An excuse to discussing "Great Color" and getting to it!

[Asher Kelman]

EXAMPLES OF IMAGES needing correction with a profile for a particular LENS-CAMERA-LIGHT-CONDITION

to follow!
​

Asher

 

[Doug Kerr]

Hi, Asher,

I will shortly show how these profiles can be used in two checks of the mouse to correct one's pictures.

I'm looking forward to it.

Let's whip back to reality for a moment. The minimum you need is a matched set of camera and optics from the same MFR and a white balance card or diffuser and don't ever change color using the sliders for correcting color. These should be avoided like the plague until you have confidence as to simple, reasonably accurate color reproduction. The cameras of the last 10 years are pretty well perfect for almost everything, except you need to include that WB card in all your sets of pictures for a new set of lighting!

Are you suggesting that if you indeed have manufacturer-matched body and lens, then "all you need to do" to get proper color reproduction is take advantage of a neutral target in the scene?

Do you mean "you don't even need to have a custom profile"?

And why is having a body and lens from the same manufacturer a requisite for this "convenient" scenario? What "magic" does that evoke?

Thanks.

Best regards,

Doug

 

[Asher Kelman]

Hi, Asher,



I'm looking forward to it.



Are you suggesting that if you indeed have manufacturer-matched body and lens, then "all you need to do" to get proper color reproduction is take advantage of a neutral target in the scene?

Do you mean "you don't even need to have a custom profile"?

And why is having a body and lens from the same manufacturer a requisite for this "convenient" scenario? What "magic" does that evoke?

Doug,

It's close enough as the MFR's have matched lens-sensor-algorithms! In most cases the color is fine. For copying pictures for an art gallery, you might be OK but I would profile the camera, even with its own lenses. One profile for each lens in the various lighting you use for an important shoot. Wedding photography doesn't matter when it's going to a major color house, as they have immense experience and will give you back perfect pictures for the bride just from your out of camera jpg or raw files. For a major portrait, even then, the color out of the camera may be just what you like! as long as you are using a single light source and no reflected hues screwing things up.

Asher

 

[Asher Kelman]

Risky times to use the WB tool!

CAVEAT: Atmospheric, sunset, Dawn or Jazz/club/fashion runway lighting!

Some color induced moods require that we preserve the color hue
naturally added to the scene. for example dawn and sunset or the
green coming from a tree. So know what you are correcting. For a
sunset, with an odd camera lens combo, try taking a white balance
shot before sunset and then you will not remove the romance!​

So ask yourself why one is staying out late to do that wonderful scene. If it's the light, you need to resist the desire to normalize everything! However, you still can try out a lens profile made in normal daylight for that sunset when you use a foreign lens with a color cast. For a night club, try using an incandescent profile for your foreign lens so that the stage lighting won't be overcorrected. You can selectively blend parts of the scene to preserve wild stage colors!

Asher

 

[Doug Kerr]

Hi, Asher,

CAVEAT: Atmospheric, sunset, Dawn or Jazz/club/fashion runway lighting!

Some color induced moods require that we preserve the color hue
naturally added to the scene. for example dawn and sunset or the
green coming from a tree. So know what you are correcting. For a
sunset, with an odd camera lens combo, try taking a white balance
shot before sunset and then you will not remove the romance!​

I remember many, many years ago (when we were both little kids) when you pointed this out to me, a very valuable piece of advice.

Best regards,

Doug

 

[Doug Kerr]

Hi, Asher,

If we have constructed a custom colorimetric profile for a body-lens combination, under some specific illumination, using a Color Checker target and the associated software, then if we have that profile in effect when we process a shot taken under precisely that same illumination, can we get "proper" white balance color correction without having to "eyedropper" a known neutral object in the scene?

If so, how do we set the white balance controls of our raw development software to make this happen?

Thanks.

Best regards,

Doug

 

[Asher Kelman]

That is to follow forthwith!

Asher

 

[Doug Kerr]

Hi, Asher,

That is to follow forthwith!

Excellent! I'll look forward to it.

Best regards,

Doug

 

[Asher Kelman]

So one drags the RAW files to Photoshop or Lightroom and selects from the Camera Icon drop down menu on the right, "Standard"sort one of the profiles one has created for these incoming files.

Standard option Selected from Drop Down Menu at Camera Icon in ACR​

Theres' a cast on the image.

So let's see what happens when we use the White Balance tool, (WB) by clicking on the included great WhiBal™ card from Michael Tapes.

Standard option Selected from Drop Down Menu at Camera Icon in ACR
and then corrected with a the WB tool on a grey card in the same light​

Well, that's so much better, but not quite there yet!

Let's start again! On the right above the options for color correction exposure and so forth, there's a row of little icons. Pick the one with the camera. In the drop down menu you will see various choices such as Landscape, Standard or Portrait and the like. Below that will be a dropdown menu of the custom profiles. Click on the one pertinent to the lens-body-lighting for the RAW files, (in this case the Sony A7R-Leica Summicron 50mm), being processed.

Sony A-7R-Leica Summmicron 50mm 1.4 option Selected from
Drop Down Menu at Camera Icon in ACR and then image corrected
with reference image of Grey cord by Synchronizing the correction.
​

Voila, the image is corrected just for the nature of the lens-camera combo in the chosen illumination of the calibration. To complete color correction, click on the gray card with the WB dropper. The image has been corrected for the lens color cast, the lighting conditions and then refined using the WB tool so that that actual light in that area of the picture, at that particular time of day, is now refined to properly account for the variance in make up of the very closely similar illumination under test conditions and in the shoot.

Now select all the other images and use the "Synchronize' option and all the files will be corrected in one go. I do some 20-50 at a time!

Asher

 

[Doug Kerr]

Hi, Asher,

Now, I recently asked this question (emphasis added just now):

If we have constructed a custom colorimetric profile for a body-lens combination, under some specific illumination, using a Color Checker target and the associated software, then if we have that profile in effect when we process a shot taken under precisely that same illumination, can we get "proper" white balance color correction without having to "eyedropper" a known neutral object in the scene?​

So, it appears that the answer is "no".

Thanks.

Best regards,

Doug

 

[Asher Kelman]

H

If we have constructed a custom colorimetric profile for a body-lens combination, under some specific illumination, using a Color Checker target and the associated software, then if we have that profile in effect when we process a shot taken under precisely that same illumination, can we get "proper" white balance color correction without having to "eyedropper" a known neutral object in the scene?​

So, it appears that the answer is "no".

Actually, if it were possible to maintain the color temperature of the light from the time the lens profile shot was taken and the new pictures taken needing to reference the resulting profile, then, of course, we'd not need in addition a WB step as well.

But it's impossible to prevent the make up of what seems like constant light from shifting in composition!

Asher

 

[Doug Kerr]

Hi, Asher,

Actually, if it were possible to maintain the color temperature of the light from the time the lens profile shot was taken and the new pictures taken needing to reference the resulting profile, then, of course, we'd not need in addition a WB step as well.

But it's impossible to prevent the make up of what seems like constant light from shifting in composition!

Oh, of course.

But suppose in some case we were assured that the illumination was exactly the same as the illumination the profile was taken under.

Then we would presumably not, of course, have to "eyedropper" the image.

But still, how would we set the white balance color correction "sliders" in the raw development software? Or would they, by virtue of the influence of the profile, be "automatically" set properly when we go there?

Thanks.

Best regards,

Doug

 

[Asher Kelman]

Hi, Asher,



Oh, of course.

But suppose in some case we were assured that the illumination was exactly the same as the illumination the profile was taken under.

Then we would presumably not, of course, have to "eyedropper" the image.

But still, how would we set the white balance color correction "sliders" in the raw development software? Or would they, by virtue of the influence of the profile, be "automatically" set properly when we go there?

Yes, Doug,

Under those strict circumstances, where not just the perceived color, but also the distribution of the energetic photons was also identical, then there would be no need to correct further using the WB dropper. However, even at the same time of day, the distribution of wavelengths making up "white" light will vary and no studio flash will likely give identical components to their seemingly "white" light.

Even if we took a nice color meter and measured the "temp" of the white light and it was identical, the distribution of energetic photons, or the respective wavelengths would be slightly different and hence colored objects would look different under the two seemingly identical conditions.

Asher

 

[Doug Kerr]

Yes, Doug,

Under those strict circumstances, where not just the perceived color, but also the distribution of the energetic photons was also identical, then there would be no need to correct further using the WB dropper. However, even at the same time of day, the distribution of wavelengths making up "white" light will vary and no studio flash will likely give identical components to their seemingly "white" light.

Even if we took a nice color meter and measured the "temp" of the white light and it was identical, the distribution of energetic photons, or the respective wavelengths would be slightly different and hence colored objects would look different under the two seemingly identical conditions.

Yes, I understand all that. But my question was (and I will augment it here in blue to avoid any weakness in its premise, and will emphasize the yet-unanswered part):

But suppose in some case we were assured that the illumination was exactly the same as the illumination the profile was taken under (that is, had precisely the same spectral distribution).

Then we would presumably not, of course, have to "eyedropper" the image.

But still, how would we set the white balance color correction "sliders" in the raw development software? Or would they, by virtue of the influence of the profile, be "automatically" set properly when we go there?​

Best regards,

Doug

 

[Asher Kelman]

Yes, I understand all that. But my question was (and I will augment it here in blue to avoid any weakness in its premise, and will emphasize the yet-unanswered part):

But suppose in some case we were assured that the illumination was exactly the same as the illumination the profile was taken under (that is, had precisely the same spectral distribution).

Then we would presumably not, of course, have to "eyedropper" the image.

But still, how would we set the white balance color correction "sliders" in the raw development software? Or would they, by virtue of the influence of the profile, be "automatically" set properly when we go there?​

Doug,

Sorry if I was not completely clear!

When there is no need to use the WB tool, also there is no need to touch the color sliders for that same purpose. So, for the sake of WB, there would be no need to touch the WB tool! There would also be no need to touch the color sliders either!

Still, for "taste" or artistic effect, or as they say in film, "look", one could use the sliders to taste and so degrade the WB!

?

Asher

 

[Doug Kerr]

Hi, Asher,

Sorry if I was not completely clear!

When there is no need to use the WB tool, also there is no need to touch the color sliders for that same purpose. So, for the sake of WB, there would be no need to touch the WB tool! There would also be no need to touch the color sliders either!

That's what I wanted to know. Thanks.

Still, for "taste" or artistic effect, or as they say in film, "look", one could use the sliders to taste and so degrade the WB!

Yes, of course. And, as you know, with 3D LUT Creator, one can do "color grading" to taste and still preserve white balance as such!

Thanks.

Best regards,

Doug

 

[Doug Kerr]

Hi, Asher,

I am still a bit bothered by the matter of the "WB sliders" when using a custom camera profile.

I am of course hampered by:

• My lack of familiarity with PS and ACR

• The latest PS I have is "CS5" (12.0.4), and ACR 6.7.1.340

• Not having any custom camera profiles in my "library"

But of course I never let the lack of all the information and resources I "should" have slow me down.

Now you have said that, in the admittedly-contrived situation of a shot taken under precisely the same illumination used to produce our custom camera profile, there would be no need to touch any of the controls in the "white balance" tool panel (nor use its eyedropper).

But here's the problem. In my PS/ACR system, when I load a raw file, all the controls in the White Balance tool panel are initially set where I had them set for the last time the tools was used. They are not, for example, set to "null" ("do nothing"); in fact there is no "null" setting for these controls.

If, in the menu of camera profiles (where I have only the "generic" ones), I choose one, this does not force the white balance settings to some particular value.

Now, if I had a custom camera profile, when I chose it, would it force the settings of the WB tool to some particular values?

If not, then we of course face the prospect of a God-knows-what coloration of our developed image. So the story about not having to touch the WB tool in the situation I described above turns out to not be so.

If in fact choosing a custom camera profile does "preset" the WB tool values, I would be interested, in the case of one of your profiles (taken under illumination of some general type), what value the WB "color temperature" (CCT) slider ended up getting preset to.

Thanks for helping me make this whole notion "reach from cover to cover".

Best regards,

Doug

 

[Asher Kelman]

Great engineers attention to detail! disconcerting if the color balance of the last work session hangs around!

Notice above that either I chose "Standard" or else my Leica M lens profile for my Sony A7R camera. I have now added a title below each image to clarify this step. So I believe that one is supposed to set up the appropriate conditions desired to replicate in the "camera icon pull down menu"! Set it to "Standard" and look at the color sliders!

Hope this works for you in PS 5.

Asher

 

[Doug Kerr]

Great engineers attention to detail.

Notice above that either I chose "Standard" or else my Leica M lens profile for my Sony A7R camera. So I believe that one is supposed to set up the appropriate conditions desired to replicate in the "camera icon pull down menu"! Set it to "Standard" and look at the color sliders!

Hope this works for you in PS 5.

No. In PS CS5, setting one of those "presets" does not move the sliders at all. Does it in your setup (I forget just what application you are working with )?

However, as I select different of these "default" camera profiles, I do see changes in the coloration of the previewed scene.

If I were choosing among actual camera profiles, I assume that this would be accommodating the colorimetric response of the camera, as determined by the profile generation process.

But choosing a camera profile does not seem to preset an assumed chromaticity for the illumination of the shot in hand. As you know, my conjecture was that perhaps it would initially set that to the same illumination the profile was taken under. But evidently not (in PS CD5). And that is not at all unreasonable.

Thus, even in the contrived (or arranged) situation in which the shot is taken under the exact illumination used to generate the camera profile, we must still, for example, use the eyedropper on the model's Spectralon brooch (unless we know the CCT and Planckian offset for that illumination).

Best regards,

Doug

 

[Asher Kelman]

Now lets spend a little more effort on the fidelity of the color.

Remember we are able to process RAW files to correct for the hues imparted by a particular lens-camera combination in a certain light, made previously), and also to refine that with a WB card such as WhiBal™ used in the same shoot.

Custom Zeiss A7R-Leica Summicron 50mm ACR Camera
Profile With WB using Grey Card from That Same Shoot​

The resulting image has color closely faithful to the original subject.

BUT WHAT IF WE DON"T HAVE A CUSTOM PROFILE?​

But what if we haven't such a custom profile. What do we do then? I have emphasized previously that any camera, (made in the past 10 years) are so well color profiled that one generally doesn't have to worry about correcting color for the most past. Still including a WB reference is always a good insurance policy!

Standard ACR Camera Profile Only

Standard ACR Camera Profile and WB


PRETTY BAD!​

So let's take a very different tack. We'll try color correction using selective color in Photoshop:

Standard ACR Camera Profile. and file saved as TIFF then
simply adjusted "by eye" with "Selective Color" in Photoshop​

That's a pretty quick correction done on my laptop but seems pretty usable, considering we didn't have custom profile. But look again at the properly corrected image with our custom profile:

Custom Zeiss A7R-Leica Summicron 50mm ACR Camera
Profile With WB using Grey Card from That Same Shoot​

The take home lesson is that one should stick to MFR's sets of lenses that go with their own cameras as they have taken into account color profiles in the "in camera processing". Still, try to include a Whibal™ grey card and make profiles for foreign lenses, especially if they give a noticeable color cast.

It's far, far easier to use a custom profile than to try to undo the complex changes otherwise!

Asher


(Yes I know there are few more dust bunnies to be cloned away. Later!

 

[Doug Kerr]

Hi, Asher,

The take home lesson is that one should stick to MFR's sets of lenses that go with their own cameras as they have taken into account color profiles in the "in camera processing".

Are we still speaking of working from the raw output of the camera? So you are speaking of in-camera processing of the raw data (so it is only "a little cooked")? I wasn't aware that was done.

Or are we speaking of working with the in-camera JPEG output?

Best regards,

Doug

 

[Asher Kelman]

Hi, Asher,



Are we still speaking of working from the raw output of the camera? So you are speaking of in-camera processing of the raw data (so it is only "a little cooked")? I wasn't aware that was done.

Or are we speaking of working with the in-camera JPEG output?

We can presume that all files prepared in the camera, RAW or JPG have the basic algorithms enabling the conversion of the inflow of analog DATA to image files tagged in some chosen color space and being quite faithful to the original colors.

For JPG files, the camera can go on to add in custom corrections such reference to a white balance card that fills the frame, (or instead a frosted glass/plastic filter in place of a lens cap. In addition, at one's creative whim, one can choose levels of saturation, sharpness, B&W, monochrome in in some cameras a host of added creative filters.

RAW files however are presumed to be used by pros and so one is expected to use one's own WB shot for each light condition used in that shoot. In addition, the RAW processing software can offer choices of "standard" or for example "portrait", "landscape" or other tweaked color and saturation flavored "normal" "camera settings". There is no need for the camera to try to predict what refinements the professional photographer or his/her retoucher will end up doing today, 3 months time or 5 years from now. So beyond whats needed to translate the math and embed the necessary results of a look up table to give "accurate color" for the MFRs native lenses built for that camera model, the RAW files are just, "non-flavored", not even vanilla. For vanilla, choose "standard in the camera drop down menu.

Asher

 

[Doug Kerr]

Hi, Asher,

I'm afraid that there are several paradoxes in the way of my embracing* the notion of colorimetric correction being applied in-camera to the to-be-delivered raw data.

*That can be taken as meaning either "understanding" or "believing".​

My concerns are somewhat complicated, and I don't want to clog the waterway by presenting them here. As you know, I have sent you a separate wire in which I lay out these issues.

I am making a search of the "literature" for information that will help me to better come to grips with this concept. So far, it seems to be a deep, dark secret, never to be spoken of.

On the other side of the coin, I am very comfortable with the idea of colorimetric correction (per a custom "camera profile") being applied during the external development of the raw data into an image (then perhaps "fixed" as a JPEG of TIFF file). I am grateful to you for your explanation of this process, with which I was not previously familiar.

And of course I have no trouble imagining that the comparable thing is done in-camera during the development of the raw data into a JPEG file for output.

There is so much in this field (as for any other) that I don't know. I am always grateful to be "enlightened" on any new area. But once "enlightened", I do not rest until I can make the concept "reach from cover to cover" in my own mind. I suppose that this is a creature (some would say a "curse") of my engineering training.

Thanks again.

Best regards,

Doug

 

[Cem_Usakligil]

...I'm afraid that there are several paradoxes in the way of my embracing* the notion of colorimetric correction being applied in-camera to the to-be-delivered raw data.....

Rightly so Doug. Ttbomk, there is no colorimetric correction possible on the raw data, ie the collected number of photons in each sensel. In that sense, the raw image doesn't even have a colour.

 

[Asher Kelman]

We need to update the thread with modern software designed for photographs!

 

[Asher Kelman]

Here is a good introduction to the color grading feature in both LR Photoshop CC RAW. It’s the same engine

 

[Antonio Correia]

Here by Julieanne Kost one of the many Youtube videos about this matter I have seen.
And another one by Glyn Dewis which I found very instructive while leaning (I still do)
No need to make a tif file. Just inside LR I start saving a Snapshot for the Default photo. Starting from there I keep saving Snapshots to compare and get the what I feeel is the most of my shot

Before

After CG

 

[Jerome Marot]

All what these "color grading" software does is restrict the colors in a picture or video to a subset of selected colors, while trying to keep the image "believable". Keeping the image "believable" may, for example, mean that skin colors stay within the accepted range. For the picture above, which has no human skin, it may mean to keep the plants green and stones within earth tones.

The subset of selected colors is chosen according to known color harmonies but also to the kind of mood the picture or video is trying to convey.

If I may voice an opinion: this system works relatively well in video, because the mood is given by the underlying story. In still pictures it is very easy to overdo the effect and then it gets tiring very quickly.

But then, of course, tastes evolve and are formed by surrounding culture. Movies we watch change the perception of color grading because they use it so much. So, maybe what I find a bit overdone is considered a necessity by modern viewers.

 

[Antonio Correia]

"...this system works relatively well in video, because the mood is given by the underlying story. In still pictures it is very easy to overdo the effect and then it gets tiring very quickly..."
Thank you Jerome for voicing your valuable opinion. Much appreciated !
In the above photograph I was trying to make a common shot in a more valuable one.
Cheers !