estimate the value of Kiril's suggestion to make color corrections by eye. A reference card may only help in the correction of colors in the area of the card which may not represent the vastness of visual elements present in a landscape.

Landscapes may need to be treated as one would a multi-source lighting image. Near, middle, and far distances considered separately and additionally each considered for sunlight and shadow color casts...

...and as such, do not easily batch post-process.

For your last example image version, you could consider PWP5's Remap transform with its color dialogs numbers in the HSV mode to:
(1) match the hue of the upper sky to the skyline and increase the skyline saturation 10%, reducing the cyan cast at the skyline;
(2) decrease by 20% the saturation of the long building's shadowed roof, reducing middle distance shadow blue casts;
(3) decrease by 10% the sunlit high saturation of the red bluffs, reducing strong reds/orange produced from a low horizon sun light source; and
(4) decrease by 20% the saturation of the near soil/rock that is in shadow, reducing near shadow blue color casts produced by a low horizon sun light source.

[doubtingtom]

Den, my weird sense of humor and reaction ("easy for you to say... ") is superficial and not at all indicative of the level of interest and concern I have to produce images depicting scenes .... as I saw them. You and select others have shown an amazing level of understanding of complex light issues involved as well as the ability to use powerful tools to deal with them. Much of this is beyond me, in spite of a frustrating sense of broader image flaws compared to a 'picture in my mind'. I will study your posted comments repeatedly (as I always do) and do my best to apply them and learn to improve my results. I do respect and appreciate Kiril's help and try to apply his suggestions as my abilities allow.
Thank-you very much for taking time to analyze these image files and teach steps to deal with their various flaws.

Best regards,
Tom B

[den]

Tom...

eh! eh! more to consider... what of gray tone brightness/contrasts?


Just as there are millions of combinations of colors/tones, there can be choices for each one based upon one's preferences, rememberances, and vision as a camera is only a recorder of approximations...

Certainly, reference cards, their software, and even camera profiles can be of help to further approximate "what is seen".

I am just encouraging you: (1) to not blindly accept the results that are produced using them in a non-studio environment; (2) to not ignore your hiking companion; and (3) to not be consumed exclusively as an observer but also a participant in the great landscapes that surround us each day...

As for using the tools of PWP, if I can be of help to you or others.... ask... and there is no single image version solution that is the 'right one'... particularly the one's I suggest.

den

[tomczak]

In principle, if the scene illumination is uniform, the standard 24-patch Macbeth card should do a better job in bringing the lighting and camera-sensor/RAW processor peculiarities to a standard, well defined starting point, than a single-point calibration with a gray card as reference. Match Reference transformation automates the process by first using Colour Curves on the six bottom gray patches to a) remove overall colour cast, even if it varies by brightness; b) correct gamma irregularities; c) correct the exposure by contracting or expanding the dynamic range of the corrected image based on the blackest and whitest patch of the Macbeth. It then (optionally) applies a nonlinear multi-colour adjustments using Colour Correct transformation and the 17 remaining colour patches of the Macbeth, to map the colours of the patches closer to their expected values (the 'dark skin' patch is ignored, I believe, for computational reasons as it's almost identical in hue and saturation to the 'fair skin' patch, just different in brightness).

In practice, Macbeth with Match Reference works fine in many cases, but I'm having trouble with: 1) clipping pixels outside the Macbeth black/white patch range (which I find quite common an occurrence); 2) using it to bring processed RAW files to a standard starting point.

The latter is because Colour Curves step, operating on RGB channels, while standardizing brightness also modifies saturation, which then makes Colour Correct having to 'nullify' saturation changes made by the Colour Curves step. In other words, Colour Correct remapping vectors will have different lengths (and possibly direction) still depending on the initial exposure, even though the Colour Curves step has already brought the exposure to a standard.

I described the issues here: http://www.dl-c.com/board/viewtopic.php?f=4&t=196, but I'm running out of ideas on how to make it work more smoothly, in those cases.

[doubtingtom]

den, your 'suggestions' are quite appropriate and deserve a brief review with each new hiking experience. Fortunately, we came here over 20 years ago with a mindset to never let these surroundings become commonplace. As of today, at least, we have managed to maintain this attitude and continue to enjoy immensely what surrounds us. You have caused me to move into another set of PWP tools and I will enjoy the challenge of learning and applying them to many landscape scenes. There are such a vast number of directions one can explore in PWP and it is very helpful to be stimulated to try a specific few which apply to regular image captures.

tomczak, yours is the first specific encouragement and comment on use of Greytag Macbeth ColorChecker for landscape images and I appreciate your perspectives. It is a good thing to work through the several posts here, which have set more realistic expectations if I acquire and use the X-Rite card and/or Passport software. I still do not have a clear idea of my gains using just the card and Match Reference Transformation versus adding the Passport software from X-Rite. I noted Den's comment about batch processing these images and that would seem to negate any 'profile' for a set of image files taken over time and at varying angles during a given hiking activity.

Many thanks for your guidance and ideas to carry me through a few mores steps forward. This is such an expansive and stimulating process to enjoy planning, 'participating', ... then evaluating and sharing. Neat stuff !!

Best regards,
Tom B