Hue Clock

[mikemeister_admin]

I am having problems with skin tones. First, when I use a skin tone and it gives me a reading, I can not seem to adjust the reading when moving curves.  I then tried the other hue clock on the same spot.  Different readings. It is difficult to describe, maybe the snap will help. It must be a seniors moment!

[mikemeister_admin]

I am having problems with skin tones. First, when I use a skin tone and it gives me a reading, I can not seem to adjust the reading when moving curves.  I then tried the other hue clock on the same spot.  Different readings. It is difficult to describe, maybe the snap will help. It must be a seniors moment!

BARRY,
INDEED I DO NOT UNDERSTAND!
wHAT i SEE ON YOUR PICTURE IS NOT A SKINTONE, BUT A HUE CLOCK WICH GIVES THE HUE OF THE FACE.
THE HUE CLOCK ASIDE IS A OTHER PLACE i THINK.
wHY YOU CAN'T USE THE SKIN TONES ?

Frits
p.s. I see know afterwards that the color of the clock not match the skintone.
that is strange.

---

have you made the left clock active by click on the small circle on the face ?

[mikemeister_admin]

Barry,
load down this .avi file:
http://members.chello.nl/f.egelie/Barry.avi
Can you do exactly as I do ?
Yes then you have no problem.

Frits

[-default]

I don't have an explanation for what you are seeing, nor can I make it happen on my system.  I am very interested in chasing this down, since this is an apparent bug that could affect a lot of people.  Here are a few things that might be relevant. 

The behavior you mention regarding the skin tone pin may be normal.  A skin tone pin will "lock in" the the curve control points so that you cannot move the relevant curve control points normally, unless you "un-pin" the icon in the lower right corner of the pin's sample palette.

The floating hue clock operates by sampling the screen directly, and might be indicating the combined red mask plus the underlying skin tone.

I don't have an explanation for the sample point continuing to indicate the old value for the baby's face - on my system your screen capture shows the baby's image with a good skin tone, although the contents of the palette shows the original magenta.

One next step to track this down would be to simplify the situation.  See if the problem happens with a small image containing just the baby's face, then add a selection, etc, until the problem shows itself.

I've enclosed the original, somewhat larger, image of the baby.

[mikemeister_admin]

When I used the original sample, every thing was fine as per hueclock 2. This was true with or without a selection.  You can see in hueclock 3, the readings are different thatin hue clock 2, and it was done as before. This time I could not get the floating clock to register and snap it.
I am not sure this helps you, but it looks to me that the rose backgorund is being read some how when the selection on the large photo is used.

[mikemeister_admin]

Barry,
I have tried again now, again this sequense.
But al worked right with me.
So it looks it is a bug.
So far I know , you are the first person with such a problem.
I have often participate, that prgrs could conflict with others.
A big example is for example Perfectdisk 8.0 defragmenter.
But it is not Perfectdisk 8.0 wat causes the problem.
Perfectdisk 8.0 is Microsoft  certified.
they are other prgrs wich conflicting.
I say not it is CM. It could.
Perhaps the frst thing you can do, is completely deinstall cM3.
also clean the registry on that point.
Restart the computer and install CM3 again.
Perhaps it is also wise to uninstall photshop X or elements.
Also close down all other prgr's could sometimes help.
The system loads several drives and also .dll files.
You could instruct your system to unload all the .dll files it used not on that moment.
there are lots of possibilities what could be the cause.
give it a try.


frits

[-default]

This is probably a bug involving selection processing for eyedropper samples.  If I can get it to happen on my system, it *should* be pretty easy to fix. 

In the meantime, it would help to know if the problem changes depending on how the image is zoomed or panned.

[mikemeister_admin]

i will try it again.
but the orginal baby is no a big picture.
You have bigger. ?
I could zoom of course the smaller one.

frits

p.s. i have tried it agai non several ways. It stays good on my system.
so what to do. If Barry is the only person with  this problem then his system configuration included the prgr's are responsible.

---

P.S 2 Barry good als do the following. Uninstall his CM3 version and going back one version.
when the problem is vanished then it is probably CM3 wich is the cause, in conjuction with a other unnown prgr conflict.

---

When i have a big problem and can't  find the solution, then I have also a other way tio find a solution.
I  make use of First Defence-ISR. a sort of improved multiboot prgr.
I make no use of partitons.
So my first HD is 250 G. I could install maximum of 10 O.S systems.
I have always a basic naked system at hand.
I copy that and install on it the challenged prgr. in this it should be photoshop or elements and of course CM.
So no disturbance of other prgrs and parts of it in the windows sytem directorys.
It is then easier make conclusions or the failor exists or not and so on......

[mikemeister_admin]

Mike,
experimenting, i discovered that when I unlock a skincolorpin in the Barry situation,  I could not move the points on the curves.
( in labmode)

Frits

Mike, when i place a skin-pin on the childs face it is locked. When  I try to move the locked pin on the curve, it could.  not much, but it could)
the intern value of the skinpin changed. When I unlock the skinpin and lock again, I see the value is changed.
Is that possible?  I thaught the values are stable.? ( this all in unmasked mode)

In masked mode it is also possible to change the locked color pin-color, with the up or downward arrow. ( not changing the curves on the outside of course)

[-default]

Yes, actually Curvemeister will allow you to move the point for a locked pin, but it does so by changing the "target" color value associated with the sample point.  The original pin, of course, is not changed.  You can see this value by double clicking the sample palette.  I intended this to allow you to fine tune the effect of a pin manually.  Now that you mention it, it is a bit confusing.

In the case of a hue only pin, the behavior in RGB and CMYK resembles that of a floating neutral, with several curves moving at once to maintain the target hue value.