Adobe Photoshop 3: A Guided Tour >>>
This summary is based as closely as possible on the topics covered in my training course and the order I would have discussed them. Left hand margin entries relate to the code numbers of sample images which can be used in conjunction with the notes. These images are all to be found on the Photoshop installation CD in the stock photography folder.
These notes were written with Macintosh users in mind - references to the "option" key refer to the "alt" key and references to the "command" key refer to the "control" key under Windows systems. Photoshop versions 4 and upwards are not covered in this tutorial, but they use slightly different keys which are outlined elsewhere on this site.
We start with some basics.
The basic format of Photoshop is raster based. Unlike Postscript and other vector-based programs we are dealing solely with a large map of pixels, not objects as such. We refer to this as a canvas.
The canvas in Photoshop will be a fixed size - usually the dimensions with which the file will be printed or otherwise output.If this area is then created with a high density of pixels it is known as a high resolution image. A lesser amount of pixels would be low resolution. For example, the most common resolution for printing is around 300 dpi, based on a file with physical dimensions the same as those to be printed. This means that the individual pixels on screen are smaller then the dots on a printer's screen. To calculate this, we would normally double the screen frequency used by our printer and then round up the figure to the nearest 100.
Working in Photoshop we can use either an empty canvas set up to our requirements (eg. for painting) or we can import a ready-made base image (eg. a video grab, a cd library image or a scan). Once this is on screen we can paint or apply effects to the entire image or we can choose to mask off a selected area in order to work on that alone. It is the ways of doing this that we will be looking at first.
The main toolbox will open on screen automatically and this is where we start work. The top group of four tools are those with which we make selections and they are frequently used in conjunction with each other to provide complex shapes.
The first tool (top left) allows us to draw rectangles and eclipses. When we create a shape we will be effectively covering the rest of the image in masking tape, thus allowing us to work on it in isolation. If we need a less geometric shape, we can draw with the lasso tool (top right) as if we were holding a pencil. If we fail to complete this shape, the end point will draw itself back to our start point.
Having made a shape, the cursor will turn into an arrow when we are inside it. We must be careful not to move it, since this will leave an empty space behind. On the other hand, we will lose the shape entirely if we accidentally click away from it.
In order than we can do further changes to our shape we make use of two essential keyboard buttons. Throughout Photoshop, SHIFT acts as an additive key whilst COMMAND acts as a subtractive one. In the case of selections we hold SHIFT to add to the basic shape and we hold COMMAND to subtract bits from it. We can happily switch between rectangles, eclipses and lasso shapes as w do this.
Another useful variant is the use of COMMAND and SHIFT together. This makes a new selection containing any already selected areas that intersect the new one. COMMAND is still behaving as a subtractive tool, but it is the areas outside our drag that become de-selected as a result.
The third and most powerful selection tool in the toolbox is known as the magic wand and can be found in the second line down to the left. This allows us to click on an existing part of the image and select that area based on its colour values. This too can be used with the keys in conjunction with the other selection tools and we can extend or reduce its pick-up range by increasing or decreasing its tolerance using the options palette.
To activate the options palette we must go to the window pull-down menu, then palettes - show options. To shortcut this procedure, simply double click on the tools icon in the toolbox. If you type 64 for example into this tolerance box, the tool will pick up a range of colour 32 shades lighter and 32 shades darker than the point you click on.
This tool can often be used to pick up backgrounds quite easily where an exact cut-out shape might be awkward. Having selected a background partially, we can also use the select pull-down menu to modify the shape with greater accuracy.
The magic wand tool can also be selected whilst working with the rectangle/eclipse tools. We simply hold down the CONTROL key on the keyboard and the icon will change.
With a partial backgound area slected, we go the select - grow. This will read our current tolerance value and grow the selection outwards by the same amount again. We can continue this until we reach the boundary we want. If are increments of jump become too great, we can reduce our tolerance setting at any point in this process.
If our background contains several major colour groups (eg. sky, but with cloud areas) we may be better off holding SHIFT and clicking on each distinct area in turn. We are then able to grow them as a group until they join into each other.
We may have to hold COMMAND or SHIFT to subtract and add odd details around the edges, but we are able to select a background with relative ease. We then go to select - inverse. This will cancel our selection but will make a new selection from everything previously unselected. This gives us an accurate cut-out from our background.
Another pull-down command is select - similar. This will scan the image and find all areas that include the colour range/ranges of anything already selected. An example might be yellow flowers in a green field. We can use the magic wand to select the first, then use this command to find and select the others.
We should also remember that the human eye often sees things by association. Photoshop does not. Using the above example, it may infact be necessary to select a bright yellow flower and then SHIFT select a dark yellow flower, in order that select - similar actually finds the whole range of yellows we want.
If we click away from any active selection it will de-select. We are now able to create a new rectangle selection. If we go select - modify - border we will get a dialogue box. This will create an entirely new selection for us around the border of our present selection. The width value we type into this field will be applied as an equal spread from our current outline's position. We OK this. We now see two flickering guidelines and the new selection is the area between them.
We de-select this and make yet another selection rectangle. This time we choose select - modify - expand. This allows us to grow our selection outwards by an exact number of pixels. (Unlike the grow command which is based on current selection and tolerance values.) In a similar fashion, select - modify - contract allows us to shrink the selection by a precise number of pixels. Both of these are useful for determining our exact selection boundary before applying the feather effect which will be discussed later.
We now need to use the lasso tool to make an irregular selection shape. We then go select - modify - smooth and enter several pixels in the dialogue box. When we OK this Photoshop will smooth the outline of this shape for us by redrawing any abrupt edges within the radius we specified. We can also try this again on an even more irregular selection obtained with the magic wand tool.
Also available is select - colour range. This extends the idea of the magic wand tool, allowing us to select multiple colour ranges at one time. Going here gives us a dialogue box where we would initially drag the fuzziness slider down to zero. We can then choose a colour range from the menu in the Select box. Alternatively, we can choose sampled colours and use the eyedropper icon to click in the image on the colour range we want. We can then add more colours by clicking with the eyedropper+ tool, or remove unwanted ones with the eyedropper- tool. Increasing the fuzziness with the slider will then extend the spread of these colours in a similar fashion to the grow command.
We can preview the results of this by choosing quick mask where it says Selection preview. The red tinted areas will be masked off - the clear areas being our selection. We can also see a black and white negative of this mask under the fuzziness slider. When we click OK, Photoshop will turn this mask into a normal selection outline, at which point we can modify its shape just as with the previous selections.
There is also a load/save function in this dialogue box. Like many things in Photoshop, we can save the chosen settings for future use - important if we want to match the process later.
We will discuss some additional aspects to making selections later, but these tools and functions alone allow for great diversity.
There is also a major distinction between floating and non-floating selections. As yet, all our selections have been non-floating - simply representations of our masks. If we are cutting and pasting component between images, or merely duplicating an area of a single image, we will need to float the selection in order to position it. Pasted-in selections will automatically float above the canvas' surface until we deselect them. For selections we have made on the base image however, we need to go select - float in order to put them on a plane above the surface. When we do this we also leave a duplicate behind underneath. A shortcut for this is to hold down option-key whilst moving a selection which gives us a floating copy.
When a selection is floating, Photoshop provides a host of effects which are totally dependent on having an underlying surface to inter-relate with. We can change the opacity of a floating selection, dictate which of its component values are actually pasted down or force various underlying attributes become visible through the floating selection itself.
Some of these effect are now only possible using the layers palette, but others will work with any floating selection. An example of this might be the previous cut-out repositioned by option-key move to a new location.
Select - matting - defringe will now allow us to delete a given number of pixels around this selection in order to remove any ghosting effect.
Select - matting - remove black matte will remove any elements of pure black in the selection. Select - matting - remove white matte will remove any elements of pure white in the selection. These are of particular use when we have prepared cut-out and other montage items using a pure black or white background in a temporary file.
Photoshop allows us to make quick work of generating a temporary file since the default values for a new file are the same as anything currently on the clipboard. Hence we can go COMMAND C (to copy), COMMANDN(for new file), RETURN key (for OK) and COMMANDV (to paste). When we finish making changes we simply go COMMANDC (to copy out again), click back on our original file and COMMANDV (to paste back in). Alternatively we can use COMMANDX to cut the original selection rather than copy it. Both options send the selection to the clipboard.
Window - palettes - show layers will give us access to a slider which controls the selection's opacity. The sub-menu to the left of this slider also gives us a library of effects modes which are discussed on the course and summarised on an accompanying display chart. The default should read normal. The name Floating Selection also appears on the layers list in the palette since it is a temporary layer by nature.
Any floating selection can be made into an actual layer. We do this by clicking in the layers palette on the name floating selection and then dragging it down to the new layer icon (on left of wastebin icon). We then name the layer and choose an opacity/mode for it. Once again, the default is 100%/normal. This process can also be done by going to make layer in the palette's sub-menu.
(A new layer can also be made from scratch by double clicking on the new layer icon or going to new layer in the sub-menu. This would however be empty.)
If you have a selection on the clipboard, it is also possible to go edit - paste layer which will generate a new layer to contain that selection automatically.
Layers can be deleted by clicking on the name and dragging them to the wastebin icon in the palette. They can be permanently pasted into the main image through the palette sub-menu.
Layers are transparent, just like sheets of acetate. The default, selections put on layers are opaque although many of the effects modes will replicate transparent inks or combinations of opaque and transparent ink.
The opacity slider and mode menus will affect the whole layer once it is made. They will affect a selection only whilst it is still floating. Previous versions of Photoshop contained an edit - composite controls (paste controls) facility. This has been removed but most of its functions are available as layer options in the layers sub-menu.
Any selection on the clipboard can also be pasted inside a currently active selection (floating or otherwise) by edit paste-into. In previous versions this masked selection could then also be adjusted using composite controls. This is no longer possible although the effect can be replicated by saving the destination selection and later applying it to a layer mask, in order to use layer options. A quicker variation on this will be discussed shortly.
Each layer also shows a preserve transparency box. When on, this restricts our working the the exact areas containing our selection(s). When off, it is possible to paint and otherwise add new information to the clear areas of the acetate.
We must click on a layer in the layers palette to go to it and at that point we are only able to work on that layer itself. Some of the tools we will be discussing later to however offer a sample merged facility which means they can be used on a layer whilst simultaneously affecting everything underneath that layer.
The order of layers can be re-arranged by dragging them individually up and down the palette's list. Dragging the cursor over any two or more adjacent layers will effectively 'group' them together. Grouped layers will display a '+' symbol and the entire group can then be moved together. Layers can also be linked using the palette's sub-menu in such a way that the linked layers are controlled by the mode, opacity and other attributes applied to the lowermost layer. Our base image will always be at the bottom of the palette list.
Any layer can also be assigned a layer mask through the layer palette's sub-menu. If we than click on the layer mask icon (to the right of the layer itself in the palette) we can then use painting tools in greyscale to create a new customised shape through which that layer will be applied to the underlying image. Loading and filling a previously saved selection on this layer mask will also replicate the process of paste-into with composite controls used in previous versions of Photoshop.
If we have an item on the pasteboard with a selection active and go Edit - Paste Into, a new floating selection appears, but it remains masked by the shape of the previous one. When we then drag 'floating selection' to the new layer icon, the mask is removed from the selection on the layer itself, but a new layer mask is created to represent it automatically.
The edit - paste behind facility from previous versions of Photoshop is no longer available, but the effect can be achieved by selection - inverse followined by edit - paste into.
Any effects obtained when creating a layer mask will be visible on screen as we do them. When the results are right and we have the desired effect we go to apply layer mask in the layers palette sub-menu to make the operation permanent or cancel if necessary.
It will be possible to understand this more fully after we have looked at channels. Channels are where we can save and modify our selections masks, even introducing tonal values into our outlines where needed. To do this however, we must use our painting and effects brushes, so we must examine those first.
Next, we should look at the four tools on the third and fourth lines down in the toolbox. These are the hand (grabber) tool, the magnifier, the cropping tool and the text (type) tool.
The second, the magnifier, is used to enlarge the image on screen so the we can work with accuracy on detailed areas. A click with this tool will zoom in by two increments on Photoshop's scale. An OPTION (key) click will zoom out by the same amount. It is however a lot easier to simply use the magnifier tool to draw a rectangle over the desired area, since Photoshop will then calculate the magnification needed.
When we are working in the greatest depth, it is commonly referred to as "pixel edit" mode. This is when the ratio displayed (by the file's name) reads 16:1 and the eraser tool has also become the size of an individual pixel. Here and in other levels of magnification we use the hand tool to pull the canvas around the screen, avoiding the need to zoom out and in again too frequently. The other function of the hand tool is achieved by double clicking on its icon in the toolbox. This tells Photoshop to zoom out to the nearest ratio that will display the whole image on screen.
Double clicking on the magnifier itself will restore the screen to display the ratio known as 1:1. This merely represents the mid-point on Photoshop's sliding scale and has no consistent value due to variations in resolution and file dimensions. With high resolution images it will be a magnification value because there are more pixels in the canvas. With a low resolution image it could merely be a "thumbnail".
The cropping tool allows us to draw a rectangular shape over a portion of an image in order to cut everything else away. The crossed-circle icon allows us to cancel whilst the scissors icon allows is to confirm the operation. This is the first time we see Photoshop's preview feature, which also allows for fine-tuning with any of the individual corner handles before completing the operation. In later instances, the scissors will be replaced with a gavel icon.
Double clicking on the cropping tool brings forward the options palette again. Here we can choose to crop our image to an exact size and shape. This option also allows us to redefine the resolution of the cropped file, but bear in mind that any increase will merely provide extra blending pixels - not actual detail. If you need the current canvas values for reference, click the front image box.
The next tool is the type tool, represented by the letter T icon. We must remember that Photoshop is not a postscript program and that type, like all other selections, will use a technique called anti-aliasing to generate smooth edges. This blurs the character outlines slightly in order that they blend into the composite image.
When we click in the image with the type tool a text dialogue box appears. We choose a font and size from the pull-down menus, then click on the style boxes for additional effects. You would only take anti-aliasing off for very small point sizes and the result will be very hard edges.
If you are entering more than one line of text you should use the RETURN key to specify each line, since there is no text-flow facility. You can also enter a value for leading and spacing before choosing an alignment option button. To see how the text will look before putting it onto the image, click the font and size buttons at the bottom.
When you OK this box, the text will appear on the canvas with each character selected and filled with the foreground colour. These now count as selections like any other except that they are floating. This means we can alter the transparency and mode like before, or convert them into a layer for more advanced editing. We can also use the characters with edit - paste into to put a clipboard image inside their outlines, fill each individually using the paintbucket or fill as a whole using edit - fill. More on these options later.
Before looking at the painting tools which follow next we should jump ahead to the bottom of the toolbox where there is a large square box containing colour. (The default is actually black) This shows what will know as the foreground colour. Offset and behind this is another box. This displays what we know as the background colour. The foreground colour is the colour used just now by the type tool. It is also the colour used by most of the painting tools that will will cover next. The background colour is the colour used by the eraser tool and is also the secondary colour that will be used by the blending tool.
Photoshop does not recognise any real background colour. Every component of an image, even the waste edges on a scan for example, are read as pixel values on the canvas. When we accidently moved a non-floating selection earlier we were left with a hole. The colour of that hole was totally dependant on what colour we had in the background colour box at that point in time.
If we double click on the foreground colour box a dialogue box appears which allows us to mix a colour. If we do this by eye we must first choose a general hue from the long narrow bar. We can then pick an exact shade from the big box to the left.
Some colours that can be displayed on screen and in video/film are not possible to mix with process colour for printing. If we select one of these colours a small triangle with an exclamation mark will appear slightly left of the OK button. Beneath that will be a tiny box. If we click on this tiny box, Photoshop will automatically take us to the nearest printable mix.
If we choose an unprintable colour it does not matter for actual printing purposes. Photoshop, or any destination program, will still convert the values before generating separations. This cautionary exercise is simply for the benefit of seeing the best replication of printed colours on screen. We will be looking at more exact calibration procedures later.
Changing from the H button to the S button gives us a colour bar based on saturation. We then pick a hue from the big box. Changing to the B button gives us a bar based on brightness. We can then choose hue and saturation from the big box. You can also mix from Red, Green and Blue values. The two small patches at the top right of the bar also show your previous and current colour. This is useful if you are trying to simply mix a variant on a previous colour.
If we click on the "Custom" button we can access a number of colour libraries, including the popular Pantone sets. We choose the book required from the pull-down menu and then scroll down the bar for the approximate range. Then we click on the exact colour. Alternatively, we can simply key in the number we want if this is known.
With the exception of duotone images, it is not possible to export Pantone and other library colours from Photoshop. Instead, Photoshop works out the most appropriate colour values for the operating mode we are in. To the right of the bar Photoshop displays CMYK figures which represent the nearest mix using process colours.
When we are happy with our colour we usually click the OK button. Alternatively we can click on the picker button to return to the previous mixer. Here we will see our library colour converted and located in the mixer palette. We can make further adjustments at this stage by using the value fields to the right hand side and type in a specific mix using HSB, RGB, Lab or CMYK values. More on what these represent later.
We click OK and the final colour now appears in the toolbox as our foreground colour. We then repeat the procedure to mix a background colour by double clicking on the offset and behind box.
This brings us to our painting tools, starting with the paintbucket itself, fifth down on the left in the toolbox. We double click on this to bring forward the options palette, where we can choose an opacity and mode for the paint. This tool will fill the area we click on with our foreground colour and we decide on how extensively this spreads by keying in a tolerance value. This is almost identical in nature to the behaviour of the magic wand tool except that we are applying paint rather than selecting. (We may even be applying the bucket to a previously made selection that we created with the wand tool. In this case, the tolerance values should infact match.)
As mentioned previously, if we are on a layer paint can only be applied to that layer itself. Checking the sample merged box in the options palette will however affect the whole image. If we want hard edges we can take the anti-aliasing off, whilst we can change the contents pull-down menu to fill with a pattern.
To fill with a pattern we must first make one. We do this by dragging a selection rectangle over an area in an image then go the edit - define pattern. This will load that part of the image into a buffer memory where it will be available until we over-ride it with another pattern. More on this later.
It is not a particularly good idea to use COMMAND - S or otherwise save files in Photoshop until we are wholly finished with them. We can however choose edit - take snapshot at any point. This will also make a buffer memory file, un-named, which we can obtain when needed. Although a pattern will stay in memory across files (until we quit Photoshop), a snapshot will remain locked to a single active file only and it too will be overidden by any subsequent snapshot we may take later.
If we make a selection active there is another way we can fill an area or, indeed, the whole image if we select - all. This is through edit - fill where a number of extra options are available. In the dialogue box that appears, the use pull-down menu now includes additional choices of background colour, the last-saved version of our file and any snapshot version we have made. This is ideal for restoring parts of an image in isolation from the main image. We can also fill with black, white or 50% grey - particularly useful for masks when we are working on channels. Once again, we have opacity and mode settings, plus the option of keeping our transparent areas intact if we are working on a layer.
Note: Do remember that a layer itself may also have an opacity setting. If the layer was 50% transparent and the fill was also 50% transparent, the end effect would be one of 25% transparency. If a doubling-up of mode settings is used too the results could be very unpredictable - although possibly quite interesting.
Next to the paintbucket is the gradient tool. This is another variation that allows us to fill a selected area with a blend between our foreground colour and background colour, or between our foreground colour and transparency. Double clicking on the gradient tool's icon will again bring forward the options palette.
The options palette now offers the usual mode variations on normal and the opacity slider. It also offers a pull-down style menu where the default is foreground to background. If we drag left side to right side across our selection we see a line which represents the extent of our blend. When we release the mouse, Photoshop creates a smooth transition between the two colours.
We also have a midpoint slider. This dictates how much sooner or later in the blend the centrepoint of the transition occurs. A low midpoint thus gives us a hurried beginning and a slow fade out. A high midpoint gives us a slow build up and a hurried end. In addition, we can limit a drag of the line itself. Any area of the selection not included in our drag will contain solid foreground or background colour, depending on which end of the blend it lies.
At the bottom of the options palette is a type pull-down menu which allows us to choose a radial style blend rather than the default linear blend. In this case our start point remains the foreground colour but our drag dictates the radius for a circular transition to the background colour. The radial offset, which now becomes active, dictates the percentage of the blend occupied by the foreground colour before the actual blend starts.
In the style pull-down menu, we can choose to blend the foreground colour to transparency instead of the background colour. We also have the option of each of the opposite directions. The clockwise spectrum and anti-clockwise spectrum options continue to use the foreground and background colours, but instead of calculating a middle blend colour they make the transition by including all the hues in the colour wheel that fall between the two chosen colours when moving in the direction specified. The dither button probably gives the smoothest blend but can be taken off if required.
Under the paintbucket in the toolbox is the line tool which we use to create straight lines of paint. As with all paint tools this will be in the foreground colour. Besides our normal options we can set a width for the line in pixels and create arrowheads for the start, finish or both. The arrowheads each have their own width and length settings which are now decided as a percentage in relation to the width and length of the line itself. The default is a pyramid shape arrowhead, but we can introduce a curvature by increasing the concavity setting. Holding SHIFT when using this tool constrain the angle of the line we draw to increments of 450.
To the right of the line tool is the eyedropper tool. This is neither a painting or effects tool. Rather, it is used to click in an image in order to pick up the colour values of that part of the image. By default, it is the exact pixel we click on that gets picked up and this colour then appears in our foreground colour box. Repeating this procedure whilst holding the OPTION key enables us to select a background colour.
If we look in our options palette there is a sample size pull-down menu. This allows us to select a colour based on an average value from a clump of 9 or 25 pixels (3x3, 5x5) and where the point we click is the centre pixel of the group. If we are using any of the paintbrush tools coming next, holding the OPTION key will give us the eyedropper while we are working, allowing us to change paint colour mid-stroke.
Beneath the line tool we find the eraser tool. In previous versions of Photoshop this has always been a block shape which becomes a single pixel at 16:1 magnification. It normally erases the image to reveal the background colour, but holding the option key whilst erasing allows us to erase to our last saved version of the file instead.
This version of Photoshop has introduced three additional options which allow the eraser to take on the attributes of paintbrush (the new default), pencil or airbrush. Its behaviour on one of these settings is the same as the attributes given to those tools where they appear elsewhere in the toolbox, except that the paint applied remains the background colour. The OPTION key function for the last saved version still operates, but an additional box for erase to saved in the palette now allows us to set this for an extended period of time. The fade and wet edges boxes will be covered shortly.
To the right of the eraser is the pencil tool. This allows us to draw hard-edged strokes with our foreground colour. It has one option unavailable to other tools - the auto erase box. This allows us to access the background colour whenever we make a secondary stroke over the foreground colour and is useful in pixel-edit (16:1) when we are editing close-up details.
The pencil and all the tools that follow are totally dependent on settings made for them in our brushes palette. We must now look at that before continuing down the toolbox.
By default, the brushes palette and the options palette are set up in the same palette group. These are probably on screen but if not we go to window - palettes - show brushes to activate them. Otherwise we click on brushes (behind options) to bring them forward. This displays a library of different shapes and sizes, which can be customised to our requirements.
We are going to need both brushes and options on screen shortly so we will now give brushes their own palette group. To do this we simply click on the word brushes and drag it into an empty space. When we release the mouse, a new palette group appears with the brushes as its sole content. We can now put the brushes and options palettes up side by side.
With the pencil still selected in the toolbox we notice that the brushes palette contains three lines of hard circular shapes. If this is not the case, we should go to the arrow at the top right of the palette, press and choose reset brushes from the pull-down menu. We can then click on the paintbrush icon (below the pencil in the toolbox) and notice the changes in the brushes palette,
Two things should happen here. Firstly we notice that the lower lines of brushes now a soft (or furry) edges. This is because we have changed to a tool that is flexible, unlike the pencil which is hard. Secondly, the chances are that a different size brush is now active. This is because Photoshop remembers the last palette positions for each tool separately, together with any additional attributes we set in this palette.
The first brush (top left) is always a single pixel brush and as a result is always square. All the other brushes are round by default. The brushes on the third line have numbers simply because their diameter is larger than would fit in the palette window. We can now go to the one numbered 100 and double click on it to see it in more detail.
A dialogue box appears which allows us to modify the attributes of this particular brush. We can start by reducing the diameter with the slider until it becomes 25 pixels. We can then increase the hardness slider to 100%. Notice that it now resembles the upper line of default brushes. We can now reduce the spacing to 1%. This represents the numbers of times the brush stroke will repeat itself as we paint. For example, this brush with a setting of 150% would paint as a line of individual circles with a half-circle empty space between them. We want the lowest figure here because we are about to create a very narrow brush.
Next we go to the icon in the bottom left of the dialogue box. Here we see a circle outline with a left/right line and arrowhead, plus a vertical line with two knobs on. We take the top knob, press and drag in downwards - effectively squashing the brush flat. Whilst we do this the roundness field reduces in value. We want 0%, which could also be typed into the field directly to save time.
The final adjustment we make is to key 1350 which could also be done by dragging the arrowhead on the icon around. This gives us a calligraphic style brush - slanted, with tight cuts and nice thick swoops. We can customise our own libraries of brushes this way. For example, a hardness of 50% would give us a semi-soft edged brush. A roundness of 40% would give us an eclipse-shaped brush, slightly more towards narrow than round. When we OK this brush it will replace the 100 brush in the palette.
We could now make more brushes without actually editing an existing one. We simply double-click in an empty space and a similar dialogue box comes up. We can also get both these options through the pull-down menu (arrow at the top right of palette) as brush options and new brush. If we are finished with a brush we simply click on it than choose delete brush from the same sub-menu.
We can extend this idea even further by creating our own brushes from scratch. For our first example we make a small (25 x 25 pixels or so) selection rectangle. This could be in the image but would normally be in a separate file. We then go edit - fill and choose contents - black, blending - 100%, mode normal and OK. This gives us a square black shape. We keep this selected and go back to the brushes palette, then the sub-menu where we find define brush. That shape now appears in the next empty position in the palette.
Next we could make a circular selection about the same size. We then hold the COMMAND key and draw another circle offset and slightly overlapping the previous one. The second chops out from the first and we are left with a crescent moon shaped selection. Fill this and define brush as before. Another technique might be to use the magic wand to select a completely unusual shape and then to fill and define that as a brush too.
There is a final variant in all this. The previous brushes were all given a black fill, but this need not be so. We can draw a selection shape anywhere in an image and bypass fill altogether. When we then go to define brush the result will be a picture brush. The one limitation here is that brushes can only be in greyscale so we need to use a distinct range of tonal values when we create them.
If we now double click on one of the brushes we have made we see that the options in the brush itself are slightly different. Obviously we already have our shape so we are limited to adjusting the spacing and removing the anti-aliasing effect. (The latter create hard edges that are rarely desirable, but...). With true 'picture' brushes we would usually need at least 100% spacing. With shapes only brushes, the spacing rules for normal brushes would apply.
Before leaving the brushes palette, we should look at the other entries in the palette's pull-down menu. When we have customised a set of brushes we may wish to save them for future use. We simply go save brushes in the sub-menu, give them a name and save them in a convenient place. To load them at a future date, we simply go to load brushes. When we load brushes into a palette, they will normally wipe any previous brushes away. We can, however, choose append brushes, which will join a new set to the bottom of any set already open. To reset the palette to Photoshop's default brushes, we simply go reset brushes.
We can also load the default brushes by going to load brushes and then load default brushes from the brushes and palettes folder inside our Photoshop folder. This was the procedure in earlier versions of Photoshop and is still the way to access the sample palettes that also come with the program.
We are now ready to try painting with these brush shapes. We simply choose the pencil for hard edges or the paintbrush for softer edges, select which shape we want and stroke in the image. Then we can try changing paint opacity in the options palette. Using the chart elsewhere, we can even try out the effects in the mode menu under normal.
We are now ready to return to the remaining tools in the tools palette where we will start with the paintbrush itself. We click on this tool and then look at the options palette where we have the usual opacity and mode list options, We also have two boxes - fade and wet edges. The wet edges feature is new to this version of Photoshop and simulates to effect of magic marker pens with the edges becoming more saturated than the stroke itself. It also renders the paint slightly transparent to obtain its effect, even when the opacity slider is still at 100%. We put this effect on by clicking on the box, but we make sure to turn it off when we are finished to avoid getting it by mistake later.
The fade option was also available to the pencil and will be repeated on most of the tools from now on. We click on the box to put fade on and then enter a number of steps for the effect, In previous versions of Photoshop this procedure relied on a distance in pixels for a fade-out of the paint to occur. Now, however, we key in a number of steps. One step is effectively the distance of the spacing you have set on that brush in the brushes palette. The default spacing on most brushes is 25% so we will need 4 steps just to occupy one real stroke of the brush. A sample fade is best observed with a steps setting of 50 using the 4th brush on the top row of the palette. If we keep this setting and use a bigger brush the fade-out takes place over a greater distance, on a smaller brush - much less distance. This operation certainly requires practice until it becomes familiar.
The default fade effect is to transparency, but we can set it to fade into our background colour. In this instance we are able to continue with our brush stroke indefinitely, but using the background colour. With the transparent option we would have to click again with the mouse to re-apply extra paint, Once again, we must remember to uncheck this box when we are finished with the effect.
On the left in the toolbox and next to the paintbrush is the airbrush. This applies paint in a way that simulates spray although the effect is negligible until you reduce the pressure slider. Infact there is not much difference between a hard-edged brush on the airbrush and a soft-edged brush on the paintbrush. Note that the terminology in the options palette changes to accommodate the conventions for a particular tool. In digital terms these distinctions are often negligible.
The options palette also includes a number of buttons for 'stylus pressure'. This refers to settings that can be used in conjunction with pressure sensitive drawing tablets - of which there are several variants. Each tablet manufacturers supplies their own software, although this frequently includes 'plug-in' functions that are available to Photoshop. The airbrush is one tool that makes good advantage of this technology and it has different effects with each of the other tools. Further examination of this particular aspect of working is not featured in this course.
Four tools remain in the toolbox which we have not yet looked at. Three of them have multiple functions which means there are actually ten tools in all. They all have one thing in common - they apply effects rather than paint, but behave nonetheless as if they were true brushes. For example, we can apply a pattern to the image just as if we were applying paint - altering its opacity, the mode, giving it a fade-out, giving it soft or hard edges,etc. The combinations are complex enough that even experienced users are prone to confuse exactly which functions they are on at any given time. This course aims to summarise the basic attributes of each only.
Directly under the paintbrush in the toolbox is the finger tool - commonly referred to as the smudge tool. This takes the brush shape you have selected for it and smears the paint under it on the canvas as you drag. We can liken this to dipping our finger into wet paint. Since this effect tends to be rather all-embracing a sample merged option is available, This means that even if we are on a layer the whole composite will be affected. If this button is off, only the layer itself will be smudged and, dependent on the layer settings, maybe only the opaque parts of that layer affected.
Another option box is for finger painting. When on, this tool still has a smudge effect, but momentarily applies a splash of foreground colour with each click. (In normal operation it is actually sampling the first colours we click on - greater or smaller in number depending on our brush size).
To the left of this finger tool we have something known as the rubber stamp tool. This contains five different tools - two of them with variants. These are now listed in the options palette under option.
We have previously discussed defining a pattern, taking a snapshot and the advantages of accessing our previously saved version of a file. This tool allows us to load either of the three onto an actual brush.
From saved and from snapshot simply load the relevant image which is locked into position on the canvas. We then paint and restore, or maybe paint with a mode variant or opacity setting, to get montage or ghosting effects.
Pattern aligned loads our pattern onto the brush and when we start painting the first click of our mouse determines the position of the pattern, We can then release the mouse and come back to continue painting outwards from that point. This is particularly useful when we need to pause during the painting of a large or delicate pattern area, or indeed, when the pattern is actually a large image in its own right. Pattern non-aligned also loads the pattern onto the selected brush but each new click of the mouse starts re-applying the pattern. This is more useful when the pattern in memory is small, especially if it is being used for textural effects with a low opacity.
The Impressionist option has no agreeable function where the writer is concerned. It purports to simulate the feel of an impressionist painting when we stroke over, say, a photograph. Unfortunately, it is near impossible to execute exact brush strokes, which means we end up doubling the effect in places to the extent of illegibility. Fortunately a number of art effect filters are available to do similar tasks with a degree of uniformity.
The final two option for the rubber stamp tool are actually the first on the menu list and concern clones. When we make a clone we are effectively taking a point in an image and loading that image onto a brush for painting in elsewhere. When we starting painting with that clone, the point we originally clicked will be the start point.
To make a clone in the first place we simply take the rubber stamp tool (regardless of its tool setting) and go in an image to the centre point required. We then hold the OPTION key and click on that point, thus defining the clone. This will then be available to the rubber stamp's clone options until we next define an alternative clone by repeating the process.
The clone aligned option will then paint in the clone in an offset position on the canvas as dictated by the first click of our mouse with the brush. We can then safely release the mouse and and take more care as we spread paint outwards from the source point. The clone non-aligned option will also allow to paint in the clone at an offset position but we must keep the mouse pressed until we finished. If we release the mouse and resume painting, a new copy of the clone will be begun.
The original function of the clone was to provide a way of copying part of an image and putting it on a brush to paint in elsewhere in an image. Subsequently, when we use the clone to paint, a guiding cursor also appears in the source position. Over time this function has extended and the clone tool now works across files, This means we can take a clone from one image and paint it into an entirely different image. The guiding cursor will remain on the sampled image, the real cursor on the new one.
When doing this with opacity and mode variations the effects are diverse, but there are two important points to note. Firstly, file dimensions and resolutions may vary. The clone will always have and keep the resolution of its source file. Thus it may appear considerably larger or smaller when applied to its destination file. Secondly, the rubber stamp tool loses its clone when the source file is closed. To do these tasks we have to keep both the source image and all destination images open on screen at once. This can take up considerable memory.
Directly underneath the rubber stamp tool is the blur/sharpen tool.
Brush strokes with these simply provide a blurring or sharpening effect to the image. Whilst blur is extremely effective sharpen can often generate too much colour noise and needs to be used with only a minimum amount (say20%) on its pressure slider. In addition, each time we stroke back over a given area we are effectively doubling-up on the effect. For this reason, precision blur and sharpen operations are best done by isolating selection areas and then using the more uniform blur and sharpen menus in the filter library. This will be covered later.
We change between blur and sharpen using the options palette. Because these are generalised effects there is a sample merged option which we can choose if working on a layer. Each tool has its own brush palette setting (as do all the rubber stamp options) with one exception. We can toggle between these two tools by using the OPTION key. However, when we do this, the brush stays the same as to tool we first started with.
The final icon at the bottom right of the toolbox is for three actual tools - dodge, burn and sponge. Once again, the OPTION key allows us to toggle between the first two whilst remaining on the first one's brush. Otherwise we would change between them using the tool menu in the options palette. In additional we must use this method to obtain the sponge tool.
The dodge tool basically lightens the image where stroked with the brush. The burn tool darkens the image. Instead of opacity or pressure, the options palette now offers exposure - terminology which, like the tools themselves, originates in the photographic world. The usual mode options are also absent for these tools. Instead we have midtones, highlights and shadows. Midtones (the default) means that the tools will generally affect the whole image. Shadows means that only dark areas will be affected. Highlights means that only light areas will be affected. Photoshop seems to distinguish shadow and highlight by the amount of gray or black in the colour mix, so the react effects are not always the same as the human mind perceives them to be.
The sponge tool is the latest addition to Photoshop's toolbox. This offers two options - saturate and desaturate, each with pressure adjustment. Desaturate, in particular, has a special use in conjunction with mode - gamut warning which will be discussed later.
We must now look at some items in the edit menu. We have already discussed methods of copying, cutting, pasting and pasting into, but this menu also provides two ways in which we can apply paint to selections. Edit - stroke lets us give an outline to a selection using either a pattern or the foreground colour at whatever width and transparency we choose.. We can also set this outline inside or outside the flickering selection line or centred on it. If our selection is 'floating' we must remember that any portion of the outline applied outside the boundary will affect the base image itself.
Edit - fill provides a quick way of filling for restoring large selected areas. Once again we can choose a mode and opacity for the paint, then select from various options. Foreground colour, black, white and 50% grey simply paint the area with these values. Black, white and mid-grey are particularly useful when preparing channels. Snapshot and pattern apply any pattern or snapshot to the area, as discussed next. Saved restores that area to the last saved version of the file, as we did earlier using the rubber stamp tool's option.
Edit - define pattern allows us to make a selection and then define it as a pattern. It will then be held in a buffer memory and be available as both a fill (above) and for use with the rubber stamp tool. It will be lost when we quit Photoshop or over-ride it with another pattern.
If the area selected to sample as a pattern is relatively small, it will paint or fill with a 'tiled' effect where its edges join. This can be compensated for by making a 'floatingf copy of the selection before defining it and then going to filter - other - offset. Here we can repostion the copy in a new postion slightly down and to the right using the 'wrap-around' option. This will cause two 'seams' to appear, but these can be normally disguised using our brush and smudge tools, but being careful to avoid the edges in doing so. We can then safely edit - define pattern, knowing that the top/bottom and left/right edges match each other - thus eliminating the 'tile' effect.
Edit - take snapshot allows us to 'save' an additional interim version of a file at any time whilst we are working. Like the pattern, this is loaded into an invisible buffer memory and will be available both as a fill and for use with the rubber stamp tool. Taking a later snapshot will over-ride it.
In the toolbox and beneath all the actual tools are the two colour boxes we looked at earlier. At the lower-left and top-right of these are two additional icons. The one with arrows allows us to switch the order of the foreground and background colour at any point. The small black and white boxes allow us to return to the default black and white colours at any time. This is particularly necessary when we are working with masks as we will now see.
Having studied the behaviour of our brushes we are ready to use them as a way of modifying the types of selections we made earlier To start, we would make a basic shape somewhere in the canvas with one or other of the selection tools. Then we go Select - Save Selection. A dialogue box appears and we must check the information in the two pull-down fields. This time we should stay with our current document and choose channel - new, then OK.
We can now click away from our selection to de-select it. That shape has now been saved to a channel for later use, but this has also provided us with a means to further modify it.
Next we go window - palettes - show channels. On the other hand, if the palette group with channels in it is already on screen we can simply click on the channels tab to bring it forward.
In the palette we see a list of currently available channels. The top one represents the composite image we are working on. This will be followed by three or four named channels which represent the colour separations that make up our image. Last of the list will be a numbered channel, the icon for which will contain the shape we saved a minute ago. We click on the name of this channel and Photoshop takes us to it.
All channels containing masks are in negative. The black area represents the part masked off - the white area is the active zone. When we make and modify selections in the image proper this is what we are actually creating. Coming to it directly means we are able to paint in additional changes to the shape. An example would be to use the paintbrush to bring the border in a bit. Then we could use the eraser to extend the white area out some more.
Next we click back on the composite (top) channel. This takes us back to the main image where we can go select - load selection and choose the number of the channel we were just working on. The changes we have made are apparent.
We will now develop this further. We click away from the selection to de-select it, then go back to the channels palette and click again on the name of the channel containing the mask. This time we will choose the smudge/finger tool, selecting a medium sized soft-edged brush with the pressure slider reduced to around 40%. We can then smear some of the white area out into black, then some black in the white. We continue with this effect until the mask starts to contain distinct elements of grey. Then we return to the top channel and select - load selection again.
This time the shape has changed again, but may look slightly different from the mask where the edges were grey. Until now, all of our selections have been simple outlines with an inside area and an outside area. Where we have added grey however, the border now contains varying degrees of transparency. In the absence of a hard edge, the flickering guideline will now be resting on the 50% opacity pixels.
To see this effect we double click on the foreground colour box and choose a new colour that is a complete contrast to the selected area. The we go edit - fill - foreground colour - 100% to fill it. If we then zoom into the area we will see that some edges have a fading out effect - with the more transparent pixels outside the flickering line. We then use COMMAND Z to undo this.
Back in the toolbox and underneath the colour boxes we find two icons that contain a circle in a rectangle. In normal working it is the left one that remains underlined. Keeping our present selection (now without its fill) active, we can click on the icon to the right. The image then fills with a semi-transparent tint whilst our selection area remains clear. This is known as quick mask mode and is exactly like going to a channel without having to save the selection first.
Choosing the paintbrush tool again, we are able to paint on this just as if it was a channel, except that black translates into the tint colour instead. The eraser will equally erase to provide clear areas.
After making some extra changes we click back on the left-hand icon to return to the normal image, where our selection differs yet again.
This will hopefully better explain another selection technique known as feathering. If we make a new selection rectangle, we can go to select - feather. In the dialogue box that comes up we type 20 pixels in order to apply the effect 10 pixels on either side of our flickering guideline. We OK this and at first glance nothing appears to have happened. Now we hold the OPTION key and drag the selection to a new position. As we saw earlier, OPTION move gives us a floating copy, but now we notice how different the edges are.
A feather is essentially a fade-out zone that can be applied to a selection. It is not visible when first applied because Photoshop is extracting the transparency information from the background - which itself remains visible underneath. When we copy or float the selection and move it elsewhere to effect is profound.
We can now click on the quick mask icon again the see this fade-out on the mask itself. Like our smudge adjustments earlier, there are elements of grey. In this case they are a smooth blend.
Clicking back on the normal mode icon, we can also go to the channels palette, then its sub-menu (arrow top right) and choose new channel. This will give us a new and totally empty channel where we can paint our own customised mask from scratch. This requires some experience and we must remember that the end result should be in negative to read into the image as a positive selection. We can even paste greyscale images onto channels to act as selection masks.
Channels also give us access to separations and blending facilities. We will be discussing these at a later stage.
Although Photoshop is not a Postscript language program, it can import and export the EPS file format. Most people working in a print-related environment on the Macintosh platform are therefore used to the techniques of object-drawing (and type-setting) using the Postscript page description language. This language was originated by Adobe systems and they have provided one feature in Photoshop that is designed to resemble the Postscript interface. This is the Paths palette.
By default this is in the palette group with channels and layers. We now click on paths to bring it forward. Alternatively, we can go to window - palettes - show paths. Here we see five additional tool icons which inter-relate with each other. (In some earlier versions of Photoshop this palette does not exist and a single pen tool can be found in the main toolbox.)
First we click on second icon from left - the ordinary pen. We then click somewhere in the image. We then move without dragging the cursor and click somewhere else. A line appears. We then click on a few more points. A shape is forming. We then continue this process to create an object-like shape and click back on a start point to finish it. We then change to the first tool icon in the palette.
This shape is known as a path. Using the first tool (arrowhead), we can now click on any of our points to adjust its position. We can also add a new point by clicking on an empty part of a line with the pen+ tool, or remove an unwanted point by clicking on the point itself with the pen- tool.
The final tool in this palette shows two lines at an angle. When we click on a point with this tool we can access two handles which will control the curvature of the line of each side of that point. These are known as bezier controls and will be familiar to users of Adobe Illustrator or Freehand. To see them, we must drag the point a little and then return to the first palette tool.
Next we click on each of the handles on either side of the point itself and drag them to see how the shape changes. This is a common way of getting curves into a path. Another techniques involves the first pen tool itself. If, instead of simply clicking on individual points to make the path, we drag and click, these 'handles' will come up automatically. This is useful if we want a highly-curved shape.
One reason to create a path is to get precision control over an area which we need as a selection mask. To demonstrate this we must go to the sub-menu of the paths palette (the arrow, top right as usual) and choose make selection. Here we can ignore the feather facility for now. We check that anti-aliasing is on and click OK. Underneath our path we can see that Photoshop has worked out a flickering selection guideline. We return to the palette sub-menu and choose turn off path. We can now return to the main menu safely,
Next we will hold COMMAND and drag a rectangle selection over half of the new one. This chops a bit out. We can now go back to the paths palette, then go to the sub-menu and choose make path. In the dialogue box that appears, a low tolerance (0.5 - 1.5) will give the best path. (A higher figure will generate less points on the outline but will have the advantage of taking less memory.) We OK this and Photoshop makes the path for us.
As we can see, this interaction between paths and selections works in both directions. As a result it also provides a way of saving selections without creating whole channels. (Each channel counts adds the equivalent of a separation to the file's memory.) To save a path we use the first palette tool to click on it and make it active. We then go to the sub-menu and choose save path, where we name it. A library of of path names will then build up in the palette under the path tools.
Also at the bottom of the paths palette are some small icons. These have been introduced to short-cut some of the most commonly used sub-menu commands. From left to right, they represent fill path, stroke path, make selection, make path and delete path. To delete a path it must be active and we simply drag its name to the bin icon. If we have a selection active in the image, clicking on make path will create one for us.
Fill path and stroke path are also dependant on having a path active and this terminology is slightly confusing. In Photoshop we cannot fill and stroke our paths themselves. Fill and stroke actually refer to effect that will be applied to the image proper based on the outline information of the path we choose. A stroke gives an outline to the area. A fill is much the same as that obtained earlier with edit - fill.
Paths that connect up to make objects/shapes are known as closed paths. It is also possible to draw an incomplete shape (eg. a wavy line) and add a stroke to that. For simple outlines on a selection area we would normally go edit - stroke. The facility in the paths palette extends this. When a paint or effects tool is active in the main toolbox, stroke path will adopt the style of that tool together with its brush size, shape and other attributes. This allows for a range of outlining effects. When no paint/effects tool is active, stroke path will bring up a menu asking us to choose which tool to use.
Using the paths palette sub-menu gives the full range of functions for these operations. Using the icons at the bottom of the palette relies on default settings which may differ. The sub-menu also allows for duplicating paths and choosing an individual path to be a clipping path.
A clipping path is the name given to a path that will act as an outline border for the entire canvas when exported to a Postscript program. We donate which path will be used for this by choosing clipping path in the sub-menu, then selecting the appropriate path from the pull-down list of all the paths we have available. Thereafter, any copy of the file saved in EPS format will be confined within the borders of that path. This is the way we would prepare cut-outs for a layout program like Quark Xpress or Pagemaker.
It is not possible to export outline-shaped raster (pixel-based) files from Photoshop in any other format that EPS, although it is possible to generate an alpha channel to act as a mask for the TIFF file format and to export path shapes on their own to Adobe Illustrator. Exporting an Adobe Illustrator file from Photoshop will not contain any pixel-based information. Rather, it will be a file of outline shapes that correspond to all the paths we have made in our file at that time. To make this Illustrator document, we choose file - export - paths to illustrator.
At the bottom of the main toolbox there are three remaining icons we have not yet mentioned. The left hand icon represents our normal screen. Moving to the centre icon removes all Macintosh interface facilities (including title and scroll bars) from the screen. This prevents us accidently clicking on the desktop and out of Photoshop. The right hand icon will remove all menus from the screen except for the control palettes. As with most videographic programs, these palettes can them be toggled on and off with the TAB key.
If we go Window - palettes - show commands a new palette appears. This is a quick-key facility which we can use to automate frequent tasks. Many functions (like grow and float above) have command keys built in. These are listed elsewhere and can also be found in the pull down menus by the commands themselves. Other functions do not have keystrokes but we are able to make them. To do this we must go to the sub-menu of the commands palette which we do by clicking on the arrow in the top right of the palette itself. We than choose edit commands and a dialogue box appears.
We click once on the command we want to change and then click the change button. We then go to real pull-down menus and actually perform the operation we want to automate - eg. inverse. This function is memorised and will respond to the corresponding F-KEY on extended keyboards, or the command button itself when the palette is on screen. Previous versions of Photoshop also allowed this facility in the preferences section. The functions were however limited to keystrokes since there was no palette.
It is possible to make 15 keys/buttons that will respond this way, plus another 15 that will respond to SHIFT in conjunction with the keys/buttons. We can extend this by saving a set of commands with save commands in the sub-menu, which will prompt us to name them. We will later be able to load commands or append commands in a similar fashion to that used with brushes. We can also reset commands for the default set. New command allows us to create a fresh command rather than edit an existing one, but is dependant on an empty space being available. It also allows us to assign a colour to a command. An additional function of edit command is the option of deleting a command and also setting column for the palette's display. Two columns is best if the palette is going to be merged into a palette group.
Window - palettes -show info brings up the information palette. This cannot easily be added to any other palette group and works best on its own. It has no function other than to provide us with information. By default this will include two sets of colour value read-outs for our pixel position and the x/y co-ordinates of the position. It will also display the width and height of any active selection. The read-out options can be changed by clicking in the tiny arrow icons beside the main read-out icons. Alternatively, they are in the pull-down menu as info options.
Later, when (for example) we are scaling, distorting or rotating our selection, this palette will extend to include new information relevant to that operation. More on that when it happens.
Window - palettes - show picker will bring up the colour palettes group which now contains three separate palettes.
The picker allows us to mix colours as an alternative to double-clicking our foreground and background colour boxes. We can select from the long colour bar at the bottom or mix using the sliders above. We can additionally change the colour mode of the sliders by going to the sub-menu, where we can also change the mode of the long bar itself.
The swatches palette displays a default set of colour patches. We can add a colour of our own by going to an empty space and clicking. It will then fill with our current foreground colour. If we click on an existing colour it becomes our new foreground colour. If we hold the COMMAND key and click on a colour we delete it. Holding SHIFT allows us to override a colour and replace it. The sub-menu allows us to save and load swatches for future use. We can also append swatches to each other and reset the palette. Some other colour libraries can be loaded as swatches. Note: It is best to avoid loading large libraries like the Pantone sets into the swatches palette. They are vast and impossible to scroll down!
The scratch palette is where we mix our own colours using brushes and the finger tool. It's like a miniature file for splashing around in. The sub menu allows for cutting and pasting between the palette and main image, It also has a locking facility and the usual load/save palette functions.
All the palette groups in Photoshop can be folded-up on screen to save space without closing them. We do this by clicking on the tiny box in the top right corner of the palette group. (The box at the top left corner is used for closing down the palette itself.)
We are now ready to look at the pull-down menus at the top of the screen. They can be summarised as follows.
The file menu is where we open, close and save files. This includes all import and export functions. We also have entries for tagging information to the file, printing and page set-up for printing. Most important of all, this menu contains our preferences sub-menu.
The edit menu has the Macintosh standard functions for cut, copy and paste. It also includes Photoshop's own variations on those themes. We also come here for filling and stroking (outlining) a selected area, defining our patterns and taking a file snapshot. Previous versions of Photoshop also contained something known as composite controls (paste controls in very early versions) which were also accessed here. These functions are now part of the layers palette and we discussed them earlier.
The mode menu is where we decide how the file is generated on screen and for print or video output. It affects how many channels (separations) make up our composite image and which colour values if any are available. Modes are exclusive for the most part although we can obtain some graphical effects by montaging files from different modes together manually. We also come here the access the colour table of a file (not usually available), to obtain a printing preview and to check-on/modify non printable colours.
The image menu is where we find the vast bulk of our image processing facilities, which are extensive. It is also where we find controls for scaling, distorting, rotating and flipping selections. One assumes they are located here since rotate and flip also function for the canvas itself. We also come here to resize our canvas or our file resolution. A trapping facility if available for pictures in CMYK mode. The image menu has additional features that let us duplicate a file and blend multiple files together using mathematical calculations.
The filter menu contains a number of effects libraries, including a extensive range of 'plug-ins' which have been developed by other software houses using Photoshop technology. The program ships with around 50 filters covering 9 fairly essential categories. Most filters give preview boxes to show us what's going on, others require specialised technical expertise. We will be looking at the most important filters at the end of this course.
The select menu contains the commands for modifying selections we have made. We have already looked at these.
The window menu simply gives us access to palettes, levels of magnification and rulers. It also has a new window facility that allows us to open a second copy of the same file where we can make quick changes to blend into the same image.
We start with the file menu. New allows us to open an empty canvas. We would use this for painting directly, painting (drawing really) by making selections with fills and strokes, or a combination of both. Since the default settings are the same as the clipboard we would also use this to make a temporary file - which, as we saw, can be used to modify a selection. Lastly, this feature can be used to set the size and resolution of a base image in which montage components will later be added,
New file contains fields for both the physical dimensions of the canvas and its resolution. The result will be an exact intensity of pixels and a subsequent file size in megabytes. All the pixels will be in a background colour ready to work on. At this stage we can choose whether the background colour is the same as that currently in our background colour box, or one of the other options like white. New to this version of Photoshop is the transparency option. No real background will ever be transparent - this choice simply gives us a first starting layer.
It is best to set up a canvas with the dimensions of the image just slightly larger than the final printed size. We then consult our printer to find the size of the screen frequency used for printing, double that figure and enter it into the resolution box. In practice we prefer to round this figure off to the nearest 50 (but by no less than 30) since output will probably be from another program and we may later need the freedom to re-size slightly. This is also the reason why we add some extra to the dimensions itself.
This set up also functions to calculate the file size for certain scanners. If we have set up the information above (exactly in this instance), the figure given for the file size in megabytes will be enough to generate a correct resolution file based on dimensions alone. This must however be done with CMYK as the mode option, since the number of separations in the file is a final variable.
The entry for open allows us to open virtually all existing files. In the dialogue box we locate the item we want and Photoshop tells us its format. We then OK to open. There is also a 'find' facility to search for items we know by name. Previous versions of Photoshop also contained an open as dialogue which enabled us to decode unknown format files by trying out variations, whilst open was more limited. Open now combines the two. If the file we are looking for is not a file type immediately known to Photoshop it will not be on the list. If however we click the box show all files, Photoshop will display everything in the given folder and offer under format a list of options we can try to open it.
Many years ago and long before it offered paint/processing facilities, Photoshop was conceived as a program designed to convert computer file formats from one platform to another. It still retains some of that function which may even be a format known as raw on the list. This is merely a vast array of computer binary code numbers describing pixels and colour information - mostly used by scientists and researchers on dedicated computer platforms. So long as the file has not been compressed in any way and you know its physical dimensions and the number of channels it contains, Photoshop will do its best to decode it. If we have only part of the information, we can enter what we know and click on guess. We may have to click on guess several times since Photoshop will cycle through a range of possible calculations each time we do so.. In case we made entered the width and height wrongly, there is also a swap facility.
The place facility allows us to open a file as a floating selection on top of any image already open. Only object-based files that recognise an empty background can be obtained here and the most common format is EPS.
Close is an alternative to the standard close box. Save will simply update any changes made to an already open document and is not recommended until we are sure there is no further need of our last saved version. (see earlier) Save as is where we would normally go to name and save our work. Here we have another format list with all the available file types Photoshop is able to write. It is most practical to save a master file in Photoshop 3.0 format, since only by doing this will we be able to retain information relating to channels, layers and other items. An alternative is version 2.o format, but this will flatten all layers.
Saving a Photoshop 2.5 file is not necessary since that version is able to read version 3.0 files if Photoshop 2.5 compatibility has been checked in the preferences section. In doing so, however, it will be forced to flatten the layers when opening.
Having saved a Photoshop format version of an image, we would normally save an additional copy in the file format we want to export. (eg. PICT, TIFF, EPS) We will return to this later.
A new feature is the save a copy function. This will save a duplicate of our image without affecting our master file. This could for example be sent to a client for approval as a composite image. If the copy is saved as a Photoshop file we will see two boxes giving us the option of fattening our layers and/or deleting our channels - particularly useful if our client needs a Photoshop file and we don't wish them to see our working data. Other formats are available too.
Revert is a facility that allows us to restore the screen to our last saved version of the file. Unfortunately it means we lose not only the changes we have made, but also any channels and layer information. It is often better to and fill - with saved to achieve the same effect with our working information intact.
This brings us to acquire which displays a pull-down menu. These are the various sources that we might be using to obtain an image directly - a scanner, a video frame-grabber or a camera for example.
PICT Resource is a previously made PICT file that may be 'embedded' in another application like a multimedia demonstration. Photoshop will tell us how many PICT format files are available wherever we are looking and give us scrolling facilities to cycle through them. When we find the one we want we OK and it opens as a new untitled Photoshop file. This is great for stealing other people's artwork!
Anti-aliased Pict allows us to apply Photoshop's anti-aliasing technology to certain kinds of PICT format files before opening the, This prevents us decoding the object-based hard edges that would otherwise appear if we has simply gone to open. The result will open as an untitled file.
Decompress JPEG allows us to 'unsquash' this format. We can save in this format under export shortly and it is primarily suitable for long term storage when we want to economise on disk space.
TWAIN acquire and the related entries here are built into Photoshop but only apply to scanner and other devices that support the particular TWAIN interface. For those that do, it replaces or loads the standard scanner plug-in that would appear here.
'OTHER SCANNER' plug-in will usually be found on this list. It allows us to scan an image and open it directly in Photoshop. This saves the time taken to launch dedicated scanner software and also the file conversion often necessary afterwards. The plug-in will normally be a duplicate of the scanner's own operating interface.
Acquire - Quick edit is a new facility which allows us to open a portion of any large file provided it is a non-compressed TIFF format. This includes files saved as Photoshop files using previous versions of Photoshop. To use this, we choose the file we wish to open and then select the portion we wish to edit with the rectangle selection tool. That smaller area of the file will then open on screen in isolation. When we finish working on it we will then go export - quick edit save and the master file will be updated. The quick-edit image remains on screen until closed.
The Acquire menu will often include additional plug-ins for external file formats like Photo CD or hardware like video grabbers.
This brings us to Export. This is where we save our quick-edit changes and also where we archive files using JPEG compress. Both of these have been mentioned since their opposite functions are in the acquire menu.
Also here is export - paths to illustrator. Going to this allows us to generate an Adobe Illustrator file containing any paths we have created and saved using Photoshop's own Paths palette (Pen Tool). All paths in a document are exported this way and will open in Illustrator in their entirety, but without any fill or stroke. Individual paths can also be used as export masks (clipping paths) but this is done elsewhere. (See Paths Palette or Save As - EPS).
File Info is another new facility which allows us to enter categories and keywords by which the image will be located when using graphic database software like Kodak's Shoebox, Corel"s Mosiac or Aldus/Adobe's own Fetch. This information is tagged to the file when saved and will read across most formats.
Both Page Set-Up and Print replicate standard functions and are not discussed in detail here. It is however worth noting that the page size chosen in the set-up dialogue box also affects the graphic thumbnail we see when option-clicking the file size box at the lower left of the image. Screen angles and transfer functions can also be set here but would usually be applied in preferences following. When used they will over-ride the file preferences which may be useful if printing a proof with different attributes to the final output.
File - preferences contains a group of sub-menus which have greater or lesser importance depending on the situation. We are considering here only those which bear on virtually all environments.
Preferences - general allows us to switch to the Apple colour mixer if necessary, choose a faster but poorer screen redraw (bicubic is the best, but slowest), opt to view colour separation channels in their actual colours and select various icon arrangements for our tools.
The more section (button) contains options for thumbnail previews and various checkboxes. Of these, four should normally be on. Anti-alias Postscript will ensure any EPS files used in Photoshop will arrive with smooth edges. Export clipboard will ensure that anything on the clipboard is converted to Apple's PICT format when we quit Photoshop. Short PANTONE names will ensure that software like Quark Xpress can identify which Pantone colours we have used (where appropriate - ie. duotones). 2.5 format compatibility will allow us to open version 3 files in version 2.5 of Photoshop (although we should also remember that any layers will be 'flattened' if we do so.).
Preferences - Gamut warning simply allows us to choose a colour and opacity for the display of out-of-gamut colours when in Gamut Warning mode.
Preferences - plug ins allows us to choose which folder (including any sub-folders and their contents) will be scanned next time we launch Photoshop. This will dictate the size and extent of the items available on the Photoshop interface. In some situations it is obviously more economical to limit the size of our plug-in menus to those required for a particular job. This is how we do it. We always single click the folder chosen and then click on the select... button underneath.
Preferences - scratch disks allows us to choose one or two external drives which Photoshop can use for its virtual memory requirements. As with plug-ins, the settings chosen here will be used next time we launch Photoshop. Drives used for this purpose must be writable and should not be prone to fragmentation. If using a Syquest drive for example, we should ensure that we not only have free space on the cartridge, but that it is previously unused. The primary drive should be the one with the most space available. The secondary drive can be either be one with lesser space as back-up or our system's regular hard disk. When not using this facility, we should always check that our hard disk is returned to position as the primary drive.
Preferences - transparency allows us to choose a pattern and colour scheme for the transparent areas on Photoshop's new layers. Alternatively we can choose none.
Preferences - units allows us to select a principal measurement for our rulers, opt for traditional point sizes on our type and setting an invisible column grid which can be accessed with our regular rectangle selection tool.
Preferences - monitor set-up will give us a list of known monitors and we should select the one we are using. The parameters will then be correct. A gamma of 1.8 is the default for printing, but this should be changed to 2.2 for videographic output. The ambient light setting should be low for a darkish room and high for a bright room. It is best to have consistent lighting at all times of day. Where this is not possible, we may have to change this setting frequently.
If our monitor manufacturer had not provided Adobe with the information needed, it would be necessary for us to refer to our monitor manual to obtain the parameter information. Having done this once, we would choose the save button and name the file. The settings would then be available with the load button at a later date.
Preferences - Printing Inks Set-up provides us with a library of popular printing inks and again offers us facilities for saving and loading our own. This information directly affects our screen display when we move to CMYK mode and should be applied to the file whilst still in RGB mode.
The above information can also be exported with the file, but is not advisable except where approved by our printer. All printing presses have slight variations due to age and machine registration, so printers will often need to adjust the dot gain compensation to their own requirements. Since consistency is their over-riding concern, these attributes are frequently applied to an entire document (publication) in one operation and using high-end pre-press software. It would be inappropriate to apply conflicting information to an individual Photoshop file in such cases.
Preferences - separation set-up and preferences - separation tables are not covered in this course since they relate to the press situation above and do not normally concern designers. Nor do they affect the screen representation of the image for working purposes in the way that monitor attributes and inks do.
In dedicated working environments such as newspapers, Photoshop is used to apply printing information defaults to images and then exported to page make-up software using EPS formats, When exporting, two checkboxes are available - Include Accurate Screens and Include Transfer Functions. If these are put on, no other application will be able to over-ride the Photoshop attributes. They should subsequently be used with caution.
In such situations a dedicated calibration operation is often performed which involves the making of several proofs in order to fully optimise Photoshop and our screen itself for the exact output form in question. One aspect of this is the use of a gamma control to create a screen illusion of the paper stock to which we would be printing. This facility is of additional advantage to anyone using Photoshop but is not found directly under preferences. Instead it is a control panel.
When we first install Photoshop this gamma control will not be available. To locate it we must go to the Photoshop folder, then the 'goodies' folder and then the 'calibration' folder. The item named gamma should then be dragged to the desktop and then into our system folder, where it will automatically be placed in the control panels folder.
Back inside Photoshop we open a suitable image and then choose control panels from the Apple menu. We then double-click on the Gamma icon. Using the button at the bottom left of the dialogue box we switch it on and then adjust any controls on our monitor itself to suit our general eyesight. It is then advised that we place masking tape over these controls to prevent further adjustment.
In the gamma control we keep our target gamma at 1.8 for print (2.2 for video) and look at the line of alternating grey bars below the buttons. Moving back from the screen and possibly out-focussing our eyes a little, we then move the gamma adjustment slider left or right until these alternating patches appear to blend together into a solid line. Since everyone's eyesight is different there is no real consistency in this operation.
Next we obtain a sample of the paper stock for which we are working (ideally we will have a proof of the image on screen printed on that same stock) and hold it up next to the screen. Since there is significant light coming from the screen, we should angle this paper to reflect the maximum amount of available light at our eyes. We then click the white point button and use the three coloured sliders to replicate the appearance of the paper itself. Where we have a proof available, we can also adjust for the appearance of solid black and any illusory colour cast in the balance.
When the screen resembles the paper as best possible, we may find that the alternating greys at the top have shifted again. In this case we must make a final gamma adjustment to restore the bar to a neutral shade of grey as before. We then click on save settings to give these attributes a name which we can access with load settings for any future work on the same paper stock. We then repeat this procedure to create a library of all our frequently used papers.
We should bear in mind that this gamma control does not affect our files in any way. It simply adjusts the screen to give us the best possible impression of the printed result based on any other attributes we have previously applied. It's advantage in being a control panel is that we can also use it in other programs such as page make-up software. We can also turn it on and off at any time.
We are now ready to look at the remaining pull-down menus. These are mostly concerned with file modes and image adjustment. To see some of the basic functions we must start with a greyscale image. If one is not available, we would choose a colour image with a good tonal range and then go the mode - greyscale. We then OK the discard colour data box and Photoshop will redraw the file with 256 pure grey shades. If we do this we must remember to save the file under a different name unless we want to lose those colours permanently.
A greyscale file is a single channel document since there are no colour separations. We can still make our own channels for masks or artificial colour overlays, but these are not in the defaults. It is the most basic of all modes in which Photoshop's tools are available.
We can now temporarily choose mode - bitmap. The bitmap mode is even more basic in that it generates of file containing only pure black and white pixels. The choices define the way the current greys will be converted to black and white. 50% threshold is most radical in that all light greys become white and all dark greys become black. The other options will produce better graphical effects but any appearance of grey will only be an illusion. We try these in turn using COMMAND Z or File - Revert to return to our greyscale image.
Since most of Photoshop's tools do not work in bitmap mode there is little reason to use it. Even the graphical effects can be achieved in greyscale. It is worth noting however that when exported, the white areas of a bitmap can be read by a destination application as 'transparent'. A lot of the more primitive 'clip-art' packages are also simply bitmap graphics. If we were to open these in Photoshop to enhance them, we would need to convert them to Greyscale in order to activate our tools and menus.
We will stay in greyscale whilst we look at the first group of image adjustment options.
First we choose image - map where we find 4 options. Invert will turn our image into a negative. Equalise will redefine the darkest shade as black and the lightest shade as white, then estimate an average well-balanced tonal range for the remaining shades of grey. Posterise allows us to key in a specific number of grey shades, simulating the effect of silkscreen printing or maybe reducing Photoshop's 256 greys to a more readily identifiable 100. Threshold allows us to turn the image into pure black and white, giving us a slider with which we decide which greys go to black and which to white - a similar graphical effect to the bitmap mode seen earlier.
The effects of each of these map adjustments are obvious in greyscale. In a colour mode they will slightly more variable since they can be set to affect an individual separation or the grey component in the composite colour mix. The actual operation will relate to that seen here.
Next we have image - adjust, where there is a mixture of established and new features. Many are 'greyed-out' on the menu since they only apply to colour images and will be looked at later. First, we look at the third down, brightness/contrast, which can best be described as resembling the knobs on a black and white television set. Contrast is useful for making an image appear crisper or more muted. Brightness affects the overall look of an image and will always end up removing either pure black or pure white depending on our drag direction. We should check that the preview box is on to see these effects as we experiment,
We now go to image - adjust - levels which offers a more powerful way of adjusting the range of greys in the image. Once again we should put the preview checkbox on to see the changes as we make them.
The first thing we see is a display like a graph - this is known as the histogram. Above the black triangle (in the top half of the dialogue box) there is a 'tower' of black representing the number of solid black pixels in the file. Above the white is a "tower" representing the number of white pixels. In the centre is a grey triangle which will be directly underneath the 'tower' representing the 50% grey pixels. In many cases however, there may be no "tower" at all above the solid black or white. this will be because the image does not contain the entire range due to a faulty scan or otherwise.
We start by dragging the black triangle inwards until we reach the start of the darkest grey pixels. We then bring the white triangle in the meet the first of the lighter grey pixels. We note how the image changes to display a wider range of tones across a well-balanced spectrum of greys.
The middle grey triangle will also have re-positioned itself in the new centre position. This could have caused our original middle grey to have shifted to a considerably lighter or darker position. We can now compensate for this by re-adjusting the mid-grey point by moving the grey triangle itself. A drag leftwards will increase the spread of lighter greys. A drag rightwards will increase the spread of the darker greys. A wide range of tonal variety is thus available.
When we have used these input levels to obtain the best effect for the histogram/map itself, we can additionally use the output levels triangles. Here, white represents general highlights and black represents shadow areas. We can mute either or both of them by dragging the triangles concerned inwards towards the middle of the slider. Going beyond the centre will be rendering some parts of image into negative - reversing the black and white triangle positions will infact create a total negative. Usually a subtle change is all that is required.
This levels dialogue offers other ways of achieving a similar result. We cancel this box and then return to it. This time, we click the button marked auto. Photoshop will then find the darkest and the lightest shade shade of grey, then redefine them as black and white automatically. Closer inspection of the histogram will also should that the actual point of definition is actually slightly further in than the end point of the map. This is because auto additionally calculates how many shades of light and dark greys are needed to generate a decent percentage of pure black and white in the image.
Beneath the auto button are three eyedropper icons. We select t the one with black in it and click in our image on a dark shade of grey. Photoshop turns it black and redraws the rest. We then do the same for white with the white dropper. The mid-grey dropper is strangely unavailable for this procedure. It will appear when we move to colour modes.
Levels also offers a load and save facility. If we work out a careful image correction (compensating for a bad scanner for instance) we may want to apply to exact same changes to another image. Save allows us to name the settings and save them for future use. we use load to get them when the time comes.
This save and load facility will be available in many subsequent dialogue boxes but will be by-passed from now on. Third party plug-ins have also been developed which allow users to automate the task further by applying any settings we have made from the main interface rather than by opening the dialogue box. We can also create a function/command key for the box itself by using the commands palette discussed earlier.
Image - adjust - autolevels is a new menu feature. This simply allows us to press the aforementioned auto button without opening either levels or curves (see next). This will obviously redefine the black and white position, but the effect is immediate and no additional dialogue box is available.
Image - adjust - curves provides yet another perspective from which we can adjust the greys. This time, instead of 'towers' we have a positional graph defining the spread of our original greys without reference to the quantity of pixels of each. The dropper icons and the auto button have exactly the same functions as before, so we can try each of these to compare what happens to this graph in relation to what happened with levels earlier. We notice how the extreme ends of the tonal range come inwards as we redefine black and white. In cases where we have a well defined black and white point, pressing the auto button also allows us to restore our original without cancelling the dialogue box.
In Curves, dragging the lower left handle to top left and then the upper right handle to lower right will create a negative. Any graph we create for ourselves will thus be somewhere in between that frame of reference. We now click on the centre point of the graph and a new handle appears showing our mid-grey. A dropper tool also becomes active in case we want to identify a grey using the image itself. We select a random grey and its position is shown with an icon on the line of the graph. We can now drag our centre point to that position and affect the image on each side of that point. To demonstrate this we leave the left side handle alone but drag the top right handle down to the bottom. A curve appears as Photoshop automatically smooths out the corruption we have generated. Every grey tone darker than the one we positioned ourselves at has now effectively become a negative - even though all the lighter tones remain in positive. A wide and wild variety of strange and metallic effects can be generated this way, although the real purpose of this dialogue is to make a slight adjustment to the curve as a way of enhancing the image.
Changing to the pencil tool in the curves dialogue box also allows us to draw our own graph. Clicking back on the curve tool will give us handles again. Whilst still on the pencil tool, there is also a smooth button. This enables Photoshop to generate a smooth curve from a jagged line. As with levels, the curves dialogue is available for individual separations or the composite image when we are in a colour mode.
Image - adjust - replace colour is another new menu item which we will look at shortly. It also provides a limited function in greyscale, We choose the plain dropper tool to select a shade in the image we want to adjust, add more shades with the dropper + and remove unwanted ones with the dropper - tools. Setting the fuzziness slider to zero, we click the selection button which displays a selection mask of the shades we have chosen. We can then increase the fuzziness to spread the selection outwards (This is identical to our colour range selection process in behaviour). Lastly, we affect the lightness of those particular areas using the slider at the bottom. The changes we make are isolated in those areas.
We are now finished with the greyscale specific functions and can return to the mode menu. We go mode - duotone. Duotones, tritones, quadtones and indeed montones are available under the duotone category. All require a greyscale image as a starting point since the process involves transparent overlaying duplicates of the same greyscale image, but printed in different colours and with a slight tonal variation for each. Monotones are simply a greyscale printed in a coloured ink. Tritones and quadtones involve three or four inks and adjustments are also made for the order in which the inks are printed.
In this instance we are limiting ourselves to duotone, which we select from the type pull-down menu in the dialogue box. We then get options for two inks. Double-clicking on each of the right hand squares allows us to select which colours will make up the image. The procedure is identical to loading our foreground and background paint colours and is most commonly used here with Pantone colours since Duotone (generic) is the only mode that will allow us to export Pantone colours intact.
Note: All other modes convert Pantone colours into fascimilie colours appropriate to their own colour separation values. We can only create a selection mask to represent a Pantone (ie. overlay or knockout) and export or print that channel separately, assigning it a real Pantone value in the destination application. Sadly missing from Photoshop is the ability to generate a specific two-tone mode which will accommodate individual Pantones as separate channels. In duotone we export the Pantones intact provided we use the EPS format and check short Pantone names in preferences where appropriate.
In duotone options, having mixed our colours, we now double-click in the corresponding left hand boxes make a curve which will define the tonal range and balance for that particular colour's printing. We can also type into the fields any new values we specifically want and we delete the information from any field we want Photoshop to calculate for us. For example, on one colour we can type 0 into 10% and 100 into 90%, which will spread out the greater range of greytones whilst knocking out the previous extremes to pure black and white (Colour or none in this case).
In duotone mode the file is a mixture of the two colours we used, although there will be little if any trace of each colour in its pure form. For example, if we chose a blue and a yellow the resulting file will be green. There will however be an enhanced dynamic range within the green itself gained from the duotone effect and impossible to create using a single colour in isolation. We can now select, adjust and paint as normal except that our colours palette will be limited to those mixable from the colours we have chosen. We are also able to copy duotone elements across modes for composite effects, even though the colour values technically change in that process.
We must remember that duotones are also defined as single channel documents with only one composite channel. We can make our own additional channels but it is not possible to access the real separations except by going back up to mode - duotone and changing the curves. We can use the normal image - adjust - curves but this will only affect the curve for the entire composite image.
Note: Under no circumstances should we attempt to split channels using the sub-menu in the channels palette. If successful, this will result in a mode - multichannel file which effectively means Photoshop doesn't recognise it anymore.
We now go to file - revert which returns us to our greyscale image. Before looking at colour adjustments there is one final effect we can use for greyscale images that is particularly effective in montages. We go the mode - RGB which will covert our greyscale to RGB without losing the greyscale appearance of the file. Then we go image - adjust - hue/saturation. Because we are now technically in a colour mode all the additional adjustment menus are available, even though we have no real colour values to play with. This can be remedied by clicking on the colourize checkbox at the lower right of the dialogue box which appears. Photoshop tints the image for us.
We notice the position of the sliders. Photoshop has given us a mid-range hue. We can change this by dragging the hue slider left or right until we find a suitable colour. We can then reduce the saturation to something more acceptable by using the saturation slider. One difference between this effect and that of monotone/duotone is that the colourize process keeps pure black and white intact from the greyscale image, as if we were looking through tinted sunglasses.We are also able to make further changes with the lightness slider, but this will finally remove pure black or pure white from the image and may mute the effect somewhat. An element processed this way could then be cut and pasted into a proper colour montage as before.
Before continuing, we should close our greyscale file and open a full colour image. It is best if we choose something with a wide range of colour and contrast.
Next we go to image - adjust - colour balance where we can make basic changes to the balance of colours already in the image, channel or selection. With the midtones button on, changes affect the overall balance, with shadows on the darker areas are those predominantly affected and with highlights on the changes mostly affect the lighter areas. Both shadows and highlights define themselves in relation to the amount of grey or black in the colour mix, as if it were the pigment in paint being affected. This is particularly apparent in RGB mode where there is no separate black or lightness channel with all the sliders affecting the grey component to some extent. A new feature of this dialogue box is the preserve luminosity button which addresses this by omitting lightness values from our changes.
To implement changes we simply drag the sliders to increase or decrease the colour cast represented by that particular colour channel. It is similar in nature to lightening or darkening the individual separations themselves. (For more precise control over this effect we could actually go the the separations via the channels palette and adjust them individually with levels or curves, but this would be time-consuming.)
Next we have image - adjust - hue/saturation which we have seen briefly before. This dialogue box offers much greater control over the colour appearance of an image and includes load/save dialogue. When the master button is selected the changes affect the entire image. When any other button is selected, only the ranges of colour relating to that colour are affected and the changes are seen in both the image and the target colour box. We could for example change the yellow to green by dragging the hue slider to introduce an appropriate colour value. Radical changes will then be seen in yellow parts of our image, other elements containing yellow will be slightly affected and parts of the image with no trace of yellow will remain unaffected. We can additionally use the eyedropper to select a sample colour from the image and gauge our changes from the sample box at the bottom.
This dialogue can be used to change the overall look of an image when set to master. For example we can alter the look of a photograph to simulate a different time of day, introduce sunshine into a photograph taken on a dull day.
Next is Image - Adjust - Replace Colour which is another new feature. The behaviour of this dialogue is virtually identical to that of colour range discussed when we were making selections. This time however, we are making changes directly inside a selected area without ever accessing the selection mask. We gauge the area being affected by clicking the selection button and initially reducing the fuzziness slider to zero. We then use the plain dropper tool in the dialogue box to pick a colour from the image that we want to replace. If we want to replace more that one colour range with the same adjustment we continue picking colours using the dropper + tool, removing any picked by accident with the dropper - tool. As an alternative we can use the SHIFT and COMMAND keys to add or subtract colours whilst still using the plain dropper tool.
When the black and white preview mask shows all the areas we wish to change, we can 'grow' these out by increasing the fuzziness slider's setting. We then use the hue/saturation and lightness sliders as we did with our previous dialogue box, but this time only our chosen colour range will be affected.
Yet another new feature is image - adjust - selective colour and this allows us to modify our image by changing the effective amount of any individual ink being used to create it. The red, green and blue channel separations also count as inks for this purpose, but the main function of this facility is the match high-end printing processes which obviously use cyan, magenta, yellow and black. The colours pull-down menu also offers neutrals, which will introduce a more general bias of change, and white, which acts for RGB mode as the inverse of black in CMYK.
Having chosen the colour we wish to modify, we then use the sliders to increase of decrease the amount CMYK inks creating that colour area. This is calculated with percentage values and method shows two options for this.
Absolute is the easier to comprehend since it adds the percentage we choose to the total amount of that ink (ie: 30% original ink + 10% adjustment gives us a 40% level of ink, although this will obviously never exceed 100%). It will also work with any value we may have selected.
Relative offers more technical control in that we implement a percentage of change to the percentage values of existing inks. (ie: areas of 30% ink + 10% will now become 33%, although areas of 75% ink will become 82.5%.) This method will not have any effect if either blacks or whites are chosen for correction, since they do not contain any colour values on their own and this relative calculation would have no information to work with.
Image - Adjust - Desaturate provides us with a quick method of coverting a colour image to greyscale without changing modes. This command effectively removes all colour values from the image, leaving us with just the grey components. It will not however work as a substitute for the sponge tool when in gamut warning mode.
Image - adjust - variations gives us a rapid method of making many of the image adjustments we have just seen - albeit with less accuracy. The dialogue box displays a grid of 'thumbnails' with an original at the top and a current pick in three central positions. We can then click on any of the other options to change the current pick. These are arranged to display opposite values - ie. more green equates with less magenta. The fine to course slider allows us to adjust the increments of change from subtle to radical.
The default for this dialogue shows a midtones button. We can also make changes based on shadows (dark areas) or highlights (bright areas) and saturation (overall). When opting for these other variants it is often difficult to see the changes on the 'thumbnails' themselves, so a show clipping box is provided. This will display an overly saturated 'neon' effect on the 'thumbnails' which represents the highlight or shadow areas being affected. It is for guidance purposes only and can be turned on or off. It does not represent a wild change to our image.
The variations facility offers enough control to make general changes to images very quickly. There is also a load/save facility as with many other functions. In most image adjustment cases it may only be necessary to use one other menu item for fine tuning, Variations is actually a 'plug-in' module for Photoshop. Where it does not appear on the menu, we should check that it is in the chosen 'plug-ins' folder. If not, it should be installed from the Photoshop disks/cd.
Image - duplicate allows us to create an identical copy of a file. This is often necessary for both image - apply image and image - calculations which follow on the menu. It is also useful when we need an extra copy of the file to work on keeping an original available for the rubber stamp tool's clone facility. (This technique effectively squares our restoration facilities when used in conjunction with from saved and snapshot, but has to a great extent been superseded by the multiple 'undos' possible by working on the new layers.)
Note: In earlier versions of Photoshop there was a single entry for image - calculate. The command for duplicate was still available, but contained within that sub-menu.
Image - apply image and image - calculations now replace the previous image - calculate menu. These offer some of the most powerful features in Photoshop since they enable us to blend entire files together, either in their own right or using different masks and pasting modes. These commands were originally designed simply for blending channel masks and have developed over time to accommodate multiple colour images. One restriction remains however. To execute these blends we have to prepare the files as if they were the channels (albeit in colour) of one document. Each component file must be identical in physical dimension and resolution, plus they must be in the same colour mode, with the exception of masks which can be greyscale images of the same physical size and resolution. The best way to prepare these 'base' components is to image - duplicate to make multiple duplicates of our largest component, select - all with each of the duplicates and delete (key) to a white background. This will provide us with any number of 'blank' canvases into which we can copy, paste and position the individual items or files we want to blend together. We should also ensure to save as and name all these duplicates in such a way that we can identify them easily when we see them listed.
Image - apply image generally offers a set-up for overlaying one image from this list on top of another - as if the top image were a floating selection or a layer. The bottom image is the result file and is referred to as the target. The overlaid and applied image is known as the source. The target file is the file we are using when accessing this menu. The source file is whichever one we select from the source list. We must then choose which channel and layer of that source file will be applied to the target, although this will often simply be the composite channel and background layer. We can also choose whether to invert (make a negative) of the source file before applying it.
In the bottom part of the dialogue we must then choose a mode for the blending and an opacity percentage for the source file. The default would be 100% normal. If the source were a layer, the preserve transparency option would also be available. (Refer back to layers and painting modes for details, or see modes chart elsewhere.)
A final option is to click on the mask button which will give us yet another list. Here we would choose another of our duplicates to act as the 'masking-tape' area through which the source file would become visible. Since most channels behave in negative we would normally invert this mask file. Additionally, if the mask chosen is a colour file we must choose a channel to use. Grey will take a greyscale version of the composite colour image involved. The other channels will take whichever separation we select.
Note: Images used as masks should be considered just as if they were selections we had made ourselves and then saved to a channel. White areas will count as active area. Black areas will be masked off. Shades of grey will adopt varying transparent values when applied as a mask. Using photographs or other images as masks can create a wide variety of effects, but the images used are not always as conspicuous in the composite result as we may have hoped.
Once we have used image - apply image to make an initial blend, we could return to the same menu and apply yet another component from our list of available files. Or we could apply the same component again, but using a different mask or different settings. There are countless combinations available with practice.
Image - apply image has one drawback. The source file is put on top of the target file (sometimes of course, through a mask) and the result overwrites the target file we started with. Image - calculations remedies this and will be more familiar to users of previous versions of Photoshop.
Image - calculations provides a slightly different dialogue box where source 1 represents the lower file and source 2 represents the upper file in what might be considered the 'stacking order' of our blend. The blending settings are also identical to those before, together with the option of a third file/channel acting as a mask (checkbox). The main difference in this dialogue is the information we key in for result. This result can be either one of the component files or an entirely new file. The latter obviously keeps all our working components intact. We can additionally choose to make this result a channel of either an existing file on the list or the new file mentioned. If we do this however, it should be remembered that the result will be a greyscale image like channels themselves.
Both Image - Apply Image and Image - Calculations now offer a preview facility so we can see the effects as they happen. This is a major enhancement on previous versions of Photoshop and can be achieved by clicking on the preview checkbox in each of the dialogue boxes.
Image - flip allows us to reflect the entire canvas. This can be to the upside down position or the sideways position. This is useful for correcting scans that are already upside down by accident or scans that may have been made from transparencies facing the wrong way. This command is also used for modifying the direction of selections when they are active - preferably floating.
Image - rotate offers us three set rotation options for the whole file, together with arbitrary, which allows us to key in our own angle. Since all Photoshop screens are rectangular, the latter will obviously include areas of empty space and increase our file size. The space will be filled with our current background colour. If a selection was active, these commands would simply act to modify our selection and would include the extra option of free, allowing us to drag and rotate the selection's angle.
Image - effects is unavailable for flat images and works only for active selections. When a selection is active we can generate size and shape changes with four options. Scale gives us size-based distortions - holding the SHIFT key with scale acts to constrain the dimensions of the selection as we drag. Skew allows us to distort the shape by moving a single plane in either a left/right or up/down direction. The direction is dictated and subsequently locked by our first drag and we must repeat the procedure to change direction. Perspective allows us to move two adjacent handles on the selection on the opposite direction from each other. Once again, the direction is dictated by our first drag and the effect is one of generating distance at one end of the selection. Distort allows us to warp a selection manually with any one of the four individual handles.
Image - Canvas Size allows us to create new space around the existing image in which we can work. We simply type in a new measurement for the canvas itself which would be larger than the current size indicated at the top. A choice of measurements are available. The position of our current image in the new canvas can also be set off-centre (centred is the default) by clicking on one of the additional squares offered as options in the placement box.
Image - Image Size allows us to resize the image itself and may or may not generate an increase in canvas size at the same time. Our current size and resolution is indicated at the top, together with the present resolution. Sizing down our dimensions will normally result in an increase in the resolution. Sizing up will always result in loss of any real resolution since no solid information is available for new pixels. At best we could create artificial resolution by duplicating existing pixels (turning a single pixel into 4,9,16,25 etc adjacent pixels), providing the 'fabric' in which to introduce texture or some other compensatory disguise.
As we have discussed, resolution is all important to our final output, especially when this is print. The image size dialogue box therefore contains an auto button to calculate this based on the screen frequency our printer will be using. Clicking auto provides a field in which we enter this screen value and various quality options. We will then know the new resolution required for our output. The new value should then appear in the resolution box and we key in the required new dimensions. We can constrain the shape of our image to avoid distortion using the proportions checkbox. We can lock the current resolution to the scale by clicking on file size.
Note: When the 'screen' value is entered and 'best' quality chosen above, Photoshop doubles the figure to obtain the resolution value. This will be fine for output from Photoshop. Most designers/artworkers/production persons are actually more likely to be preparing files for export to another destination program for page make-up. As a result, it is best to create extra resolution that will allow for minor resizing and adjustments at that destination. A common print resolution based on actual size images would be 300 dpi.
Image - Histogram is a passive dialogue box which simply gives us a read out of values for the composite image or any chosen channel. The graph is a quick display of the values seen earlier in levels. A limited read out of values for a partial segment of the histogram is available by click-dragging over the relevant portion of the histogram.
Image - Trap is only available in CMYK mode and we will move to this now. We go to Mode - CMYK
CMYK mode regenerates the screen based on channels which contain the separations for the four colour printing process - cyan, magenta, yellow and black. If our file contains a significant amount of cyan or magenta, the difference will be most visible on our screens. The actual impression of the image we see also reflects the type of printing inks we chose earlier under preferences, since their attributes involved a certain figure for dot gain.
Dot gain occurs when an image is printed using those inks on a specified paper. The paper itself absorbs the ink and expands the dots of the printer's screen in the process. Sometimes this is marginal and sometimes it is very noticeable, That is why we may need to calibrate and increase the dot gain allowed for. This way Photoshop will compensate to decrease the amount of dot gain on the real print, making that noticeable expansion of dots less evident.
Colour images are made by overprinting these dots at different angles for each colour. The colour separations are also normally printed in a set order, the the lightest first and a black overprint last. Black subsequently masks out blemishes in the colour overlaps. Sometimes the information from the black separation is also superimposed on the cyan separation. The gives a bluish underprint behind the real black which improves that masking effect even more.
Unfortunately, these measures don't always work and there will be small 'cracks' in the reproduction where colours don't quite touch. A technique called trapping is used to compensate for this by spreading some colours outwards by an amount corresponding to the size of those gaps. There are two types of trapping - spreading and choking. Choking has the effect of locking the outlines on one separation in order the spread the outlines of another underneath. The facility to do this is not offered by Photoshop.
Image - Trap does however offer the facility to trap using the 'spread' method. Since the amount of spread required is often tiny, a width of one or two pixels may be all that is required. For more precise or awkward jobs, point sizes and millimetre measurements are also available. When we apply this we see a visible spread around the main colour transitions, although to witness this properly we would need to be in pixel-edit (16:1) mode. Applying trapping is not advisable unless we are specifically asked to do so.
Whilst in CMYK mode is is worth a further look at our channels palette. Window - palettes - show channels will bring this up if it is not already on screen. We look at the eye icons next to each channel and click on the one next to black. This will make the black separation invisible, allowing us to see and possibly work on the colour in isolation. The same can be done with the other channels. The facility to switch off a channel whilst keeping it visible has been removed from this version of Photoshop. Rendering a channel invisible now switches it off for working purposes too.
Before returning to RGB mode we will go Mode - LAB colour. This mode was introduced as a working mode by Photoshop relatively recently. It has always been used however as the intermediate stage between RGB and CMYK. It is at this point our separation and printing ink information is applied when Photoshop is converting automatically.
In Lab mode we look again at the channels palette and switch each channel on and off in turn. Then we specifically turn off the lightness channel. As in CMYK when we turned off black, all the greyscale information has been hidden. This is because lightness contains all the attributes we would normally consider to be black or grey.
The remaining 'a' and 'b' channels contain all our colour information. The display differs from what we saw in CMYK (minus black) only in that the lightness values of the colour are also absent. Painting and editing in Lab colour mode 'feels' intuitively like RGB. It is useful to know however that the grey/black values can be isolated. An example might be a pen and ink or pencil drawing that we have scanned in and wish to colour. We could safely do so knowing that the scanned information would remain isolated on the lightness channel as we applied pure colour.
We now move back to mode - RGB and then change again to mode - Indexed Colour. Indexed colour is a single channel 8-bit colour mode where our palette is restricted to 256 actual colours. The dialogue box defines how we will create those 256 colours or limit them still further by lowering our resolution. This mode has no real function when working for print, but it is used extensively for preparing work for videographics programs where the limited 8-bit colour is the norm. It is also likely to be used increasingly as publishers and studios take on multimedia work. The art used on interactive CDs is often saved in this format.
The Palette options contain 5 variants. Exact will use our existing colours if possible, but since there are more than 256 in the current file we can't use it here. System will re-map/re-draw the image using the 256 colours in the Apple system palette. Adaptive will allow Photoshop to build the colours for us based on the existing colours and is usually the best option. Custom allows us to create our own colours and previous will offer our previously used palette when we have one. We now choose adaptive.
There are three dither options. This affects the way extra colours are artificially created using the basic 256 we decide on. None will restrict us to the exact 256 colours. Pattern will allow the display to include tiny black and white pattern elements to darken the pure colours where required. This is similar to the effect sometimes seen on 256 colour restricted monitors. Diffusion gives a kind of blur-mix effect and creates the best appearance. We choose this.
On screen we now notice that some harsh colour transitions now look rather rough. This is unavoidable due to the palette restrictions and we would ideally modify the image to compensate. Blurring is not an option since most filters are unavailable in this mode. Many tools will not work either. Most work will be created in RGB before being converted to Indexed Colour.
We now go to Mode - Colour Table. Colour table is only available in indexed colour and is a facility rather than an actual mode. It offers us a way of changing our image by changing the colours used to make it. The default is custom, which shows the colours used in the file at present. We double click on one of them to change it and follow the procedure used when mixing foreground/background colours. We can change as many as we wish. Alternatively we can load a different palette through load or from the pull-down list, but this will seriously alter the look of our image in most cases.
The quickest way back to our original file is now file - revert. This returns us to RGB where we last saved it. We will now look at two new items in the mode menu.
Mode - CMYK Preview is a new facility that allows us the see the CMYK interpretation of our file on screen whilst still working in RGB mode. This will not stop us using unprintable colours or provide any other normal function of CMYK, but it does allow us to work with all functions available to us, whilst actually still seeing a replica of the printed image.
Mode - Gamut Warning provides highlights scattered all over our image that indicate non-printable colours. The colour used for these highlights is the one we chose earlier in preferences - gamut warning. If only minor areas are highlighted we are now able to use the sponge tool set to desaturate to brush over these areas and remove them. We do not need to remove everything perfectly since Photoshop will remove out-of-gamut discrepancies automatically when converting to CMYK. We can also monitor this correction by going to window - new window.
New window gives us a replica of our main window which allows for operations on a file to be monitored separately. We could for example be working in a highly magnified view with a second window showing the changes as they happen in a full image view. In the present case we change the new window to Mode - CMYK Preview as we did previously. We now watch the changes to our printing file as we sponge-correct the window still set to gamut warning.
This completes the general summary that hopefully provides an overview of Photoshop 3.0. A summary of principal filters is available elsewhere.