Black and White Images(techniques)

     Before you start reading it,I just want to say thank you to Sebastian Craciun who spell checked and corrected the blog for me! Send him a big hug on Facebook;)
1. Grayscale mode
    Go to Image/mode/ Grayscale ,usually the result you end up with is not bad,but some more tweaking is always welcome I guess.Before I forget,if you take a look at the info panel(window/info),depending on where my pointer rests on the image it is reading a so called k value in percent,that's because we now only have one channel and it is called gray(window/channels)K stands for keying(as far as I know) and it reads the grayscale values in percentage,from white(0%) to black(100%).

    As I was saying,the result is not that bad,but some local adjustments will make it look better .My opinion is that nondestructive adjustments are the best way to work so I will create a levels adjustment layer(because it is easier to use than curves)

    As you can see in the image,if you don't know that already,you have to click on the circle that is half black half white at the bottom of the layers panel,it may not be located at the exact spot on screen as mine because it depends on how your workspace is set up but it will be at the bottom of the layers panel
and choose levels from the pop-up menu that appears on screen .The levels dialog will appear in your adjustments panel if you have version cs4 or cs5 as I have,or it will pop-up if you use cs3 or earlier .Then all you have to do is play around with the three triangles,the black ,white and gray triangles to adjust the contrast of the image.In the layers panel,that is our levels adjustment layer we just created(where the top arrow points at).You can double click its thumbnail to reopen the levels dialog and readjust settings .that is why it is a nondestructive adjustment,you can reedit setting or just delete the adjustment layer,because the settings don't get dropped directly on your image.
     So far we only made a global adjustment tho I was talking about some local adjustments,so lets get to it.
     1.a. Color Range
     A dialog box will pop up .White areas in the small preview window show parts of the image that are selected,that means will show parts of the image that get selected if you first click on an area of the image,color or in our case shade of gray. Shift+click(and or dragging) on other parts of the image it will add those to the selection and fuzziness controls the range of color or shades of gray that get selected.

      "The marching ants" represent our current active selection,but since I managed to accidentally select a part of the shadow I first want to get that part of the selection out of the way. That is why you see the lasso tool circled with red in the image .I can select that tool by clicking on it(obviously) or by hitting "L" on the keyboard .If you see an other of the nested tools there,the polygon or magnetic lasso tool,you have to hit shift+L a few times until you see the lasso looking thing.Holding down the alt key or option key if you use a Mac machine while you circle around the unwanted area will subtract from the selection,if you hold down shift you can add to selections .One thing that I forgot to say is you should make sure you select the background layer(in my case it is still called the background layer because I did not unlock it,the point is if you have an adjustment layer selected once you choose color range and hit OK it will turn that selection into the adjustment layers maskso make sure you select a normal layer by clicking on it .That is why you see that red arrow pointing to the highlighted background layer. And you should do that before you enter the color range command.Now with the selection in place you can create an other adjustment layer,levels or curves,to darken or lighten selected parts of the image .Since we have no color in our image that's all we can do to make it look better .Because of that selection you will now see that the new adjustment layer has a mask .The white areas in the mask represent the areas where the effect will be visible and black shows where the effect is masked,and it looks the exact way as it did in the Color Range dialog.And you can do that as many times as you need to,going back to Color Range,selecting different shades of gray.

     Just in case you might ask yourself why do we need to darken/lighten parts of an image,well lets say you have a picture with to apples,a red and a green one .If you just desaturate the image(we will be getting to this one soon) there wont be much of a difference between the shades of gray in the red and green apples .If I would want to represent red and green in shades of gray I would choose a darker gray for red and a lighter one for green .That is why we would want to do that .Or to put an accent on some parts of an image.
Before we get to the Desaturate command,and then the channel mixer and the black and white command(which is probably the best way to create B&W images) and probably I will show how to do it in Photoshop Lightroom and Camera Raw,I would like to show two other methods for local adjustments because even if you use the black and white command sometimes you may feel the need to locally adjust some areas to.
And before you ask me why is this post so long,I'm just hoping you will get more out of it this way,or you could just scroll down to the Black and White command.
     1.b. Burning and Dodging
     There are three ways that I know of that you could use to dodge and burn(darken/lighten) images .First of all with the use of the dodge and the burn tool,obviously,but you could do that with a new blank layer set to soft light and then you could paint with black/white to darken/lighten areas.
An other way would be the use of adjustment layers,just as we have seen in the example above in the section about color range,but you could skip the Color Range part and use just a brush for masking.
     1.b.a. The Burn tool and the Dodge tool
     Even if hard to use tools especially on colored images,tho since version cs4 we have the protect tones option,that is inactive if used on a grayscale image,because we have no color in our image,these tools may be just the thing you need .Tho I rarely use it and don't really recommend it.

     First of all you should hit cmd/ctrl+J to create a duplicate of our image,this way if you mess up you only have to delete that layer then duplicate again .The red box on the right shows our new layer named Layer 1.Then select from the toolbox the tool that is marked with red,well in my case it is the dodge tool but it could be the burn tool or the sponge tool because those tools are nested together. You can press "O" or "Shift+O" to circle trough them .By the way the sponge tool is used to saturate or desaturate areas of an image.
Up in the tool options area Range and Exposure are the options you should focus your attention on .Range defines the tones you want to affect,those can be either midtones ,highlights or shadows .Exposure defines the strength of the tool,so most of the time you want to keep that at a low setting then brush over an area more times if you want to lighten it more. That pop-up box on top of the image appears if you right click with the tool on the image .This is where you can adjust the size and the hardness of the brush .You can set these parameters from the tool options too but it is more handy this way in my opinion .There are some shortcuts for these options too,the left bracket key makes the brush smaller and the right one makes it bigger,shift+the left bracket key decreases the hardness of the brush,shift+the right bracket key increases the hardness. In my case hardness should be set to 0.When you dodge and burn it would be a pain to always have to go to the toolbar to change tools,instead of that just hold down the alt or option key while you are brushing .If you have the dodge tool selected it will switch to the burn tool or the opposite way if you have the burn tool selected.
     1.c. A dodge and burn layer

     
     Another way would be by simply creating a new blank layer by hitting the new layer button at the bottom of the layers panel,that will create our new layer tagged Layer 1.At the top of the layers panel set the blend mode to soft light by clicking the down pointing arrow in the area marked with red and choosing soft light from the menu that appears.Next take the Brush tool from the toolbox or by hitting "B" or "shift+B" in case you have an other tool selected from the tools nested with the Pixel Brush. Set its Opacity to a low value in the upper tool options area .Down at the bottom of the toolbox we have the foreground and background color swatches. We want that set to black and white. You can do that either by hitting "d" on your keyboard or by clicking on that smaller foreground/background symbol above it .To swap your foreground and background colors between each other you can click that small double pointing arrow next to it
or by hitting"X" on your keyboard .Why do we need this? if you paint with black on the new layer we created(make sure it is highlighted) you will see the image darken under the area you painted on .hit "x" that will make white your foreground,now you will see the image lighten where you painted over it.
    1.d. Adjustment Layers

   Create a Curves adjustment layer by clicking the adjustment layer button at the bottom of the layers panel. click on the middle of the line in the dialog that appears and drag the point you created on the line to the left or right to darken or lighten the image .Hit cmd/ctrl+I to invert its mask .Now the image reverts back to its original state because the effect is masked away .Those black rectangles next to the adjustment layers thumbnails represent its mask .When you create an adjustment layer its mask will b white,so your adjustment will b visible .Cmd/ctrl+I will invert it.Create another curves adjustment,set a point in the middle of the line,drag it to the opposite direction you dragged the first one .Hit cmd/ctrl+I .Now you only have to take the brush tool and paint with white to reveal the adjustments only on areas of the image you want them to appear .You should be careful at the edges .The beauty of this one is that the adjustments are re-editable so you can double click the thumbnail of each adjustment layer and tweak your settings again to your liking.

NOTE
     You cant use any color on an image that has been converted to Grayscale because it has only one channel that contains the gray value of each pixel .Converting that to RGB or LAB will fix this problem(image/mode/RGB)It will not magically turn your image into a colored image because the color information was lost when we converted to Grayscale but it will allow you to use color.

  
    For example I converted my grayscale image to an RGB(image/mode/RGB),created a new layer with the new layer button an sat its blend-mode to multiply .With a soft brush I painted red over my apple .Now it looks like crap but if you double click on the new layer(not the name not the thumbnail of it,somewhere on this empty space a dialog box pops up.

     At the advanced section of the dialog box we can see to bars with gradients from black to white each with a black and a white slider at its end .We focus now on the lower bar where it says underlying layer .If we drag the white slider to the left we are telling Photoshop we don't want our red color to show up on the lightest parts of the underlying image .At first you will end up with jagged edges but if you hold down alt and start dragging one of the edges of the white slider it will split it in two. the further you drag one side from another the smother the transition will be. The upper bar would make the selected layer's(the one we painted red on) lighter or darker shades disappear depending on witch slider you play with .But in this example it wouldn't do to much because we only have one solid color on our layer.
     This is a good example for colorizing black and white images in a convincing way .Just with a brush you would have a hard time trying to paint on all the details of the apple.

     2.THE DESATURATE COMMAND 
     Go to Image/Adjustments/Desaturate or you can just hit cmd\ctrl+shift+u and you have a desaturated image.You can see two images below with two different results tho I used the same desaturate command on them.I will explain why that happened in just a second.

                                          Desaturate in Lab

                                          Desaturate in RGB

     One of them was desaturated in LAB mode the other one in RGB .While RGB(and CMYK) are device dependent color modes,LAB is an independent mode and it works kind of in the way our eyes work.(At least that is what they say)LAB is also the only mode that has a separate channel just for the lightness of an image while in RGB they are sort of embedded in the three channels(red,green,blue) as we can see in the histograms below .

The histogramm shows the density of pixels in luminosity values from black(0) to white(255).In LAB color mode it only maps the lightness channel because that is the only one that stores the luminosity values of pixels in an image.That is the reason why in RGB mode correcting tones can cause color shifts. You will find yourself lots of time having to perform color corrections after tonal corrections in an RGB image .The differences between the color modes cause the desaturate command to work with different algorithms in different modes,that's why the different results .Lets take a look at the algorithms.
     2.a. DESATURATE IN RGB

   
      In RGB mode I placed a sample point with the eyedropper tool,shortcut "I",and shift clicked on an area of the image,or you can use the color sampler tool(shift+I) to place the point. In the info panel we can see the numeric values of the color sampled .In my example R=190,G=175,B=30 in the colored version while in the desaturated version this same area becomes R=G=B=110.In RGB mode whenever R=G=B it creates a shade of gray,so R=G=B=0 means black and R=G=B=255 means white,all the other possibilities between mean shades of gray.So how does Photoshop end up with this 110 value in my example? It takes the sum of the highest numerical and lowest numerical value and divides it by 2. In my example it would be (R190+B30)/2,so it just ignores the green channel .That is 220/2=110. Then makes all three values in a pixel equal with that number,we would have no gray if the three channel values are not equal.Sometimes it may be just what you need but you have to agree our apple looks like crap in RGB mode if you just desaturate it .Lets see a better way of doing it in RGB mode.
 

     
     As we see in the image above,first create a duplicate by hitting cmd/ctrl+J then hit shift+cmd/ctrl+u to desaturate the duplicate then set its blend mode to Color .Now we have a better result .Then you could use some local adjustments on it if you need to .So the color blend mode tries only to affect the color of the layer below so not the tones .If you want to affect only the tones you would use the blend mode luminosity .This would be an alternative to color corrections after tonal adjustments in RGB .They don't just work as a charm tho so if using luminosity you might find yourself having to saturate the colors in the image .You may have the same problem in LAB tho you only affect the lightness channel when you do tonal adjustments.Keep in mind that you can change the blend mode of an adjustment layer as well .Most of the time I just correct the colors tho without using the luminosity blend mode.
     2.b. DESATURATE IN LAB
     Go to Image\Mode\LAB.Lets place a sample point just like we did before,while the image is still in color and the desaturate it by hitting cmd\ctrl+U.

     As we can see in the image above L=71,a=-5 and b=68.L is the lightness channel and a and b contain the color information .After desaturation L=71,a=0,b=0.So obviously the algorithm is really simple in this case .It just turns the color information to zero and uses just the lightness channel to create our grayscale image. after this you could use some other global and local adjustments as seen in the examples above.
     Now lets see some more advanced tools photoshop has to offer to create black and white images.
   
    3.CHANNEL MIXER

     
     This one only works in RGB mode! You cant use the channel mixer while in LAB mode .Create an adjustment layer by choosing channel mixer this time .There isn't much I can say about the channel mixer,it just does what you may guess because of its name.I don't really use it only to create black and white images in case I am in the mood of using the channel mixer .The way it works is kind of like this,if you check the Monochrome check-box it will create a grayscale image .The Red ,Green and Blue sliders control how much of each channel you want to use to create the black and white image .If you take a look at your channels(window/channels)You will see by default your channels in grayscale by default .In the preferences dialog(cmd\ctrl+K) you may change that to color,then you will see your red channel in shades of red and the green and blue in green respectively blue shades but I don't see no reason you will ever want to do that.
In channels black means that that primary color is missing from those pixels,in the info panel it would show up as R or G or B=0,and white areas show that color at full intensity(255 in the info panel)Channels by their nature contain transparency. so black does mean transparency too. Whenever you create a mask it will create a new channel for it in the channels panel .This is how masks work .So dragging the red slider for example to the right will bring more red in the image,that means in the red channel the lighter areas will get even lighter so with monochrome checked the pixels that contained more red will start to get lighter .Drag it to much to the right all the image will start to get lighter as it start to add red to pixels that had no red originally.
That is why you should watch out that the total of the channels should be 100%.That's why that total is for under the three sliders .Obviously is the total is not 100% the lightness of the image will start to get messy,it will lighten or darken depending on the settings .I sometimes use the constant slider to compensate for that in case I like the image more when the total is not 100%.With monochrome unchecked you will notice that you can choose an output channel,Red,Green or Blue .And it will mix the other channels only into that channel .If you take a look at your channels while working in the channel mixer you will see that the channel chosen in the output is the only one affected. I never used the channel mixer on colored images so I don't know any reason why you would want to use that(Yes i am ashamed to admit that!).In monochrome all channels will get affected because we can only have gray in RGB while R=G=B.

    4.BLACK AND WHITE ADJUSTMENT
    This one only works in RGB mode too! There we have it! The easiest and most efficient way to create cool black and white images.

    Why these adjustments only work in RGB? My guess is that we can only use the color information of an image to manipulate the tones when we convert it to grayscale is if the lightness of the image is somehow nested in the color information as in RGB mode.So create a new adjustment layer,this time a Black and White adjustment layer .This was introduced in version cs3 so don't search for it if you have an earlier version of photoshop .We see a bunch of sliders in the panel .Red,green,blue and the complementary colors of R,G,B cyan,yellow and magenta .Depending on the direction you drag these sliders it will either darken or lighten pixels in the image that fall in the range of that slider.So in my case i darkened the reds and lightened the yellows.If you check the tint box you can ad a tint to your image .There you have one of the ways to create a sepia effect.

    How to do it in Photoshop Lightroom,Camera raw and the free to use soft,the GIMP,I will post that in a few days,so stay tuned if you want to read it!;)