Making Layer Masks Using The Curvemeister Plug-in:

Masks are a very versatile tool provided to users of Photoshop and Photoshop Elements; they provide a means of protecting parts of an image from being changed by adjustments to the entire image. The number of potential uses is only limited by your imagination; if you can imagine something as a black and white cut out then you probably can make a mask of it.

Do you remember the “body books” from grade school?  These books allowed you to peel off parts of a body so you could see what is hidden beneath.  If you wanted to see the muscles you peeled back the skin layer which was just painted onto a clear piece of plastic.  When you moved the skin layer the muscles below were there for you to see.

This concept explains layers as well as masks.  Layers are the clear plastic and the painted on surface was the mask hiding the information held on the layer below.  The mask could be opaque as in the example of the skin or it could be slightly transparent allowing you to see into the layer below without completely blocking.

Curvemeister provides the means to create a mask from the color channels of an image being edited in Photoshop (PS or PSE).  You can use the mask for your current session of Curvemeister or you can copy the mask to the clipboard using the copy Icon and then paste it to a Layer mask in PS or PSE.

Once you open an image in Curvemeister you need to activate the Mask Pallet in Curvemeister (CM) by clicking on the mask cart icon.

This will open the Mask Cart and display all of the channels that you can select a mask from.

So, why ten choices; I just need a mask from the image right??

Short answer is yes; you just need a mask but it is always much better to do more with less especially in Photoshop. If I have a yellow flower on a green background how will I mask out the green so I am just adjusting the yellow?  Sounds easy but if I just look at each channel there are some very distinct things going on that might just help me. Below is the image I am going to make masks from.

The Red Channel:
The Green Channel:

The Blue Channel:

So do we have anything we can use as a mask?

Not really, or should I say not yet?
The Red and Green channels have some potential that we’ll need to look at. Let’s curve those channels first to see if there is enough detail to make a decision.
First the Red Channel:

If I curve the channel so the curve line is vertical; then I get can create a high contrast mask, but as you can see there are areas of the image that are not yellow that also make it into the open areas of the mask.  If I adjust the color based on this mask I will get results that I do not necessarily want; as seen below.  Notice the yellow blotches in the middle of the Green areas?  These are areas that are not supposed to be adjusted.  Obviously I need a better mask.

How about the green channel?  Again, areas of yellow are appearing in the background and some parts of the stems and buds that should not be yellow are getting too much of that color when this channel is used as a mask.

It appears that neither Red nor Green will serve as a good mask without some additional adjusting and painting in Photoshop.  Let?s look deeper.

An inverted Hue mask like the one shown above might work; but it is a very coarse mask and we would have problems with the image looking spotty; the edge detail of the flowers is severely lacking.  Again, this is not a good choice.
The A channel in Lab might be a starting point after all it has green in it?  Unfortunately, there is again too little detail in the Channel to make an effective mask.

How about the B channel?  When a vertical curve line is applied as before the channel seems to provide a clean, sharp mask that we can use quite well as shown below.

As you can see from the example above masking with the right channel can make fast work of Photoshop adjustments.  Choosing the wrong channel can lead to frustration and poor results?

How do you know which channel will give you the best mask?  This is the area of masking that is the hardest to describe.  Most people try to do this using trial and error. The real answer is that you need to use your color theory knowledge to help you make fast choices that narrow down the possibilities.

 

First things first; what are you trying to do?

Are you enhancing a single color?
Are you working on shadows or highlights?
Are you trying to limit saturation in specific areas of the image?
Are you keeping skin tones protected?
Are you enhancing the sky?

Your choices for a mask are based on the answer to one of those types of questions.  Figure those out and you have a good start on what channel to use to make your mask.


Are you enhancing a single color?

Think color opposites.  Magenta/Green, Red/Cyan, Blue/Yellow are all color opposites.  What about channel opposites?  In LAB Red and Green are on the same channel so an Image that ahs lots of reds and greens might be better masked in the A channel or as in our example the Yellows from the B channel make a excellent mask for this image?


Are you working on Shadows and Highlights?

Think in Black and White.  The two channels where we have only black and white information not tied to color are the L channel in LAB and the K channel.  Both are strong candidates for a mask but if you are dealing with shadows and highlights the K channel might be a bit weaker since it is calculated from the color values rather than a representation of the tonality of the image.


Are you trying to limit saturation in specific areas of the image?

This is a more difficult area to discuss.  The accepted practice currently is to use some type of process to boost color and create color contrast by differentiating the colors using LAB.  The trouble with this process is that you also increase the saturation at the same time. You then have to limit the effect of the adjustment to keep the image looking normal.  Garish greens, over blown reds, unreal blues are all the result of the color contrast enhancement.  The Saturation channel is available to you for masking but it is very coarse and has to be heavily blurred in order to have any use.  This makes it difficult to use for fine details and it is generally not done.  There are some L channel adjustments that you can try to help you control the saturation of your image.  Remember that you might have to make a curve that looks very wild and out of place but if the mask is correct you can have some of your cake while eating it too.


Are you keeping skin tones protected?

You can use the skin mask to help you a great deal with this problem.  Make sure you have a reasonable correction for the skin first and that you adjust the mask as tight as possible to the skin tones.  You risk looking cut out if you are too aggressive.


Are you enhancing the sky?

The K and L channels make excellent Sky masks.  The difference between clouds and sky is primarily a brightness issue and brightness lives in those channels.


Using Blur:

Curvemeister has a Blur Slider built into the interface.  Adjust the slider to help blend the mask into your image so it looks less cut out and more natural. In general use; I find myself using the blur setting and later applying the mask to a layer if I feel I need to blur the results further.