X rite colorchecker passport magenta tint

Started 4 months ago | Discussions
Iliah Borg Forum Pro • Posts: 28,807
Re: X rite colorchecker passport magenta tint

ggbutcher wrote:

The guy at MaxMax just piped the monochromator output into a fiber optic cable and pointed it directly at the sensor on the lens-less camera... ??

Not exactly 'just'. Dan wrote: "We split the monochomatic light output between two fiber optics cables. One was routed to our spectrometer and the other was routed to the camera."

One of the questions here is the uniformity of the sampling area. However, for analyzing converted JPEGs one doesn't need much accuracy.

https://maxmax.com/faq/camera-tech/spectral-response

-- hide signature --
ggbutcher
ggbutcher Senior Member • Posts: 1,194
Re: X rite colorchecker passport magenta tint

Iliah Borg wrote:

ggbutcher wrote:

The guy at MaxMax just piped the monochromator output into a fiber optic cable and pointed it directly at the sensor on the lens-less camera... ??

Not exactly 'just'. Dan wrote: "We split the monochomatic light output between two fiber optics cables. One was routed to our spectrometer and the other was routed to the camera."

Fair 'nuf. But, the salient point is, didn't use an integrating sphere.

One of the questions here is the uniformity of the sampling area. However, for analyzing converted JPEGs one doesn't need much accuracy.

Yes, was thinking along those lines as I posted the response. Could insert a diffuser and essentially take a picture of the blob, but then there's the diffuser and lens in the path to factor into the measurement. Splaying the light directly on the sensor, with only the fiber to account for, seems to be the least complicated way to accurately present light for measurement.

https://maxmax.com/faq/camera-tech/spectral-response

If I were wanting to go into the business of measuring cameras' spectral response, a monochromator would be my first choice. But I only have three cameras here in my basement, and not much motivation to chase the measurement of many others. For this use case, the homebuilt $130US single-shot spectrometer approach renders data of sufficient quality, within 2 dE (24-patch ColorChecker) of a monochromator-based profile for the same camera.

 ggbutcher's gear list:ggbutcher's gear list
Nikon D50 Nikon D7000 Nikon Z6 Nikon AF-S DX Nikkor 18-200mm f/3.5-5.6G ED VR II Nikon AF-S DX Nikkor 18-140mm F3.5-5.6G ED VR +2 more
Mark Scott Abeln
Mark Scott Abeln Forum Pro • Posts: 18,147
Re: X rite colorchecker passport magenta tint
4

Steve BB wrote:

Mark Scott Abeln wrote:

Steve BB wrote:

Are colour profiles meant to be 'exact'

No.

ok good to know

The biggest problem is that any particular visible color may be made up of a large number of different varieties of wavelengths. For example, a dark gray may be made up of all visible light frequencies equally, or it may be made up of only three narrow frequencies, or any combination of frequencies that balance so that it *appears* gray. If we are talking about gray paints, if they all appear to be the same under a given light source, then we say that the paint colors are metamers of each other, even though the precise combination of spectral frequencies they reflect happen to be different. As colors become more saturated, the range of potential metamers becomes quite narrow, until you get to spectral colors where there is only one.

The trouble is, real-life objects may very well not have the same spectral reflectance as the color patches on the ColorChecker target, even though they *appear* to be the same color, at least under a particular lighting. It's entirely possible that under a different light source (let's say in shade versus incandescent lighting) that these two colors won't match, and that's called metamerism failure. In consumer products, where components may be made of different materials, but it is desired that they always appear to be the same color, then it is important that there is a metameric match at least under common lighting conditions.

Cameras don't have the same spectral sensitivity as human vision, and this can lead to mismatch of colors, which is called device metamerism failure. Two materials, identical to the human eye, may appear different to the camera, while two materials identical to the camera may appear different to humans. This is the basic color difference between different models of camera, even if you take great care to shoot them under controlled conditions with custom processing. You can force colors to be anything that you want, but there will be metamerism failure that varies between models.

All of this color matching must be done under controlled conditions. A small neutral patch surrounded by a large brightly colored surrounding will *appear* to be different from the same patch surrounded by a large black area, which will in turn appear to be different from the same patch surrounded by a white. Also, colors, say in a print, will look different depending on the light levels and color of light falling upon the print. These effects are the source of a lot of optical illusions, but unfortunately all of this is hard to quantify and put into a general model of human vision. Cameras don't take "color appearance models" into account, and how could they? They can't predict the final viewing conditions of the image, and instead that is a part of the art of photography.

As mentioned, cameras don't have the same spectral sensitivity as does human color vision, and converting the raw values from the sensor to human color is not something that can be done precisely, at least without adjusting spot colors, which has its own problems. It's something like trying to fit a line to data in statistics: some values are going to fit well, and some are not, and it is something that requires a bit of engineering and artistic judgement. Do you want to fit most of colors as close as you can, even if some will poorly match, or do you want to minimize the worst match, even then no color is really good? Or do you want to force all colors on the ColorChecker to match exactly? This last solution is very likely to lead to problems with *other* colors not found on the ColorChecker, or it may lead to major hue shifts with exposure (this is called a 'twisted' profile). The only way to get a good match for all colors is to have a sensor that satisfies the "Luther-Ives condition" and none of them do, nor is this foreseen anytime in the near future.

or do they just help, and manual color corrections are to be expected afterwards?

Yes, unfortunately.

It helps if you use a "spectrally flat" light source, such as skylight, incandescent bulbs, and flash, which is close enough. You'll definitely have problems if fluorescent or LED bulbs are used.

Im using Bowens XMT 500s

Strobes are a "gray body" light source, but they are close enough to a spectrally flat light source for government work.

 Mark Scott Abeln's gear list:Mark Scott Abeln's gear list
Nikon D200 Nikon D7000 Nikon D750 Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF Nikkor 28mm f/2.8D +4 more
Mark Scott Abeln
Mark Scott Abeln Forum Pro • Posts: 18,147
Re: X rite colorchecker passport magenta tint

Steve BB wrote:

If I shoot against a white gallery wall and use the XMT500s as the only light source, and then set the white balance to the colour card, I find that this often gives warmer white walls than if I dropped the white balance on the white wall.

Cameras, raw processors, and camera profilers will typically saturate colors over what is strictly accurate. There is a good scientific reason for doing so, as images are typically viewed—in a print or on a monitor—far dimmer than how they were photographed, such as photography in sunlight, and human vision becomes less sensitive to contrast and saturation differences in dim lighting. So a brightly lit scene in a photograph needs more saturation and contrast to look good in a print when viewed under dim lighting.

But you are photographing a scene that is *already* dim, so you don't need to boost either saturation or contrast to make the scene appear to be good, as your photographs will be viewed under conditions similar to your subject. Your Canon camera's "Faithful" picture style might give you a better start, because as far as I know it does not add this additional contrast and saturation, and indeed, it tends to produce images that appear a bit dull and flat. But it is more accurate, and should give good results with your dimly lit scene.

I'm not sure how you can use the Faithful picture style as a basis for calibration, other than using special software and a lot of work. Or you just need to lower contrast and saturation, which may lead to the walls looking more accurate.

 Mark Scott Abeln's gear list:Mark Scott Abeln's gear list
Nikon D200 Nikon D7000 Nikon D750 Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF Nikkor 28mm f/2.8D +4 more
ggbutcher
ggbutcher Senior Member • Posts: 1,194
Re: X rite colorchecker passport magenta tint
1

Mark Scott Abeln wrote:

The most readable description of the difference between camera encoding and the reality of the scene I've yet read...

 ggbutcher's gear list:ggbutcher's gear list
Nikon D50 Nikon D7000 Nikon Z6 Nikon AF-S DX Nikkor 18-200mm f/3.5-5.6G ED VR II Nikon AF-S DX Nikkor 18-140mm F3.5-5.6G ED VR +2 more
Alex Marks New Member • Posts: 5
Re: X rite colorchecker passport magenta tint

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

Alex Marks New Member • Posts: 5
Re: X rite colorchecker passport magenta tint

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

To further clarify, I currently use the ICC profile for the majority of the work and splice in the regular adobe profile that gets the wrong colors much better on its own. So basically if there's a swath of red somewhere in the painting I borrow the adobe profile for that portion only and then for the greens I create my own profile with color manipulation in adobe raw to tune up the greens to the right color. Xrite gets the rest of them pretty good with a slight tweak needed for the yellows. Has anybody had better success with lumariver? Should I get that and do I need all the customization that comes with the repro level membership. Thanks in advance. Also I would be happy to pay for consultation with a repro pro.

ggbutcher
ggbutcher Senior Member • Posts: 1,194
Re: X rite colorchecker passport magenta tint
1

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

I'd dump the x-rite software. A ColorChecker with a proper calibration file should give you decent results. With that, I'd recommend Lumariver software, as it has the same processing core as dcamprof, a tool I've used extensively to do all that technical stuff. Lumariver will make either DCP or ICC profiles from your ColorChecker target shots that are colorimetrically consistent with the target.

http://www.lumariver.com/

<mind-numbing>
Now, if you are really interested in doing precise reproduction work, getting a spectral sensitivity dataset for your camera is the way to go.  With this data you won't need target shots; you'll use this data with a training color reference dataset to produce LUT-based camera profiles, which will give you better color transforms to render destinations.  Indeed, with this workflow you could make a training dataset which included patches measured from pigments of the artwork like Iilah suggested earlier.  To do this sort of thing, you'd need dcamprof and would need to learn how to use the command line.
</mind-numbing>

Disclaimer: I'm not affiliated with the Lumariver folks; I just use its core, dcamprof, and it is excellent software IMHO.

 ggbutcher's gear list:ggbutcher's gear list
Nikon D50 Nikon D7000 Nikon Z6 Nikon AF-S DX Nikkor 18-200mm f/3.5-5.6G ED VR II Nikon AF-S DX Nikkor 18-140mm F3.5-5.6G ED VR +2 more
Mark Scott Abeln
Mark Scott Abeln Forum Pro • Posts: 18,147
Re: X rite colorchecker passport magenta tint
1

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing.

That’s why reproduction photographers get paid the big bucks.

 Mark Scott Abeln's gear list:Mark Scott Abeln's gear list
Nikon D200 Nikon D7000 Nikon D750 Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF Nikkor 28mm f/2.8D +4 more
Entropy512 Veteran Member • Posts: 5,421
Re: X rite colorchecker passport magenta tint

ggbutcher wrote:

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

I'd dump the x-rite software. A ColorChecker with a proper calibration file should give you decent results. With that, I'd recommend Lumariver software, as it has the same processing core as dcamprof, a tool I've used extensively to do all that technical stuff. Lumariver will make either DCP or ICC profiles from your ColorChecker target shots that are colorimetrically consistent with the target.

http://www.lumariver.com/

<mind-numbing>
Now, if you are really interested in doing precise reproduction work, getting a spectral sensitivity dataset for your camera is the way to go. With this data you won't need target shots; you'll use this data with a training color reference dataset to produce LUT-based camera profiles, which will give you better color transforms to render destinations. Indeed, with this workflow you could make a training dataset which included patches measured from pigments of the artwork like Iilah suggested earlier. To do this sort of thing, you'd need dcamprof and would need to learn how to use the command line.
</mind-numbing>

Disclaimer: I'm not affiliated with the Lumariver folks; I just use its core, dcamprof, and it is excellent software IMHO.

Yup.  All of the bundled profiles in RawTherapee are generated using dcamprof

I'm not sure what lumariver may offer other than a GUI for those of us who are comfortable on the commandline, but for someone like Alex, lumariver is probably a solid choice.  It'll definitely be a better option than X-rite's own software - I never could get satisfactory results using it, although that was partly because I was trying to profile a camera with a fixed fisheye lens.  (There are workflows for feeding dcamprof a "defished" reference shot, but not for X-Rite's software...)

-- hide signature --

Context is key. If I have quoted someone else's post when replying, please do not reply to something I say without reading text that I have quoted, and understanding the reason the quote function exists.

 Entropy512's gear list:Entropy512's gear list
Sony a6000 Pentax K-5 Pentax K-01 Sony a6300 Canon EF 85mm F1.8 USM +5 more
OP Steve BB New Member • Posts: 22
Re: X rite colorchecker passport magenta tint

Mark Scott Abeln wrote:

Steve BB wrote:

If I shoot against a white gallery wall and use the XMT500s as the only light source, and then set the white balance to the colour card, I find that this often gives warmer white walls than if I dropped the white balance on the white wall.

Cameras, raw processors, and camera profilers will typically saturate colors over what is strictly accurate. There is a good scientific reason for doing so, as images are typically viewed—in a print or on a monitor—far dimmer than how they were photographed, such as photography in sunlight, and human vision becomes less sensitive to contrast and saturation differences in dim lighting. So a brightly lit scene in a photograph needs more saturation and contrast to look good in a print when viewed under dim lighting.

But you are photographing a scene that is *already* dim, so you don't need to boost either saturation or contrast to make the scene appear to be good, as your photographs will be viewed under conditions similar to your subject. Your Canon camera's "Faithful" picture style might give you a better start, because as far as I know it does not add this additional contrast and saturation, and indeed, it tends to produce images that appear a bit dull and flat. But it is more accurate, and should give good results with your dimly lit scene.

I'm not sure how you can use the Faithful picture style as a basis for calibration, other than using special software and a lot of work. Or you just need to lower contrast and saturation, which may lead to the walls looking more accurate.

Thanks for the explanation! Yes I've found lowering saturation and contrast does help with making a wrong colour look less wrong.

Is my scene 'already' dim? Not sure if I follow as it's shot with strobes, and also 'viewed' under bright 'daylight' temp fluorescent tubes.

 Steve BB's gear list:Steve BB's gear list
Canon EOS R5 Canon TS-E 17mm f/4L Canon TS-E 24mm f/3.5L II Canon EF 35mm F2 IS USM Sigma 50mm F1.4 DG HSM | A +2 more
OP Steve BB New Member • Posts: 22
Re: X rite colorchecker passport magenta tint

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

My reds are totally out of whack too. I don't get it. I mean they are really wrong for me.

My problem is that I shoot a painting, use the colorchart, go home, make a fresh profile, and edit the image's contrast crop etc, and then send the images. And sometimes get an email back saying the colours are 'wrong' and they send an iPhone image of what is 'right'. 

I'm really looking for a colour chart and profile making software that is 'good enough' to trust when I don't have the artwork next to me, to put faith in sending the images to the client.

 Steve BB's gear list:Steve BB's gear list
Canon EOS R5 Canon TS-E 17mm f/4L Canon TS-E 24mm f/3.5L II Canon EF 35mm F2 IS USM Sigma 50mm F1.4 DG HSM | A +2 more
Iliah Borg Forum Pro • Posts: 28,807
Re: X rite colorchecker passport magenta tint

ggbutcher wrote:

Iliah Borg wrote:

ggbutcher wrote:

The guy at MaxMax just piped the monochromator output into a fiber optic cable and pointed it directly at the sensor on the lens-less camera... ??

Not exactly 'just'. Dan wrote: "We split the monochomatic light output between two fiber optics cables. One was routed to our spectrometer and the other was routed to the camera."

Fair 'nuf. But, the salient point is, didn't use an integrating sphere.

In my early experiments I bounced the monochromator beam off a gray card, or used a diffuser. Later I bought an integrating sphere for $129 (ebay). Next, I found out a sphere is easy DIY.

One of the questions here is the uniformity of the sampling area. However, for analyzing converted JPEGs one doesn't need much accuracy.

Yes, was thinking along those lines as I posted the response. Could insert a diffuser and essentially take a picture of the blob, but then there's the diffuser and lens in the path to factor into the measurement. Splaying the light directly on the sensor, with only the fiber to account for, seems to be the least complicated way to accurately present light for measurement.

https://maxmax.com/faq/camera-tech/spectral-response

If I were wanting to go into the business of measuring cameras' spectral response, a monochromator would be my first choice.

Consider OL 490 Agile Light Source.

But I only have three cameras here in my basement, and not much motivation to chase the measurement of many others. For this use case, the homebuilt $130US single-shot spectrometer approach renders data of sufficient quality, within 2 dE (24-patch ColorChecker) of a monochromator-based profile for the same camera.

-- hide signature --
Iliah Borg Forum Pro • Posts: 28,807
Re: X rite colorchecker passport magenta tint

Steve BB wrote:

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

My reds are totally out of whack too. I don't get it. I mean they are really wrong for me.

My problem is that I shoot a painting, use the colorchart, go home, make a fresh profile, and edit the image's contrast crop etc, and then send the images. And sometimes get an email back saying the colours are 'wrong' and they send an iPhone image of what is 'right'.

I'm really looking for a colour chart and profile making software that is 'good enough' to trust when I don't have the artwork next to me, to put faith in sending the images to the client.

https://www.dpreview.com/forums/post/65477038

-- hide signature --
OP Steve BB New Member • Posts: 22
Re: X rite colorchecker passport magenta tint

ggbutcher wrote:

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

I'd dump the x-rite software. A ColorChecker with a proper calibration file should give you decent results. With that, I'd recommend Lumariver software, as it has the same processing core as dcamprof, a tool I've used extensively to do all that technical stuff. Lumariver will make either DCP or ICC profiles from your ColorChecker target shots that are colorimetrically consistent with the target.

http://www.lumariver.com/

<mind-numbing>
Now, if you are really interested in doing precise reproduction work, getting a spectral sensitivity dataset for your camera is the way to go. With this data you won't need target shots; you'll use this data with a training color reference dataset to produce LUT-based camera profiles, which will give you better color transforms to render destinations. Indeed, with this workflow you could make a training dataset which included patches measured from pigments of the artwork like Iilah suggested earlier. To do this sort of thing, you'd need dcamprof and would need to learn how to use the command line.
</mind-numbing>

Disclaimer: I'm not affiliated with the Lumariver folks; I just use its core, dcamprof, and it is excellent software IMHO.

Thanks for the heads up! I'm going to download this. I'm just looking for something that is better than xrite but not as precise as 'mind-numbing'

Getting samples of the artwork or paint to make datasets: I can't see how this is feasible realistically as a gigging photographer, as I shoot somewhere different each day, not all the artworks are paintings. And if they are paint, then there's acrylic, oil, how much thinner did they use? is the painting glazed? What colour was the base gesso underneath the paint etc etc etc.

 Steve BB's gear list:Steve BB's gear list
Canon EOS R5 Canon TS-E 17mm f/4L Canon TS-E 24mm f/3.5L II Canon EF 35mm F2 IS USM Sigma 50mm F1.4 DG HSM | A +2 more
OP Steve BB New Member • Posts: 22
Re: X rite colorchecker passport magenta tint

Not much.

Interesting that my mention of 'reproduction-grade' went unnoticed.

Out-of-the-box X-Rite profiles don't play well with default settings in Adobe converters.

One needs to modify the settings in a converter and, ideally, modify the profile itself.

https://www.rawdigger.com/howtouse/overriding-raw-converter-default-adjustments-settings and the comment where we suggest the modification of the profiles,

Hi Lilah, sorry I missed your comment. Please bear with me, just so I understand:

When you say x-rite profiles dont play with with default settings in Adobe converters, do you mean Lightroom reading/converting the profile that x-rite plug-in made?

> exiftool -ProfileToneCurve="0 0 1 1" -o new.dcp old.dcp

Oh i see now in the comments on the webpage you sent. Is the purpose of this to edit the tone curve of the xrite profile say, or to edit the tone curve of an exported image?

 Steve BB's gear list:Steve BB's gear list
Canon EOS R5 Canon TS-E 17mm f/4L Canon TS-E 24mm f/3.5L II Canon EF 35mm F2 IS USM Sigma 50mm F1.4 DG HSM | A +2 more
Alex Marks New Member • Posts: 5
Re: X rite colorchecker passport magenta tint

Steve BB wrote:

ggbutcher wrote:

Alex Marks wrote:

Wow, I'm doing repro work and this thread is basically mind numbing. Is this all to say using the large color checker and the xrite software is for the birds? My reds are hopeless and currently so am I. I only ask this because it seems as tho none of these responses address a way to get a solid profile from this color checker without getting heavily technical and using a spectrometer. Yikes. Also, greens are a problem for me. Just to make it real confusing I'm using cross polarization as well.

I'd dump the x-rite software. A ColorChecker with a proper calibration file should give you decent results. With that, I'd recommend Lumariver software, as it has the same processing core as dcamprof, a tool I've used extensively to do all that technical stuff. Lumariver will make either DCP or ICC profiles from your ColorChecker target shots that are colorimetrically consistent with the target.

http://www.lumariver.com/

<mind-numbing>
Now, if you are really interested in doing precise reproduction work, getting a spectral sensitivity dataset for your camera is the way to go. With this data you won't need target shots; you'll use this data with a training color reference dataset to produce LUT-based camera profiles, which will give you better color transforms to render destinations. Indeed, with this workflow you could make a training dataset which included patches measured from pigments of the artwork like Iilah suggested earlier. To do this sort of thing, you'd need dcamprof and would need to learn how to use the command line.
</mind-numbing>

Disclaimer: I'm not affiliated with the Lumariver folks; I just use its core, dcamprof, and it is excellent software IMHO.

Thanks for the heads up! I'm going to download this. I'm just looking for something that is better than xrite but not as precise as

Getting samples of the artwork or paint to make datasets: I can't see how this is feasible realistically as a gigging photographer, as I shoot somewhere different each day, not all the artworks are paintings. And if they are paint, then there's acrylic, oil, how much thinner did they use? is the painting glazed? What colour was the base gesso underneath the paint etc etc etc.

Yes, agreed. There isn't a snow ball's chance on that concept for me.

In regards to lumariver, if I just need to make profiles from DNG for adobe can it be done with the basic mode?

I read about the 3D LUT and I'm not sure if I will need that type of customization. Are you guys using this software in repro edition?

Is adobe problematic regardless of profile maker?

I use canon so no capture one at the moment tho I have it for my fuji.

Entropy512 Veteran Member • Posts: 5,421
Re: X rite colorchecker passport magenta tint

Steve BB wrote:

Not much.

Interesting that my mention of 'reproduction-grade' went unnoticed.

Out-of-the-box X-Rite profiles don't play well with default settings in Adobe converters.

One needs to modify the settings in a converter and, ideally, modify the profile itself.

https://www.rawdigger.com/howtouse/overriding-raw-converter-default-adjustments-settings and the comment where we suggest the modification of the profiles,

Hi Lilah, sorry I missed your comment. Please bear with me, just so I understand:

When you say x-rite profiles dont play with with default settings in Adobe converters, do you mean Lightroom reading/converting the profile that x-rite plug-in made?

> exiftool -ProfileToneCurve="0 0 1 1" -o new.dcp old.dcp

What does this code relate to? Does it go in Raw Digger?

There are multiple components of a DCP profile, and one of those is called a Tone Curve.

It appears that the default setting of most Adobe products is to use whatever the tone curve in the profile is - which means that if a profile has a baked-in "look", it will have effects on the colors.

Some raw processors make it easier to see the tone curve being applied - and in the case of RawTherapee, defaults to ignoring any tone curve embedded in the DCP profile (but will apply its own if you choose certain presets) - RawTherapee's documentation on tone curves is at https://rawpedia.rawtherapee.com/Exposure#Tone_Curves

You can pretty much assume that if you use a "typical" DCP profile which includes a defined tone curve in an Adobe product, you'll get whatever tone curve is in the profile applied, using an algorithm similar to what RawTherapee calls "Film-Like" - https://rawpedia.rawtherapee.com/Exposure#Film-Like - this will alter saturation of colors depending on their luminance.

The exiftool command that Iliah gave will take whatever tone curve happens to be in the profile and replace it with a linear mapping (which effectively disables the tone curve feature). This leads to a "flat/neutral" starting point.

The documentation on dcamprof also goes into detail regarding tone curves and tone reproduction - http://rawtherapee.com/mirror/dcamprof/dcamprof.html#tone_curves

-- hide signature --

Context is key. If I have quoted someone else's post when replying, please do not reply to something I say without reading text that I have quoted, and understanding the reason the quote function exists.

 Entropy512's gear list:Entropy512's gear list
Sony a6000 Pentax K-5 Pentax K-01 Sony a6300 Canon EF 85mm F1.8 USM +5 more
Iliah Borg Forum Pro • Posts: 28,807
Re: X rite colorchecker passport magenta tint

Steve BB wrote:

> exiftool -ProfileToneCurve="0 0 1 1" -o new.dcp old.dcp

What does this code relate to?

Forces linear rendering, using a free command-line utility ExifTool.

Start with hacking your dcp profile with exiftool, and when using the hacked dcp in Adobe converters, reset the converter curve to linear too, and switch the process version to Process Version 2. This 3 workflow modifications together make for a good starting point.

Alternatively, if not willing to hack dcp files, just modify all your Adobe converter settings as suggested in the article.

Does it go in Raw Digger?

RawDigger in that article was used only for illustrations of what is going behind the scenes.

-- hide signature --
Mark Scott Abeln
Mark Scott Abeln Forum Pro • Posts: 18,147
Re: X rite colorchecker passport magenta tint

Steve BB wrote:

Is my scene 'already' dim? Not sure if I follow as it's shot with strobes,

Yeah, that gets kind of difficult conceptually. Of course, you don't see the painting under the strobes—only the camera does.

and also 'viewed' under bright 'daylight' temp fluorescent tubes.

I still would think that actual direct sunlight is many times brighter. A very bright gallery may be 5% as bright as daylight, and more typically about 1% as bright. Art museum galleries, with light sensitive objects, tend to be extraordinarily dim, about 1/2000th as bright as direct sunlight or maybe even less.

But I wonder if the color rendering of the fluorescents might be a problem—using strobes may not give the same impression as viewing the paintings under their original fluorescent lighting.

One thing that worked with me when I did photography for color sensitive clients, was finding a matching color sample—this is pretty easy for paints—and the manufacturers often specify color values for it.

 Mark Scott Abeln's gear list:Mark Scott Abeln's gear list
Nikon D200 Nikon D7000 Nikon D750 Nikon AF-S DX Nikkor 35mm F1.8G Nikon AF Nikkor 28mm f/2.8D +4 more
Keyboard shortcuts:
FForum MMy threads