Photoshop: Tutorial: Using the Info Panel to Color Correct by the Numbers

This tutorial will be help the user understand how to color correct using the info panel and curves adjustment.  This is highly useful when you can’t rely on your display for color accuracy.

I’m illustrating this process by choosing an image that is used to portray texture, sharpness and color rendering of a camera.  It has a full range of textures, tones, colors and saturation.

The image is very warm, and will need the color cast removed.  It is effecting the white point, the gray point and the black point.

Our first step is to identify where the eye dropper should set a point sampler.  We can create 4 point samplers at a time.  

First lets examine the info panel:

In order to identify which point on the image could represent a nice clean printable white point, (that is not a spectral highlight) you may use the threshold adjustment layer, or use curves, and drag the white point slider while holding option toward the center of the curves panel.  My illustration must use threshold because I can not take a screenshot of the curves panel preview while holding option.

Drag the slider all the way to the right to see any pixels that have a reading of 255, then start sliding it toward the left until you see pixels appear in mass over an area that you know should read as a clean white.  We will choose the white of the background of the woman.

Use the eye dropper tool (i) and hold shift while clicking on the image.  That will create the first point sampler.

Observe the image zoomed in where I have chosen the white point.

Now Let us choose the black point.

Note the RGB values for our black point.

Now let us pick our gray point.  Normally this can be quite difficult, but for this image our gray point is known to be a perfect gray,

Note that the R and G values are 30 points higher than the B value.

Our next step is to adjust the white point.  
We are somewhat constrained by the white point value of the Red channel.  Note that it reads 234.  We can push it as far as 245 to maintain printable detail, but for illustration sake, lets make the Red White point our baseline, and raise the Green and Blue white points to match.

Lets choose our green channel in curves, and slide the white point over toward the left until the white value is equal to the Red white value.  Please note that the RGB values on the left of the slash show you what the color values WHERE, and to the right of the slash, what the color values are NOW.  We raised the Green white point from 230 to 234.

Now lets repeat the process with the Blue Channel; from 205 up to 234.

Lets repeat the process now, but we will adjust the black point.  We don’t want our black swatch to read as 0,0,0, because it would crunch the blacks down too far, and we would loose detail, so lets aim for R12,G12,B12.

Note that our black values are reading as 12, 12, 12.  Also note that after shifting the black point, we have also altered the Green white point a tad.  We will notice a greater shift after we adjust our gray points.

Move the Blue gray point up to 115

Move Green gray point down to 115.

Note that our Black point and White points shifted a bit.

Please shift the white and black points back to where we aimed them to be before.

Our white points, gray points and black points are all equal now.  The image is a tad dim, and could use with a white point pop, so lets push our white point up to 244.

Lets zoom in and examine the actual point sampled spots.

And finally, the Before and After.

PHOTOSHOP: TUTORIAL: Contrast Masking and Calculations Function

As we have discussed in the past, the channels panel is composed of very use information about that one’s image that can be utilized for some very powerful adjustments and alterations.

http://housetribeca.tumblr.com/post/82410168570/photoshop-tutorial-explaining-photoshop-channels
and 
http://housetribeca.tumblr.com/post/82476222159/photoshop-tutorial-selections-alpha-channels-and
One of my personal favorite uses of the channels panel is to isolate an element in an image based upon the color and or tonal contrast that it has compared to other elements.
For this example I want to isolate the blue dress and create a mask.  My Goal will be to change the color, or hue of the dress. 
I want to examine the Red Green and Blue channel that make up the image, and look for the channel or channels that provide the most contrast between the background, the dress and the skin.
image

image

image

image

You will notice that the Red Channel provides the most contrast between the background and the dress and the majority of the skin tones.  The only area is problematic is the shadow area above her left elbow.  The blue Channel is a better match for isolating that contrast difference.

What you should first do is create an alpha channel that is derived from the Red Channel, hold down the command key and click the image icon on the Red Channel.  You will now have marching ants active on the Red Channel.  Click the “New Channel” icon next to the trash can icon on the Channels Panel.  This will create a channel called Alpha 1.  You may rename it, but I will keep it as Alpha 1 for this demonstration.

image

image

Click on Alpha 1 and note that it appears as the inverse of the Red Channel.  Our goal is the make the dress solid white, and everything else solid black.

image

You can increase the contrast of the channel by using the adjustment levels.

image

Pull both the black point and white point toward the center.  Note that the contrast under the arm is too strong and starts to blend in with the dress.

image

This is when we will use the Blue Channel to create distinction.
I will now introduce the Calculations menu.  It will provide you with the ability to combine or subtract elements from one channel, to or from another to create a brand new channel. 

image

We will take the Red Channel and Divide the Blue Channel from it, and end up with a new channel that is a rather nice isolation of the skin tones.

image

That new Channel, “Alpha 2” should be inverted, and the the contrast increased so that it can be used to create a selection, which will then be filled with black.  This act will remove the skin tone from Alpha 1 resulting in a much closer isolation of the dress.

image

image

image

image

image

image

image

We will now need to clean up the selection by using a brush set to Overlay with black as the color.  All tone that is not intensely white will become black.  Then switch your brush blending mode to normal and brush away some of the remaining white pixels that appear outside the dress silo being careful not to ruin the edge.  Switch your color to White, and blend mode to Overlay again, and proceed to clean up the inside of the dress.

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

Once complete let us load the selection of Alpha 1.

image

image

image

Then Click on Channel “RGB” so that we can see our image in standard color mode.  Click on your layers panel and choose the adjustment layer, Hue / Saturation.  Slide the hue to the left and that saturation a bit to the right. 

image

image

image

image

image

image

image

You have now changed the color of the dress.

PHOTOSHOP: TUTORIAL: Selections, Alpha Channels, and Transparency

Alpha Channels and Saving Selections:

An alpha channel is 2D information that distinguishes points of transparency based upon the pixel dimensions of a 2D image.  An alpha channel is used in practical purposes to accomplish a composite of either color, tonal or pixel texture.

The real power of photoshop manifests when the use of layers is incorporated.  What a layer allows the user to do, is to stack multiple images and or adjustments while using different degrees of transparency to create an overall effect or to a specific area which could not be accomplished when altering the image as a whole.  Even a basic lasso selection incorporates the concept of transparency because it distinguishes where an alteration will take place, and where the borders of that alteration stop.

To further understand an alpha channel, let us create a lasso selection and then save the selection (call it squiggle or circle or box, etc).  When you save a selection, it notifies you that you will be creating a new channel (please read about channels in my post here: http://housetribeca.tumblr.com/post/82410168570/photoshop-tutorial-explaining-photoshop-channels ).  Examine your Channel’s panel and notice the new channel which appears below the Blue channel.  That new Channel represents the exact shape that your selection created.  If you used any auto feathering you will notice that the shape has soft edges.

The shape that you created and then saved is now called an alpha channel (make sure you De-Select your selection before doing anything else to your image).  That alpha channel translates selection information to the viewer by displaying pixels which are 100% selected in pure white, pixels which are 100% NOT selected in pure black, and pixels are partial transparency in shades of gray.  If your lasso selection had auto feather set, you will have shades of gray around the edges of your shape rather than a hard black/white edge.

This selection information, or alpha channel can then be utilized by command clicking the black and white rectangle icon of the channel.  This selection can also be “loaded” by going to Select>Load Selection, and choosing the name of the alpha channel from the drop down list.
So in short, an Alpha Channel distinguishes transparency, and in practice is usually associated with selection information.

White = 100% selected
Black = 0% selected
Grays = 1% - 99% selected



PHOTOSHOP: Tutorial: Explaining Photoshop Channels

image

One can get into a very detailed conversation about Photoshop Channels, but I’m going to describe them as briefly as possible.



Photoshop Channels are composed of black and white information that determine how Photoshop will present a color image to a viewer through a color display.



     The simplest channel image would be a black and white, or gray-scale image.  It is composed of the brightest point of the image reading as pure white, the darkest point of the image reading as pure black, and everything in-between reading as some range of gray.  If we think of a gray-scale image in terms of drawing, one would start with a pure white piece of paper, and add tone to an area till that area reaches the darkest point that one’s particular media can achieve.  In this particular case, white would represent the absence of tone, while black would represent maximum tone, i.e. charcoal, pencil, paint, etc.  This particular example is only useful if we continue the explanation in terms of tone and pigment, but I will be save that for a later date.  



     Let us then transfer the concept of a gray-scale image to that of light, where black represents the absence of light, while white represents pure light at maximum brightness.  We need to refer to this example in terms of light because we are looking at a photoshop image on a light based display; meaning, the display produces a white light behind a panel of pixels.  The color display is composed of pixels and each pixel is composed of sub pixels that produce either Red, Green, or Blue color at 256 brightness levels.  Since each color has the ability to display its self any where from 0 brightness to 255 brightness, that means an individual pixel can be any variation of just under 16.8 million colors (256x256x256 = 16.777 Million). The display attempts to produce an neutral tone by having each pixel produce equal levels of brightness of each Red Green and Blue sub pixel.  This is the starting point for understanding what Channels are.  



     A color image which is displayed by Photoshop is composed of three channels, Red, Green and Blue; all displaying some level of their color from 0-255.  Let us now examine the color Red.  We would see a red pixel if that pixel were only to show red light, while simultaneously the Green sub pixel and Blue sub pixel produce no light.
Red = R 255, G 0, B 0
The color Cyan would then be the absences of Red and the presence of equal parts Green and Blue.

Cyan = R 0, G 255, B 255
The color Green would then be the presence of Green and the absences of Red and Blue

Green = R 0, G 255, B 0
so on and so forth, (please examine the attached images)



Now that being said, we understand that Red is the presences of maximum Red and the absence of Green and Blue.



This can then be presented to the photoshop user with the concept of the black/white brightness scale.  Every pixel that reads as pure Red would appear is solid white.  Any pixel that had absolutely no Red present, or appears as Cyan, would be presented as solid black.  If a pixel is not pure red, it must be any tone other than white, which then gives us a visual translation that photoshop presents us that appears as a black and white version of our particular image.  (Please examine the examples of the color dresses below.)
So, in short, a color image is composed of a Red, a Green and a Blue channel.

Each particular channel tells the viewer how intense that particular color of a particular pixel is by whether that pixel is white. 
For a display to produce middle gray, the combination of RGB needs to be 128,128,128.  Any equal parts brighter or darker, will produce a neutral gray that is brighter or darker.

An RGB image can also store an Alpha Channel which is a channel that distinguishes transparency.  I will provide a future entry to explain Alpha Channels.
One can easily create an alpha channel by creating a selection then choosing Select>Save Selection.  The saved selection will appear in the channel just below Blue. (See illustration below).

One can then activate the alpha channel or selection information by command clicking the new channel.  Notice the marching ants reappear on your image.
TIP!!!  You can also command click a color channel and select varying degrees of opacity and transparency.  Lets say you command click the Red channel on image that has a woman wearing a bright red dress.  You will then have a rough selection of the Red dress, and a partial selection of anything else that has degrees of red in it.  Pure Cyan would absent from the selection. 

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image

image


PHOTOSHOP: Tutorial: Explaining Photoshop Channels

(all photos found online without attribution)

Photoshop: Healing Tools vs Clone Stamp vs Burn Dodge

image

As one becomes more experienced in the use of Photoshop for practical retouching one will realize the need and the advantage of one tool over another.

When I first introduce my students to the spot healing brush they think of it like its God’s finger coming down and magically fixing all of the problems they have on an image.  It is not till the student learns the disadvantage of the spot healing brush and its club-like use as a tool that they begin to want more control and precision of something like the clone stamp.  Lets explore the pros and cons of the different retouching tools.
The Spot-Healing Brush: brush based tool that will sample (do not use Option-Click, just click) pixels within proximity to the spot you wish to remove, and is best used for small dots, dust and micro blemishes.   
     
     Advantage: quick and simple, very easy to use when cleaning a scan or when getting rid of tiny black heads and white heads.
     Disadvantage: can not be used on large areas because with will mush the texture together and create an odd dappled effect across your image.  Can not be used near points of high contrast because it will bleed the texture together.
     Tip: the size of the brush tip should be only slightly larger than the spot you are removing.  Too large a brush tip and you will damage essential texture.

The Healing Brush: brush based tool that allows the user to choose specifically (by Option-Clicking) which texture is sampled and blended into the spot that you wish to remove.  Works much like the clone stamp, but will blend pixel texture with current luminosity.
     
     Advantage: allows you to choose a specific area that you wish to sample from for more control over blending.  Can be used for a larger area of the image than spot-healing brush.
     Disadvantage: Should not be used for very large areas because texture will take on an inconstant look and not flow well.  Can not be used at points of high contrast, can not preserve sharp edges.
The Patch Tool: lasso based tool allows one to select a large area of texture and either drag that selection from one place to another to either relocate the texture to that spot or to use that spot as the source of the texture for the selection.  The tool will then blend the texture with the edges of the selection and keep the luminosity of that area selected.
     
     Advantage: can be used to blend texture of very large areas, helps maintain large areas of pixel texture.
     Disadvantage: must be used upon an island of texture and tone, and can not be used upon an area of multiple tonal ranges that intersect with the marching ants of your selection, i.e. you can patch-tool a tree that lays on a blue sky, only if you first remove the bottom of the tree where the sky merges with the ground.  If you do not, the color and tone will bleed from the ground into the sky.
     Tip: before deselecting, use the quick keys Command-Shift-F to bring up the Fade option so that you can fade the blend to something that may be more pleasing.

The Clone Stamp:  brush based tool that allows one to sample (using Option-Click) a specific area of pixels.  The sample can be used for a 100% 1-to-1 copy of pixels that will fade onto the spot where the brush is clicked based upon the softness or hardness of the brush edges.  The tool can also be controlled by the opacity of the brush or the blending mode of the brush.
     
     Advantage: is used for making very specific and controlled retouches.  Greatest advantage is that the user can use it to copy/clone edge texture and maintain a sharp edge and or edge contrast.  Also that it can be used to blend and smooth texture by controlling the blending mode or opacity.
     Disadvantage: takes more user control and one can easily damage texture and pixels with this tool.
     Tip: try using a low opacity and soft brush to get used to using the tool
Dodge / Burn: brush based tool that allows the user to brighten or darken the tone of pixels with out destroying pixel texture.  You can size down the brush to effect a specific pixel or small group of pixels, and adjust them here and there over the expanse of the image to have an impactful, yet subtle change of the image.
     
     Advantage: can be used to effect subtle change and maintain pixel texture
     Disadvantage: can effect color and saturation when used heavily in a specific area.  To use effectively the tool must be used at very low exposures (1-8% exposure) so not to damage the image.  The result is that it is a very slow tool to use.
     Tip: (see my previous blog entry on Burn and Dodge - http://housetribeca.tumblr.com/post/77174506688/photoshop-tool-burn-dodge

- Jeremiah Dart