An advanced course for those concerned with the creative and editorial uses of Photoshop in a multimedia production environment.
Part One: Production & Design Essentials
The traditional areas of publishing/print and television/video have both found that separate, but linked, aspects of Photoshop have brought hitherto undreamt possibilities to graphic composition. Advances in communications technology and the economics of broadcast delivery have even led to examples of corporate merging of these otherwise different industries. The much heralded 'information super-highways' and consumer-based mass storage systems have further led to a converging of the creative talents involved in those fields. Creating products for publication as interactive CD ROM or publishing documents to be 'broadcast' through the internet involves the combining of creative discliplines from such divers areas as videographics to page editorial. Photoshop, with its many feet anchored right across this spectrum, is fast becoming even more indispensible.
We now examine its role in this new environment and discuss the basic techniques for approaching it as both a production and creative tool. Whilst Photoshop still confines itself to the limitations of the two dimensional canvas, its layer technology enables us to prepare images that can used in the 'building' of moving sequences. Its filter technology can also be used by other applications and we can import sequences of frames from filmstrip files. Needless-to-say, the things we can do to blend display components are virtually unlimited.
It is worth specifying some other software which we may mention in passing since they are rapidly becoming industry standards.
Macromedia Director is the 'page make-up' software leading the way for CD ROM development and other interactive projects. Its 'pages' are not those of print however. Pictures can be pasted in as still images but also as moving images - either live action video or animation. Text can appear on the page, but it can also scroll on and off the page, animate or distort itself. Any word of text or any picture can be 'hot-linked' to anywhere else in the multimedia document without the need for bookmarks or indexes. All this is accompanied by, if required, stereo digital surround sound - itself editable, often by the end-user, and 'linked' throughout the work. The only constraints are those of the still limited storage media for the work itself, together with the compression alogorithms used to encode its more complex parts. These documents are set to increase in size radically with the advent of both greater capacity CD storage and digital 'on-line' access to 'serving' facilities where storage of data will be largely unlimited.
Macromedia's Authorware Professional is the state-of-the-art 'elder brother' of Director. It has cross-over facilities with Director but essentially does the same things using a different interface. Director creates a 'time line' and a 'stage' whereupon different elements are mounted, frequently overlapping, but following a linear path. 'Hot links' and other interactive functions require scripting of language called 'lingo' which makes the connections that are 'invisible' on the time-stage itself. Authorware adopts a 'family tree' style of flow-chart instead of a time-line and is, in many ways, more intiutive with its 'icon-led' building system. There is also no basic need for the user to learn any form of programming language since these procedures are largely automated.
Adobe Premiere from the makers of Photoshop is the commonly-used video editing software for this new media. It supports Photoshop's filter library and Photoshop can access some of its files.
Macromedia Sound Edit montages and mixes multiple sound sources in much the same way that Photoshop does with images.
Macromedia Freehand and Adobe Illustrator are both the leading 'Postscript' illustration programs. Unlike the memory intensive and complex pixel calculations characteristic of Photoshop, these applications allow us to create graphics using lines, coloured fills, layering and automatic intersection commands for the objects. The process is fast and precision driven rather than natural. For printing purposes these images are converted to pixels by a dedicated image processor shortly prior to plate-making. This is done at high speed, based on the binary code describing the image. Given their artistic limitations, both these programs have complete synergy with Photoshop and can be used to enhance working, even in the non-'postscript' environment. This is particularly important given that the 'hard' onscreen outlines displayed by these programs are unsuitable for most video display.
Quark Xpress and Adobe Pagemaker are the leading 'page make-up' programs for print and again use the 'postscript' technology. Beyond the print arena, Xpress pages can be saved in the eps format and imported into Photoshop as images or image components - a particularly useful feature for backgrounds and colour templates. Pagemaker, on the other hand, allows for the authoring of pages that can be directly published on the worldwide web/internet by supporting both HTML and Adobe Acrobat formats. Its incorporation of Kodak's device-independent colour management system is another bonus since this allows for calibration across output for both print and video, with support for the PhotoCD standard.
Adobe Acrobat is the authoring tool for the Acrobat format designed to be usable to the end-user across platforms. It is also set to become at least a secondary standard for the publishing of Web distributed documents.
Ceneca's Pagemill and Hot Java from Sun Microsystems are both new authoring software for Internet/Web documents and include elements of interactivity. Pagemill sets the standard for creating HTML files with an intuitive graphical approach ('drag and drop') familiar to Macintosh users and Apple themselves consider that this will become the 'pagemaker' for web authoring. Hot Java boasts considerably more powerful features with programming facilities and security encoding built-in. However, some knowledge of programming is required by the user until the scheduled 'plug-in' function modules become available.
Adobe Dimensions is one of an array of applications which operate in three dimensions rather than two. Simple outlines from Illustrator or Freehand can be interpolated 180 degrees into 'wireframe' models whose surface can then be painted or textured with images from Photoshop. The end result can then be 'photographed' and brought back to Photoshop, or it can be exported in 3D for use elsewhere.
Fractal Design Painter is the closest relation to Photoshop itself with some features even Photoshop lacks. It can also create animation sequences of a limited nature. Its main characteristic is that it has been designed with the painter/illustrator (person) in mind. Textured surfaces are created prior to painting and the brush facilities are enormous. Uniquely, it reads Photoshop files (although at this time any layers will flatten) and can apply surface textures and other effects to them.
Kodak Create-It and Kodak Build-It are the standards for authoring using Kodak's PhotoCD format. Create-it is concerning with authoring presentations and incorporates other file formats. It uses basic interactive functions including sound but cannot support moving image elements. Build-it duplicates some of these functions but is largely concerned with the formatting and writing of the PhotoCD medium. See the notes on PhotoCD format shortly.
Letrastudio from Letraset is the principal text manipulation program and is renowned for the distortions it can achieve without losing typographical integrity. Text effects in Photoshop are limited in this regard so this application as a good complement. It can also access a vast number of type libraries and export for almost any destination program.
We also need to be aware of the following file formats.
JPEG (joint photographic experts group) is the quality standard for compressing image files and extensive control is offered to adjust the 'trade-off' between quality and file size reduction. Massive savings can be achieved with little or no loss in quality, whilst images can later be de-compressed and re-compressed again with no additional loss thereafter. Primarily used for still images, a video extesion, M-JPEG is now becoming more common.
MPEG (moving image experts group) is the standard for digital video compression. Absolutely massive in file saving, it remembers similarities in adjacent frames of a video sequence and records in detail only the significant changes. Encoding and decoding is time-consuming since it has to operate with a 'time-lapse' facility held in store - this is where it calculates/decodes the similarities. It also compresses sound files attached to video but only as an integrated part of the video sequence so they cannot be edited independantly.
Photo CD (Digital Negative) was devised by Kodak as a cross-platform multi-functional format for still images. At heart it is a high-speed scanning, colour calibration and bulk storage technology in one. The Kodak YCC colour model is not yet in the public domain but is largely derived from the LAB colour model used by Photoshop. The pixels of the 'lightness' or 'luminence' channel are treated with the sanctity one might expect from photo-imaging experts, whilst an extensive JPEG alogorithm is used on the two colour channels where little detail is present. The format embodies a master file and five or six simultaneous resolutions of the same image. Approximately 100 of these multi-resolution 'packs' (using the 5 basic resolutions) can be held on a single CD. 'Packs' which include a sixth 'super-high' resolution only manage about 30 images on a CD but are needed by the print industry for reproductions greater than magazine page size. The base format is mastered from 35mm negatives or transparencies, the extra resolution from larger film sizes or 5x4 transparencies. The higher resolution size of the basic 'pack' is considered the digital substitute for standard negatives and can be used, like negatives, for the developing of photographic prints. The 'pack' of five basic resolutions contains all the information needed to print the images up the magazine page size, with the lower four also used for videographics and television display. The master file (Kodak CMS PhotoCD) additionally contains information about the scanning device and the type of film the original photograph was shot on. Photoshop has Kodak's colour management system built into it (although the data is actually stored in the system folder) and allows us to open any of these resolutions through the master file information window - thus ensuring optimum quality for our output medium. A photoCD disc containing this standard information is known as a PhotoCD Master.
Portfolio PhotoCD. It is also possible to author self-running presentations (partly interactive) on the same disc format. These can include 'links' between images, index and cataloguing data, music, narrative and transition effects. These discs are known as Portfolio PhotoCDs and still contain the uncorrupted image 'pack' for normal use. Since interactive presentations do not actually make use of the higher resolution components, some commercial titles exclude them and can subsequently include hundreds more images on the disc - they take considerably less space and are still more than adequate for high definition television or monitor display. Any PhotoCD format encoded disc (including masters) can be read by virtually any CD drive currently on the market - including audio CD players if there is any sound component to the presentation. In addition to CD ROM drives (whether with Macintosh, Windows, MS DOS or other), the following players also support these discs: PhotoCD players, Philip's CD-I players, 3DO machines, Sony Playstation, Sega Saturn, Amiga players. Most offer the facility to zoom into images in detail which makes the format ideal for mulimedia titles based on quality photography or art.
PhotoCD has suffered from limitations. It is confined to the 35mm dimensional ratio regardless of resolution. Most images therefore need cropping. It is has also been a 'scanning only' format which, working at high speed, prevents image re-touching at the input stage. Good images can look excellant, but poorer ones exhibit all the flaws evident in the original negative. The master file format is shortly being released into the professional public domain and future versions of Photoshop will be able to write image 'packs' (including the CMS master component) directly. This will enable PhotoCDs to hold digitally-created works or otherwise 'corrected' images without the need for secondary scanning.
Note: A number of photographic CD ROM libraries are available which, not surprisingly, have been mastered using Kodak's PhotoCD format. This does not necessarily mean that the manufactured copies have been properly encoded to the same format. Although all titles can be accessed through a CD ROM drive, only those bearing the PhotoCD logo are certain to work on these other players.
The most fundamental level at which Photoshop can be seen to contribute to multimedia is in the production of backgrounds for the on-screen displays, so we will look at that first.
The opening 'page' of a multimedia project is a good candidate. Its colour range and textural feel needs to entice the television/monitor viewer. As with videographics (and to a certain extent, slide shows) the output medium is based on light rather than ink. For those used to working with print, this absence of ink attributes and the lower resolution file sizes will be a major change. As will be the approach to composition. On-screen backgrounds are almost always slanted toward darker or richer colours, although if the result is intended for broadcast there are colour gamut restrictions just as there are for process printing inks. Areas that are lighter therefore stand out better, so bright ranges tend to be used for text and created objects such as interactive 'buttons' which the viewer needs to be drawn toward. In the same way that good typography makes a reader feel 'comfortable' whether glancing at a page or examining it in detail, so the balance or 'feel' of a screen must be right. We must also consider the time period over which it will be displayed to determine whether detail or texture has the greater emphasis.
We already know that Photoshop opens and saves myriad file types, but it is likely that work for an interactive destination will involve using the PICT format. This may also be JPEG if the destination program supports decompression. The PhotoCD format capabilities may well supercede both when it becomes easily writable. The reason for this limitation is to do with comparitively small file size. The viewer does not want to watch the screen re-draw itself, so the image must be small enough to open inperceptably.
This is optimised further by our selective use of Photoshop's different modes. The PICT file format and JPEG variants can save images in RGB screen colours, or the mode known as Indexed colour. This latter will be of major importance since it creates a file containing one single channel for up to 256 individual colours. These colours behave as absolutes and can be likened to the 256 individual shades of grey we see in greyscale mode. Screen resolution is a relatively low one (usually 72dpi), although the canvas size is quite large so an RGB file may not be that small. Converting to a single channel mode reduces it by two-thirds. If the viewer is only going to see that particular image for a few brief seconds, this is more than enough. For images displayed over longer periods would would probably want to keep the full RGB range, although JPEG may be able to compress this further if the destination supports it.
The moving image is often used in multimedia, even if not to the extent of MPEG full screen video. Many motion-based items like quicktime video clips and animations are contained within their own window, possibly superimposed on a Photoshop screen. In designing these screens we will have to consider the area that such windows will obscure and avoid using them for essential detail. More importantly, some multimedia authoring software can create 'foreground' animations through its own toolbox. An example could be a cartoon character walking across our Photoshop background. Obviously the animation and the background are both 'live'at the same time, so it will be even more necessary to keep the Photoshop element to a managably small file size. A full RGB backdrop would effectively slow down the animation sequence except on a very powerful system.
Whilst working with knowledge of other elements in mind we can often use Photoshop's layers palette to build a working model for the 'stacking' of what may actually be exported as separate components. To demonstrate a practical example we will open a texture file at full screen video resolution. We then open a suitable logo to reside on the main image, select the background behind it, select - inverse and copy it to the clipboard. We then go window - palettes - show layers and then edit - paste layer. We OK the subsequent dialogue box. The logo comes into the picture and the palette shows layer one above the surface of the background image. The logo is now on a sheet of 'acetate' superimposed upon, but for working purposes, separate from the main image. We click the 'eye' in the far left column next to the layer's name and it becomes invisible. We click again to put it on. We now have the first two elements in our 'stacking order'. Although we will later have to dismantle this file to save the elements separately, the information will be aligned ready for the destination sequence.
Next we position the logo where we want it - maybe to the top right of the screen. We check the layer is highlighted in the palette and use the move tool to drag the pixels into position. If the logo is too large (or small) we go image - effects - scale and, holding the shift key to constrain the proportions, drag one of the corner points inward or outward. Photoshop gives us a preview of the effect and, if it isn't quite right, we can further adjust the handles. When it is right, we click the 'gavel' icon to confirm the operation. (The line-through-circle icon will cancel it altogether.)
We will assume there are four 'paths' in the destination project that the viewer can take from here. One will be to exit. Others might be to search, to see reference works or to watch demonstrations. We now need to locate or make four objects to montage onto the image which will act as graphical 'signposts' to these functions. When found, we select each object in turn, copy to the clipboard and edit - paste layer in the main picture. This will produce a total of four layers besides the background. We will edit them later if necessary.
To work on the background itself we must go to the layers palette and click on its name. We will have returned to it when it becomes highlighted. By doing so, the next layer we paste will appear directly above it in the stacking order - not at the top as before. This time we need to open an image which conveys the mood of or actual subject matter of the presentation itself. Our hypothetical project is going to be about a geographical location so we use location- style photography. Something like an exotic landscape, a city scene and some flamboyant people would be ideal for the exercise.
We locate these elements and select - all to get the complete images. We then follow the now familiar edit - paste layer sequence to dump them in our main image. These three layers will be blended and merged into one, but we will worry about that later. For the present we will just position and scale them into a good visual layout.
The final items for our working rough will be text. The chances are that when the viewer clicks on an object 'hot spot' a word will highlight telling them what the spot represents. For our purposes we use the words 'demos', 'notes' and 'find' - for demonstrations (movies, slide shows, etc), documents (text features, photo access, etc) and search respectively. The less characters in the word chosen the better for viewing on screen.
Next we click on the icon at the bottom left of the layers palette. This will create a new layer - a clean sheet of acetate to work on. In the toolbox we choose the type tool and click in position over the object representing demonstrations. We type 'demos' and choose a style/size. We confirm the box and the characters appear as 'floating' selections above our empty layer. If they are too big or small, we delete them and redo the text. Photoshop remembers our last entry so all we need is a new point size. When it looks right, we position it on top of the object.
On the same layer we repeat the same exercise with the words 'notes' and 'find', positioning again over the objects concerned. This usually looks best if all the words are centred left/right on the objects, but centred up/down to each other. We now have three text objects which will have been rendered in our current forground colour.
We may want to change this colour. If so, we click the magic wand on the background acetate area. This selects all but the letters. We then select - inverse to make a selection of the letters themselves. We can now mix a new colour and edit - fill with it.
Flat text doesn't look dynamic on screen so we create a drop shadow. To do this we simply drag the layer containing the text onto the bottom left icon we used to make it. This time it functions to make a copy of the layer. This is above the previous text, so we move down one layer and treat the underneath one as the shadow. Changing to the move tool we then drag this layer to a suitable offset position. We than alter the opacity or blending mode to subdue the shadow - or we can recolour as we did previously. Next we switch off all the other layers except the text and its shadow, then go merge layers in the sub-menu. This 'flattens' them into one, so both the text and shadow are now a combined layer.
The final touch will be to add some transparency or a blending mode to the merged layer itself. We try out combinations. In the next chapter we will return to all these layers and see how we can apply effects to various components. A good visual rough is all we need at present.
We must now look at the task of saving this file, or others, in a way in which we have access to both the entire canvas and some of the individual components.
First we make sure that the master copy is already saved in Photoshop 3 format, complete with layers. We now choose file - save as again. This time it is greyed-out for all other options. This is to prevent us flattening and losing our layers. We must go file - save a copy instead. This allows us to save the same file with the name 'copy' or a new name as a flattened composite. It will not however include the information on any layers we have made invisible.
First we switch off all layers except the background, then save as PICT calling it 'background A'. We then switch on layer one and repeat the save as PICT calling it 'background B'. Both of these are now separate images in RGB which will be available for import at our multimedia destination. To obtain a third and fourth variant we repeat the process again to make 'background C' which will include our logo layer and 'background D' which will also include the layer with our subject title on.
The remaining layers contain our objects and text information and are either simple in colour scheme or less important as graphics. They are elements we will probably have multiple uses for and which will often require a fast screen re-draw. For this reason we should change to indexed colour.
Any change of mode in Photoshop will force our layers to flatten, including indexed colour. Although we need to change modes, we also need the layer information so we must use a quick method to obtain completely new files. First we select - all and copy, then go file - new where the dialogue will automatically show us the measurements we just put on the clipboard. We confirm the dialogue box after checking that it is set to contents - white.
We click back on our main image and go to the layers containing our superimposed objects. We keep all three switched on but turn all other layers invisible. We then choose merge layers from the palette sub-menu. On the name of the single layer now created we simply press down our mouse button and drag it to the new empty file. The layer boundary will be visible so we can position it correctly centred. It is then safe to change mode - indexed colour, then save as a PICT. We could also use the same technique to save indexed colour PICT files of each individual object if necessary.
We now click back on the main image and repeat the exercise above with the layers containing the text overlay for the objects themselves. We switch on the text layers, switch off all others, choose merge layers, get another empty file - new (same settings) and drag the now-merged layer across. Then we change modes and save as like before.
Back in the main image we notice that the separate layers for each object and the text shadow have been lost. Since we have not actually saved any changes to this particular file we can now go file - revert to restore the original with all layers intact.
Part Two: Design Technique & Special FX
Photoshop has a vast number of built in effects and composition features which make quick work of creating new backgrounds, 'hot-link' buttons and special optical tricks. Many of these involve using the plug-in filter technology - some coming with Photoshop, others marketed independantly. A summary of the basic built-in filters appears elsewhere, but we will now look at particular ones which can be used to create good 'on-screen' effects.
Of those packaged with Photoshop, the filter categories of stylise, distort and pixelate are most commonly used for perspective and textural effects. The new render category includes some specific textures and the lighting effects.
The KPT series of plug-ins are purchased separately but provide power options for similar tasks. Paint Alchemy, another separate title, is a single plug-in interface that can turn photographic images into virually any natural medium required. KPT's Convolver provides an interface that does little new, but enables us to render almost any major filter effect from one single dialogue box. Another external plug-in, Squizz, does much the same for distortions using a single dialogue screen, but with the added advantage of making actual distortion brushes for us. Gallery Effects from Adobe is another natural (also some un-natural) effects library whilst the strangely titled Sucking Fish series can provide quick æbuttonsÆ artwork for mediamedia very rapidly. Specular's Texturescape offers custom control for creating any texture, while X-Textures provides a cut-price variation of the same.
Our present screens contain a number of objects which are planned to 'hot-link' to navigation commands in the destination program. We now need to create a number of other 'buttons' that will provide similar but more basic functions on every screen. In our example this is an 'exit' button, a 'forward' button and a 'back' button, although the two latter ones are most often simulated with the left and right arrows facing video symbols.
We create a new empty file by making a selection rectangle somewhere at the same size as the button required. We then copy this and go file - new using those measurements. We then fill the whole of this file with a basic colour for the button. Next we choose filter - other - SF Dekoboko (sometimes in the Sucking Fish category) and okay the defaults. A relief and shadow area is applied turning our solid colour into a botton shape. If we don't have this filter we could also try an emboss filter, with resulting greys colorised (via Image - Adjust -Hue/Saturation) and blended with a solid or a simple shape with a gradient fill.
We image -duplicate three copies of this button file and superimpose different text on each. We do this by picking a suitable text colour, clicking the type tool and typing the required word. It will look best if a consistent font is used for all basic buttons. We should also stay with the same point size and the easiest way of judging this is to do the longest word first. We then save each file.
Choosing the file with the 'back' button first, we select it and copy to the clipboard. Our first title screen will use 'exit' instead, but all our other screens will need 'back' in them somewhere. Once again, unless you actually want to disguise the viewer's escape route, consistancy in positioning for common buttons is a good idea. Lower left of each screen is the prefered place. We always ensure to edit - paste layer since these must not be stuck into the background.
We then copy the 'forward' button and repeat the exercise, this time the convention is lower right in the appropriate files. Although we do this for practice and as a method of 'proofing' our display, we will usually import these button files separately at the destination program. We would also covert to and save in Indexed Colour mode as before since they are simple colour schemes. We complete complete the simulation exercise by pasting the 'exit' button on a layer in the title page file, positioning it as with 'back' buttons.
We are now ready to create three more master screens which will be those activated in the destination program by the user clicking on one of our three graphical 'hot-spots' on the title screen. Each of these 'pages' will obviously have to covey the feel and information requirements of the section chosen.
The demonstrations section will need an interface that allows the viewer to select from movies, art galleries, photographic slide shows and possibly voice or music recordings. The search section will undoubtably be a glorified database with attractive graphics simply to cover up an otherwise boring index. The reference section may well be something of a hybrid. It will be the route to further documentation and willalso need some kind of general index. The documents themselves may include routes back to the demonstrations section or alternative ways of accessing the data used in the demonstrations. It is almost certain that all three screens will have more limited graphical display areas than our title screen, since the emphasis is now on information. To create them, we will have to rely more of texture and impressionistic ambience.
We choose a basic texture which will be common to all screens, although we will alter it subtly on each to give some variety. We copy the entire chosen texture file to the clipboard and paste it into a new file. We can use the measurements of our title screen for the template. Next we make three duplcates and save them each with different names.
We then return to the title screen and drag the first of our object layers into the first new canvas. This puts the layer on top of our texture. We repeat the exercise for both the second and third object layers in the title screen file, moving them to our second and third new canvases respectively. Lastly, we go to each new layer in each of the new files in turn, scale the object to 'icon size' and re-position it at the top left of each screen. We now have an identifying mark to show the eventual viewer which graphic he/she pressed to get here.
In one of these files (it matters little which) we click back on the background layer, then go image - adjust - colour balance. This allows us to manipulate the feel of the texture for this particular screen. In the dialogue box we click the 'preview' button to see the results before confirming them. Rather than do so, we cancel and go image - adjust - hue/saturation where a similar facility is available. If the texture is monotone or dull, using colorise will apply the 12 o'clock hue from the colour wheel, which can then be changed or de-saturated at will. Again we cancel this.
Still experimenting, we try similar test variations using image - adjust then levels, curves and brightness/contrast. This shows the main possibilities. Lastly, we select - all and copy (the background layer), then paste layer. This puts an identical layer on top of itself. On the new layer copy, we can choose one of the adjustments explored above and then use that layer's own opacity and mode controls to blend it back into the original underneath.
To create a picture-brush element for added extra detail we select an easily defined component in the background layer, select - float it and edit - fill with contents - black. Next we open the brushes palette and define brush in the sub-menu. Back in the picture we press delete to remove the black selection. We then choose the rubber stamp tool in the toolbox and set it to non-aligned clone, before picking the actual brush we just made and positioning it exactly on top of the area we made it from. Finally we hold the option key and click whilst centred on that same position to make a clone. When we paint, this tool will now give us an exact replica of that item. We double-click on the brush itself in the palette and re-set the spacing to 100 or more. This ensures a space between repeats of the image. We can now paint extra copies of this particular component as a method of enhancing the texture.
Note: Painting, whether this or any other way, can be calculated in advance - a relief for those who consider themselves less than proficient with the handwork (mousework) side of the art. We open the paths palette and use the pen tool to create an 'open-ended' directional line which we will force the brush to follow. We check we're on the same tool and drag the path's name to the 'second-from-left' icon at the bottom of the palette. Photoshop renders the stroke for us following the contours of the path but with the attributes and options set for the painting tool. Since there is no release of the mouse button involved, patterns and clones set to non-aligned are unable to use this facility since the resulting behaviour will be as if they were aligned.
Another approach to textural image adjustment is to use Photoshop's print technology to convert an RGB into process colour. We do this by opening a RGB texture file, then choosing mode CMYK. This will covert the image to process colour inks which contain a highly restricted colour gamut. We can choose CMYK preview to see this in advance. Any effect obtained this way will remain evident when we convert back to RGB and can sometimes be a way of achieving greater graphical contrast between foreground elements and the background. The effect must be done before the foreground elements are introduced however, preferably in a separate file.
When Photoshop moves into CMYK mode it also applies precision values to the printing inks based on the type of inks and paper stock used for output. There are additional variations for the age and type of printing press. Although these functions are indispensible in print production, they are redundant in multimedia and video - despite the fact we may want the process inks effect. To render them neutral we go file - preferences - printing inks setup and enter a dot gain of zero. This will reduce the colour gamut, but prevent any spread of the colours involved.
In the same texture file we investigate some more filters. Most effects can be seen in the preview window and then cancelled. Others may need to be undone before trying out the next one. Many filters need a selection active to work, but select - all for the whole screen counts.
The distort section of filters are the closest to real special effects. The image can be given a ripple or wave effects to warp it as though viewed from underwater. Pinch pulls at the surface the image is resting on whilst spherise blows out or contracts the surface from a fixed central position with the distortion occuring in a radial dimension. Twirl spirals the pixels around a central point which remains largely unaffected, to create a blurred kind of vortex, while zigzag totally fragments the image like hard-edged waves. Others are more radical still.
The pixellate section alters the whole look of the image by re-organising the existing pixels into different kinds of grouping structures, frequently 'clumping' them together to create particular 'look' for the image.
The stylise section also contains filters that give a totally new 'look' to the image. Most of these actually lose almost all pixel information except where edges or significant colour/lightness transitions occur. Effects like find edges, which creates a kind of coloured line drawing from the image, are occasionally desirable in their own right, but we use them more frequently on layers containing copies that will ultimately be blended back into the 'real' picture underneath.
Having two stylistic variations on a background can enliven the final display in the destination program. Mediamedia Director, for example, allows us to have one version visible for a few seconds, then a gradual fade or dissolve into the second. All this independant of other components that may have been imported separately in the foreground. The impression is of movement without any real animation or video, as sometimes seen in slide-making programs.
Note: It is also possible to send a still video image directly from Photoshop through an appropriate video card to a video mixing system and utilising any one extra stencil mask as a video alpha channel. Imagine we have a landscape and centred within it is an arch. We make a selection of everything except the gap within the arch and save it to a channel. In the channels palette sub-menu, we choose video alpha and then this same channel. The normal 'eye' icon is replaced with a 'television' icon and the outgoing signal for this image now has a 'hole' within the arch itself. Live action film can then be superimposed in that space. The resulting video footage might then be imported into a multimedia destination program and in turn superimposed on the original photoshop image, possibly with more items in the foreground.
Note: It is inadvisable to leave 'holes' in Photoshop files even if we know that video footage or other items will ultimately cover that space in the destination programs. Pauses between start and stop points can easily reveal discrepancies in the underlying images but, more importantly, the video footage may be prepared against a blue matte screen so that the action can be 'dropped' directly onto whatever image lies behind it.
For the remaining exercises will will return to our various screens and enhance the components. The main methods involve feathering and layer-masking, also using blur effects and fade-outs for particular edges. We will also look again at the blending modes and layer options. To start, we return to the first title screen where we still have unedited background components and graphical 'signpost' layers.
To be continued...
COMING SOON... tarting up images /using layer masks/feathering, modes /creating secondary pages to follow each hotspot /using filter effects and textures /cd rom book standards and formats /index colour-cycling and a whole lot more!