Jpeg Magenta Cast fix using Photoshop

Raw undoubtedly produces stunningly accurate colour straight out of the G1.Jpeg on the other hand has a distinct magenta cast. Shooting exclusively in Raw is the obvious answer. However if you are caught for memory and have access to Photoshop, the above fix will restore things to where the difference is barely perceptible, if at all. It is surprising that Canon has'nt fixed this annoying quirk in the firmware. However Photoshop users can record Color Balance adjustments as an Action and apply it instantaneously to any jpeg imported from the G1 into Photoshop either singly or in batch form since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default with the 'midtones' radio button checked. Change this to 'Highlights'.

Kevin

Hi Kevin,

We learn more everyday!;-) I've messed around with correcting in photoshop, but never created an action for handling this...just lazy, I guess. Perhaps I'll do that the next time I have to deal with magenta casts. Thanks for the info!
nahau

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Kevin

Click to expand...

Hi nahau,

Yes, Actions are a great timesaver for any set of instructions that you want to repeat. They are useful for mundane jobs like resizing to an often used format or for more complicated ones like Unsharp Mask in conjunction with Blending Layers which can be immensely time consuming. After a while you can have an Action for everything but making the coffee.

Kevin

nahau said:

Hi Kevin,
We learn more everyday!;-) I've messed around with correcting in
photoshop, but never created an action for handling this...just
lazy, I guess. Perhaps I'll do that the next time I have to deal
with magenta casts. Thanks for the info!
nahau

Click to expand...

Thanks for the tip. Are there any more photoshop tips that you can enlighten us with - e.g. removing objects via cloning, enhancing sky, improving indoor (internal flash) skin tones etc?

Thanks in advance.

Marc

Does anyone know how this translates in Photoshop Elements? I have the 30-day trial now.

Using my old camera, I had a set of actions that could be applied
to every shot most of the time. I find that with the G-1, every
shot is so different that no automation is possible. Your action
worked with the posted shot, but will it work with every shot?
I suspect that it will not, and as such becomes useless as a
recorded action.

Marc said:

Thanks for the tip. Are there any more photoshop tips that you can
enlighten us with - e.g. removing objects via cloning, enhancing
sky, improving indoor (internal flash) skin tones etc?

Thanks in advance.

Marc

Click to expand...

Sorry Nancy - I've got Elements on as it's a nice 'n easy alternative for lots of things but it does not do 'Actions' as far as I'm aware .. it's made more easy than PS proper so these things are not included.

FOR nahau - go one, do it, I just made one in a couple of minutes. It's not likely to be be needed very often on my pics anyway as I just don't seem to suffer much this way but it's useful to be there at a click

EJN

Nancy said:

Does anyone know how this translates in Photoshop Elements? I have
the 30-day trial now.

Click to expand...

Hi Bob,

My main interest involves landscapes and seascapes and so far I have found the magenta cast to be fairly predictable. You may well be correct with other subjects. However, you can record a vast number of Color Balance Variations and use which ever Action seems appropriate. The relationship between red and blue (2:1) seems constant at least. Removing just red solves the red problem but makes the image cold. Half the red adjustment in blue seems to correct this. Therefore you could also have -10r -5b; -12r -6b and so on. You can also apply one action twice to double the values in extreme cases. Changing Actions to Button Mode (options) prevents your Actions pallette from becoming cluttered.

Kevin

Bob Williams said:

Marc said:

Thanks for the tip. Are there any more photoshop tips that you can
enlighten us with - e.g. removing objects via cloning, enhancing
sky, improving indoor (internal flash) skin tones etc?

Thanks in advance.

Marc

Click to expand...

Click to expand...

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Kevin

Click to expand...

This is the type of posting that makes this forum worthwhile! I created the action in PhotoShop (good practice for an overlooked procedure) and on my first test picture Kevin's color correction was right-on!

Obviously, cameras and shooting conditions may vary a little, but having the groundwork figured out got me pointed in the right direction. I can always modify the settings if necessary to fine tune MY computer and printer to MY G1.

Hi Marc,

Getting the correct balance between exposure for the sky and the land is always a problem because the difference in values frequently stretches the dynamic range too far. In darkroom days this was achieved with a best compromise exposure and a bit of dodging and burning. You can achieve much the same in Photoshop with the dodge and burn tool. However Photoshop gives you quite a few more powerful tools. A couple of quick ones:

The Curve tool can make very fast adjustment to the light and dark areas of an image but needs to be used carefully. However it can very easily preserve the values for a sky whilst lifting a 'moderately' under-exposed land value. Experiment.

Using a copy of an image on another Layer is very effective. Use levels to correct the sky value on one copy and do the same for the land on the other. Then use the Eraser to allow the corrected part show through. At the edges reduce the opacity and and possibly the size of brush so that there is a smooth transition. Done carefully the transformation is seamless. Then just flatten the image.

Kevin

Marc said:

Thanks for the tip. Are there any more photoshop tips that you can
enlighten us with - e.g. removing objects via cloning, enhancing
sky, improving indoor (internal flash) skin tones etc?

Thanks in advance.

Marc

Click to expand...

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Click to expand...

Dear Kevin,

You may or may not have seen my post of June 30th discussing the magenta problem. It is in general agreement with your assesment of the problem as I came to the same conclusion about shooting in RAW. ( although I shoot exclusively in JPG )

In that experiment a Kodak gray card was inserted into the view so that actual measured values could be made. In those images the gray card showed a mgenta bias of 5 to 10 units, easily corrected with a simple increase in Green of about 10.

However ,( shooting in bright sunlight at least ) the color cast is not always magenta and your standard correction may not be applicable.
The image below shows a slight red excess, not magenta.



A -7 Red correction was made in the next image below:



The finished image with the red correction included below:



Some thoughts about the slight magenta colorcast:

1. For most pictures with fairly saturated colors you will not see any excess magenta. The image must have a neutral or near neutral to see the magenta color shift.

2. Shooting in JPG will produce just as good a final image as shooting in RAW. The correction should be done on a case by case basis.

3. If one is serious about a perfect color balance he sholud consider inserting a gray card into the field of view. This is a technique used by us old print makers where getting a good color print ( Ektacolor that is)was a time consuming effort.

Regards,
Don T

Hi Don,

No, I had not seen your previous posts but I looked them up and I can see that it is a subject that you have given some thought. I could not agree more with your remarks regarding neutral grey, and grey cards in particular in detecting colour casts.

The color correction illustrated previously is to emulate the output from raw which makes the assumption, probably wrongly, that raw is cast free in all situations. In general, raw seems to be extremely accurate.

I am not convinced that emulating raw with a colour correction is a matter of case by case but experience tells me to hold fire on that one. So far my G1 produces a 'magenta' cast in jpeg with a remarkable degree of consistency. Maybe I'm just lucky.

I got out my grey card this morning and took a raw and a jpeg in the shade with 200th at f2.2 as distinct from the flower shot in direct midday sunlight with 250th at f8. As you can see, with raw at the top, the correction is fairly close. The reason for the second correction adjacent to the raw is that subtracting color lightens the overall image slightly This adjustment just involves lowering the output Level from 255 to 250 of thereabouts and is more for illustration than practical benefit but could also be included in a PS Action.

Kevin

Don T said:

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Click to expand...

Dear Kevin,
You may or may not have seen my post of June 30th discussing the
magenta problem. It is in general agreement with your assesment of
the problem as I came to the same conclusion about shooting in RAW.
( although I shoot exclusively in JPG )
In that experiment a Kodak gray card was inserted into the view so
that actual measured values could be made. In those images the gray
card showed a mgenta bias of 5 to 10 units, easily corrected with a
simple increase in Green of about 10.
However ,( shooting in bright sunlight at least ) the color cast is
not always magenta and your standard correction may not be
applicable.
The image below shows a slight red excess, not magenta.



A -7 Red correction was made in the next image below:



The finished image with the red correction included below:



Some thoughts about the slight magenta colorcast:
1. For most pictures with fairly saturated colors you will not see
any excess magenta. The image must have a neutral or near neutral
to see the magenta color shift.
2. Shooting in JPG will produce just as good a final image as
shooting in RAW. The correction should be done on a case by
case basis.
3. If one is serious about a perfect color balance he sholud
consider inserting a gray card into the field of view. This is a
technique used by us old print makers where getting a good color
print ( Ektacolor that is)was a time consuming effort.

Regards,
Don T

Click to expand...

Kevin McLoughlin said:

Hi Don,

No, I had not seen your previous posts but I looked them up and I
can see that it is a subject that you have given some thought. I
could not agree more with your remarks regarding neutral grey, and
grey cards in particular in detecting colour casts.

The color correction illustrated previously is to emulate the
output from raw which makes the assumption, probably wrongly, that
raw is cast free in all situations. In general, raw seems to be
extremely accurate.

I am not convinced that emulating raw with a colour correction is a
matter of case by case but experience tells me to hold fire on that
one. So far my G1 produces a 'magenta' cast in jpeg with a
remarkable degree of consistency. Maybe I'm just lucky.

I got out my grey card this morning and took a raw and a jpeg in
the shade with 200th at f2.2 as distinct from the flower shot in
direct midday sunlight with 250th at f8. As you can see, with raw
at the top, the correction is fairly close. The reason for the
second correction adjacent to the raw is that subtracting color
lightens the overall image slightly This adjustment just involves
lowering the output Level from 255 to 250 of thereabouts and is
more for illustration than practical benefit but could also be
included in a PS Action.

Kevin

Click to expand...

Thanks for the quick reply.

I am am having difficulty understanding your latest images. It is not clear to me where one image ends and the next begins. But no matter, we agree on using a gray card and will both continue to develop more experience with the dreadded magenta problem.

My main point was that the bias is small and does not always have to be corrected, depending the picture content.

While not discussed in my earlier post, I shot a whole bunch of views ( in JPG) specifically looking for magenta casts. These included a series of clouds both overcast and beautiful cumulus in a clear sky.

I did not find any "pink" clouds. In general all the views were free from any unwanted color aberation leading me to believe the magenta problem is not as bad as some believe.
New Concern:

I now find after more than six months useage that the G1 has a deficient autofocus system. Until I read Kevin Bjorke's treatise on the G1 system, I did not know the G1 does not use the center of the image for autofocus,
that it uses the entire CCD .

I could not understand why objects in the background seemed to be in better focus than ,for example a flower bloom right in the cener of the image. And, placing a "target" in the same plane as the bloom to increase the object contrast wont work because the system sees around the target.I did get the concept to work by using a bright labled plastic container which was large enough to fill the whole image.Not easy to do.

I believe this is a more important area of concern than the magentta issue
because it is not as easy to correct.

Regards,
Don T

Kevin M said:

Don T said:

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Click to expand...

Dear Kevin,
You may or may not have seen my post of June 30th discussing the
magenta problem. It is in general agreement with your assesment of
the problem as I came to the same conclusion about shooting in RAW.
( although I shoot exclusively in JPG )
In that experiment a Kodak gray card was inserted into the view so
that actual measured values could be made. In those images the gray
card showed a mgenta bias of 5 to 10 units, easily corrected with a
simple increase in Green of about 10.
However ,( shooting in bright sunlight at least ) the color cast is
not always magenta and your standard correction may not be
applicable.
The image below shows a slight red excess, not magenta.



A -7 Red correction was made in the next image below:



The finished image with the red correction included below:



Some thoughts about the slight magenta colorcast:
1. For most pictures with fairly saturated colors you will not see
any excess magenta. The image must have a neutral or near neutral
to see the magenta color shift.
2. Shooting in JPG will produce just as good a final image as
shooting in RAW. The correction should be done on a case by
case basis.
3. If one is serious about a perfect color balance he sholud
consider inserting a gray card into the field of view. This is a
technique used by us old print makers where getting a good color
print ( Ektacolor that is)was a time consuming effort.

Regards,
Don T

Click to expand...

Click to expand...

I have to agree with Don's assessment. The amount of magenta appears to be variable based on the lighting temperature and color saturation. The magenta cast appears worst at 5400+ K (sunlight and overcast conditions) and negligible at other times. Photographs whose subjects contain a wide range of color saturation tend to demonstrate less magenta shift.

I have to wonder why the jpg algorithm would produce this? A more important question, though I have not yet tried this, is to see if performing a custom WB before shooting in these conditions would eliminate or reduce the color shift. At least in this way, one could determine if the shift is truely due to some jpg processing artifact or whether it is a software WB problem. Just my $0.02

Don T said:

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Click to expand...

Dear Kevin,
You may or may not have seen my post of June 30th discussing the
magenta problem. It is in general agreement with your assesment of
the problem as I came to the same conclusion about shooting in RAW.
( although I shoot exclusively in JPG )
In that experiment a Kodak gray card was inserted into the view so
that actual measured values could be made. In those images the gray
card showed a mgenta bias of 5 to 10 units, easily corrected with a
simple increase in Green of about 10.
However ,( shooting in bright sunlight at least ) the color cast is
not always magenta and your standard correction may not be
applicable.
The image below shows a slight red excess, not magenta.



A -7 Red correction was made in the next image below:



The finished image with the red correction included below:



Some thoughts about the slight magenta colorcast:
1. For most pictures with fairly saturated colors you will not see
any excess magenta. The image must have a neutral or near neutral
to see the magenta color shift.
2. Shooting in JPG will produce just as good a final image as
shooting in RAW. The correction should be done on a case by
case basis.
3. If one is serious about a perfect color balance he sholud
consider inserting a gray card into the field of view. This is a
technique used by us old print makers where getting a good color
print ( Ektacolor that is)was a time consuming effort.

Regards,
Don T

Click to expand...

Hi Don,

The image is four overlapping layers with the jpeg straight fom the camera being the largest at the bottom half of the frame. The next one is the jpeg colour corrected. The next one is the colour correction adjusted for the resulting alteration of Brightness. And the top one (lowest layer) is the raw image.

(By the way I shot another test this afternoon in the rain and set the white point. The magenta cast was exactly the same. )

Re Auto focus:

I noticed quite recently from exif data that in a series of frames of the same subject that the focus distance occasionally jumped to infinity even though they appeared to be identical in viewpoint. I thought it very odd at the time. (Also six months)

Kevin

Don T said:

I believe this is a more important area of concern than the
magentta issue
because it is not as easy to correct.

Regards,
Don T

Donald b Terrana said:

Don T said:

Kevin McLoughlin said:

Raw undoubtedly produces stunningly accurate colour straight out of
the G1.Jpeg on the other hand has a distinct magenta cast. Shooting
exclusively in Raw is the obvious answer. However if you are caught
for memory and have access to Photoshop, the above fix will restore
things to where the difference is barely perceptible, if at all. It
is surprising that Canon has'nt fixed this annoying quirk in the
firmware. However Photoshop users can record Color Balance
adjustments as an Action and apply it instantaneously to any jpeg
imported from the G1 into Photoshop either singly or in batch form
since it appears that the cast is reasonably consistent.

The numbers were arrived at on a calibrated Sony Trinitron Monitor.

Please note that Photoshop's Color Balance tool opens by default
with the 'midtones' radio button checked. Change this to
'Highlights'.

Click to expand...

Dear Kevin,
You may or may not have seen my post of June 30th discussing the
magenta problem. It is in general agreement with your assesment of
the problem as I came to the same conclusion about shooting in RAW.
( although I shoot exclusively in JPG )
In that experiment a Kodak gray card was inserted into the view so
that actual measured values could be made. In those images the gray
card showed a mgenta bias of 5 to 10 units, easily corrected with a
simple increase in Green of about 10.
However ,( shooting in bright sunlight at least ) the color cast is
not always magenta and your standard correction may not be
applicable.
The image below shows a slight red excess, not magenta.



A -7 Red correction was made in the next image below:



The finished image with the red correction included below:



Some thoughts about the slight magenta colorcast:
1. For most pictures with fairly saturated colors you will not see
any excess magenta. The image must have a neutral or near neutral
to see the magenta color shift.
2. Shooting in JPG will produce just as good a final image as
shooting in RAW. The correction should be done on a case by
case basis.
3. If one is serious about a perfect color balance he sholud
consider inserting a gray card into the field of view. This is a
technique used by us old print makers where getting a good color
print ( Ektacolor that is)was a time consuming effort.

Regards,
Don T

Click to expand...

Click to expand...

Click to expand...

Are actions doable in PhotoShop LE?

I haven't installed it yet.. but if they are I will

Or even PSP Pro 7?

CJ

EJN said:

FOR nahau - go one, do it, I just made one in a couple of minutes.
It's not likely to be be needed very often on my pics anyway as I
just don't seem to suffer much this way but it's useful to be there
at a click

EJN

Nancy said:

Does anyone know how this translates in Photoshop Elements? I have
the 30-day trial now.

Click to expand...

Click to expand...

CJ said:

Are actions doable in PhotoShop LE?

Click to expand...

YES!

Adam said:

I have to agree with Don's assessment. The amount of magenta
appears to be variable based on the lighting temperature and color
saturation. The magenta cast appears worst at 5400+ K (sunlight
and overcast conditions) and negligible at other times.
Photographs whose subjects contain a wide range of color saturation
tend to demonstrate less magenta shift.

I have to wonder why the jpg algorithm would produce this? A more
important question, though I have not yet tried this, is to see if
performing a custom WB before shooting in these conditions would
eliminate or reduce the color shift. At least in this way, one
could determine if the shift is truely due to some jpg processing
artifact or whether it is a software WB problem. Just my $0.02

Click to expand...

Dear Afam,

I have tried using custom white balance based on the gray card and there is no change from using daylight for white balance( shooting in daylight of course).

As regarding a image with with a wide range of color saturation , my point is not that there is less magenta shift, but that you just cant see it.

You can kind of prove this for yoursaelf by painting over a vivid color (in Photoshop, IE)with a a 5 unit magenta color. While you can measure that the new color has increased 5 units of magenta it will not be visible to the naked eye.

Is it possible that there is no variation from camera to camera that might explain why I don't seem to get the constant magenta shift that Kevin does?
Regards,
Don T

Kevin McLoughlin said:

Hi Don,

The image is four overlapping layers with the jpeg straight fom the
camera being the largest at the bottom half of the frame. The next
one is the jpeg colour corrected. The next one is the colour
correction adjusted for the resulting alteration of Brightness. And
the top one (lowest layer) is the raw image.

(By the way I shot another test this afternoon in the rain and set
the white point. The magenta cast was exactly the same. )

Re Auto focus:

I noticed quite recently from exif data that in a series of frames
of the same subject that the focus distance occasionally jumped to
infinity even though they appeared to be identical in viewpoint. I
thought it very odd at the time. (Also six months)

Kevin

Click to expand...

Dear Kevin,

Thanks for the reply which is very clear. Your test certainly shows a very obvious magenta bias in JPG. As to why my tests are less conclusive is a mystery to me. Could there be a variation from camera to camera ?

I will shoot some more pictures of the gray card in differing light conditions to build a minidata base and we can compare notes sometime.

Regarding autofocus, suggest you look at Bjorke's web site (botzilla.com) and review his description of the G1 autofocus situation.

My late annoyance stems from this new to me revalation that the system does not narrow its focusing area the way it does for exposure metering.
How could I have missed that in reading G1 reviews before I bought the camera?
How does the G1 system compare to the Nikon 995 and Olympus 3000 series?

Regards,
Don T

Has anyone tried this correction with the pink sky problem? I
noticed that one poster said they did not get the pink clouds.
I get them all the time. My experiments with removing the pink
using a global color shift have not been very satisfying. Editing
on a portion of each image is not a real option for every day
shots either. I can supply a clear example if needed.

Don T said:

Adam said:

I have to agree with Don's assessment. The amount of magenta
appears to be variable based on the lighting temperature and color
saturation. The magenta cast appears worst at 5400+ K (sunlight
and overcast conditions) and negligible at other times.
Photographs whose subjects contain a wide range of color saturation
tend to demonstrate less magenta shift.

I have to wonder why the jpg algorithm would produce this? A more
important question, though I have not yet tried this, is to see if
performing a custom WB before shooting in these conditions would
eliminate or reduce the color shift. At least in this way, one
could determine if the shift is truely due to some jpg processing
artifact or whether it is a software WB problem. Just my $0.02

Click to expand...

Dear Afam,
I have tried using custom white balance based on the gray card and
there is no change from using daylight for white balance( shooting
in daylight of course).
As regarding a image with with a wide range of color saturation ,
my point is not that there is less magenta shift, but that you just
cant see it.
You can kind of prove this for yoursaelf by painting over a vivid
color (in Photoshop, IE)with a a 5 unit magenta color. While you
can measure that the new color has increased 5 units of magenta it
will not be visible to the naked eye.

Is it possible that there is no variation from camera to camera
that might explain why I don't seem to get the constant magenta
shift that Kevin does?
Regards,
Don T

Click to expand...