Leeds LS2 9JT Telephone +44 113 233 3112 Fax +44 113 233 3167
A section showing the optic chiasma from the Visible Human Project
This is an experimental seminar, which will be modified in the light of audience feedback. We therefore hope that participants will complete one of the anonymous feedback forms, and leave it either with Dr Illingworth or with the General Office staff in the Garstang Building.
Computer illustration is one of those "transferable skills" that people often acquire at short notice when preparing a report. The fastest way to learn this material is to have your own pressing personal task to complete, so if you want to use your own materials for the practical work then you are welcome to do so. Alternatively I have provided some out of copyright images to download from the website that you can use for practice. This session will get you started and show what is available for free from the Leeds University network.
We will look very briefly at the different types of image file, and their pros and cons for different applications. In many cases, particularly for computer animations and writing for the web, it is important to achieve the highest possible resolution in the smallest possible file.
Practical work will be based on "Paint Shop Pro" for illustration and "Omni Page Pro" for text recognition. We will clean up some images and tables scanned from books, controlling colour depth and removing background marks, convert the scanned tables back into editable text, and paste the results back into "Word" and "Excel" documents.
Then we will look at creating new images from scratch, including line drawing, pre-set shapes, flood fills, gradient fills, cloning brush, adding and rotating text. We will do some simple image arithmetic and create a short computer animation.
Two related problems that frequently complicate academic writing are the need to avoid plagiarism - passing off the work of others as your own contribution, and the requirements of the Copyright Designs and Patents Act 1988, which are designed to protect the interests of authors and the economic basis for the publishing industry. Copyright normally exists for seventy years after the death of the author, or twenty five years after each edition of a published book. Personal copies of a limited part of a complete work taken for the purposes of research or private study do not infringe copyright: it is their publication that creates the offence. If you want to re-publish material that is subject to copyright then you need the permission of the copyright owner - usually (but not always) the previous publisher.
Neither issue will arise if your work differs sufficiently from the earlier material to make it a new contribution in its own right. If you paraphrase an article and render it in your own words that is neither plagiarism nor a copyright infringement, but if you were to translate it from English into French, then copyright would be infringed. Photographs are subject to copyright, but you can take your own virtually identical picture of the same original scene without infringing the previous copyright. Maps and drawings can be difficult, since there might be only limited ways to accurately depict the object in question. If you base your drawings on previous work that is still subject to copyright, then you must make a substantial input of your own, that goes beyond merely cosmetic alterations.
Copyright is regulated by international treaties, so the rules in most countries are similar, but not identical. Publishers often take a fairly vigorous line on copyright, because they depend on it for survival, but this enthusiasm is tempered because their own publications are frequently based on the work of others. Permission to copy is usually forthcoming for academic work, but there might be a fee. See the University Library Copyright Pages for a balanced view, and the US Copyright Office website at the Library of Congress for a description of the American legislation. The draconian remedies for copyright infringements are often moderated for practical and economic reasons: most electronic circuit diagrams appear to infringe somebody's copyright, but this is rarely enforced because (a) the original copyright owner may be unknown and (b) the electronics industry could not survive without such copies.
Useful images are often printed on very thin paper, which may allow text or images to show through from the other side of the page. The best way to avoid this is to place a sheet of matte black paper directly behind the required image, so that it is close contact with the unwanted material. A blank photocopy made with the cover left open is ideal for this. Subsequent scan quality will be greatly improved.
Microsoft office "document imaging" provides some very basic text recognition software, but at present there is only one serious contender in the market place. ScanSoft "OmniPage Professional" is so good that you will wonder why you ever used anything else. It costs over £400 for a full price copy of the latest fully-featured version. Fortunately the University has a licence and this program is available from the desktop of all ISS computers that are attached to a scanner. This is the only software that I am aware of that routinely handles complex tables and multi-column text, and saves them directly into "Word" or "Excel" documents fully formatted with no mistakes at all.
Scanned or computer-generated images can be stored in a wide variety of file types, some of which are preferred for particular applications. Very detailed image files may be inconveniently large. Most images can be stored in most formats, but it is sometimes important to use a particular format to maintain image quality while minimising the size of the resulting files. Some graphic file types are specific to particular manufacturers and are best avoided. Fortunately most drawing software includes a file conversion facility, and Paint Shop Pro is particularly good at this. Some popular file formats are tabulated below:
|Tagged Image Format||.tif||raster||optional||none||unlimited||efficient||widely used|
|Graphic Interchange Format||.gif||raster||yes||none||2, 16 or 256||small||diagrams, websites|
|Joint Photographic Experts Group||.jpg||raster||adjustable||adjustable||unlimited||efficient||photos, websites|
|AutoCAD native format||.dwg||vector||inherent||none||enough||efficient||engineering drawings|
|Protein data bank||.pdb||vector||inherent||none||enough||efficient||molecular structures|
The image on the computer screen is built up from thousands of individual pixels, normally with 256 individually adjustable levels of red, green and blue (RGB), or alternatively for greyscale images, 256 shades of grey. High quality kit can do better than this, but you are unlikely to be using it if you are attending this course. Paper images are represented in a similar way, but the image information is stored as cyan, magenta, yellow and black (CMYK) to match the printing inks. It is easy to convert from one system to the other, and you may never realise that this is being done.
Raster images reflect the pattern of coloured pixels on the computer screen. This imposes a fixed size and resolution on each image, which may not "scale" very well - an image that looks good on a website is unlikely to look so attractive when printed. Conversely, a good printed image will be much too large to view on screen. Good quality images on paper require much higher resolutions (at least 300 dots per inch) than computer screens (often 72 dots per inch). You can view very large images at a lower resolution, but fine detail will be lost in a random, quirky fashion and the results may not look attractive.
All images except bitmaps may be compressed. This can save a huge amount of space, because large areas containing a single colour need only store that colour once, with pointers to where it should be placed. TIFs offer a choice of compression systems and image representation. The compression algorithms for TIFs and GIFs are loss-free, so the uncompressed images are a faithful reconstruction of the originals. JPEG compression is mainly designed for photographs. It is inherently lossy, but the deterioration can be controlled. This means that if you keep saving and reloading a JPEG image the quality will get gradually worse, often quite noticeably so if the compression factor is set to "high". Paint Shop Pro remembers your previous compression setting by default, so you can ruin a high quality image by carelessly saving it when the previous image was highly compressed.
GIFs and (optionally) TIFs used a "paletised" representation. The software prepares a list of commonly used colours within the image, and stores them separately as a colour palette. There is an option of 2, 16 or 256 colours in the palette. This allows very highly compressed images with a faithful reconstruction of the original colours provided that there are not too many of them. It is the ideal system for cartoons and geometric diagrams, and it is a preferred format for the web. GIF images may also be animated, and browsers know how to play them by default. It doesn't work very well for photographs, which tend to have a much larger tonal range. The number of available colours is referred to as the "colour depth".
Of the various raster formats, JPEGs scale much better than the others, and GIFs are the worst because of their limited tonal range. The only way to reduce the size of a GIF is to omit some pixels completely, whereas JPEGs can handle half-tones and combine information from several pixels, thereby achieving a smoother result. If you need to reduce a raster image resolution, especially GIFs containing detailed text and graphics, it may be worth converting and saving them temporarily as high resolution JPEG images. Close the window and reload each image to force the format conversion, then reduce the image size and resolution before converting back to a GIF. One can achieve a similar effect by changing the image properties manually, increasing the colour depth and applying filters, but the JPEG method is quick.
The alternative is to store the image as a vector representation. This defines a series of coloured objects and their placement within the view. Vector images print and scale beautifully, and are used for engineering drawings, macromolecular structures and especially for maps. The image file commonly has a database structure and may consist of plain text. It is easy to attach additional information to particular points or areas within the image.
Click here to download a map for you to colour in.
We have based this tutorial around six scanned images of the human ear taken from two elderly anatomy text books, that are themselves out of copyright. Five images are from the tenth edition of Quain's Anatomy, edited by Schafer & Thane and published by Longmans in 1894, and the sixth is from A Text-Book of Anatomy, edited by FH Gerish and published by Henry Kimpton in 1899. We will study how to clean-up and extensively modify these starting materials, thereby creating your own personal presentations that are free from copyright restrictions and any accusation of plagiarism.
You can download some of these original images by right-clicking one or more of the blue links in the following table, and saving them to temporary files on your local D:\ or C:\temp drive while working on them in the class. Note the increased file sizes for the higher resolution 300 dpi images. There is no need for everybody to work on the same image, since similar operations are required in every case. Resist any proposals from your web browser to open them as pictures. You must eventually copy these files to your network drive if you wish to keep them, but please don't try to edit them on a network drive because it will be very slow.
|image||description||150 dpi||300 dpi||preview image|
The six pairs of images above show the human auditory apparatus at increasing magnifications. The first pair shows the external ear, the auditory passage (meatus), the eardrum and the middle ear, or tympanum. The air-filled cavity of the tympanum connects with the throat via the Eustachian tube. The middle ear contains three tiny bony ossicles (the malleus, incus and stapes) which efficiently transmit the vibrations from the ear drum to the sound detectors inside the fluid-filled cochlea.
The second pair shows the middle ear in greater detail, and the third shows the overall arrangement of the spiral cochlea in cross section. The interior of the cochlea is divided by the basilar membrane and Reissner's membrane into three spaces: the scala vestibuli, the scala tympani and the scala media, also known as the cochlea canal. The scala vestibuli and the scala tympani communicate at the apex of the cochlea, but the scala media is a blind-ended sack. The scala tympani and the scala vestibuli are filled with sodium-rich perilymph, which is very similar to cerebrospinal fluid, whereas the scala media is filled with endolymph containing a high concentration of potassium ions. Cochlea means "snail shell" and scala means "staircase".
Sound vibrations are detected by the inner hair cells in the Organ of Corti, which is a complex spiral structure attached to the basilar membrane. High notes vibrate the basilar membrane most vigorously near the base of the cochlea, whereas low notes (unexpectedly) cause the greatest movements near the apex. These movements bend the actin-stiffened "hairs" or stereocilia which are normally embedded in the tectorial membrane above the Organ of Corti. This distortion opens and closes ion channels in the plasmalemma, and allows the potassium rich endolymph within the scala media to modulate the hair cell membrane potential in both directions. These potential changes control neurotransmitter release from the opposite end of the hair cells, and initiate the signalling process.
In addition to the inner hair cells which are the sensory transducers, the cochlea also contains a larger number of outer hair cells whose function seems analogous to a graphic equaliser on a hi-fi system, except that they produce a very uneven frequency response. All hair cells are equipped with myosin-driven adaptation motors that adjust their sensitivity in real time. The outer cells can actively vibrate and locally modify the properties of the basilar membrane, thereby adjusting the sensitivity of the detection system at each frequency to emphasise those sound patterns with the greatest biological importance. When people speak of "straining their ears" they are doing exactly that.
The evolutionary record shows that this bizarre arrangement evolved from a modified gill arch and has subsequently been in continuous use for about 400 million years. It can detect movements of atomic dimensions and performs real-time spatial and frequency analysis that compares favourably with modern computers. Hair cell signalling is among the fastest biological processes and it is claimed that the echolocation system of the big brown bat can distinguish timing differences of 10 nanoseconds. If you would like to read more about this remarkable subject, we recommend recent review by Chan & Hudspeth (2005), which Leeds students can read online.
Paint Shop is a low-cost graphics package. It has many competitors providing similar functionality, but we find it convenient to use. More expensive programs are available with enhanced performance, but in such cases the learning curve is likely to be much longer.
Fire up Paint Shop from the START button on your desktop (it is stored in the graphics directory) and examine the image(s) that you have chosen. The open command is on the PSP "File" drop-down menu, or use the standard Windows file open button along the top of the screen. You will need to navigate to the file that you just saved in the previous section and specify the file type to be either "all files" or type "tif".
The .tif suffix to the file names indicates that files you have downloaded are in Tagged Image Format, or TIF. This is closely related to the international standard used by fax machines. TIF is one of the most popular formats for "raster graphics" on small computer systems, because it can be compressed so that large blocks of uniform colour can be defined in a few bits of information, without having to send identical details for every part. Raster images are built from thousands of individual coloured pixels arranged into hundreds of narrow stripes across the computer screen, like a television picture. They are difficult to re-size and work best for relatively low-resolution pictures where the details are uniformly distributed over the whole image. The alternative system of "vector graphics" draws each picture like an artist with a pen. Such images are easy to re-size, and work best for high resolution work, containing areas of fine detail interspersed with featureless expanses. Some MS Windows programs (including Word and PowerPoint) can use a kind of half-way stage called an enhanced format metafile, which might contain a mixture of bitmap images and vector graphics. Recent versions of Paint Shop can also keep raster and vector information in separate image layers. Use of these techniques can provide device-independence and works particularly well with images containing graphs and text.
Paint Shop can convert your files into other raster graphics formats, using the "Save As" command on the "File" menu. Vector to raster conversion is also easy, but it is very difficlt to go the other way. The most popular file formats include Compuserve Graphic Image Format (GIF) and Joint Picture Expert Group (JPEG) both of which are extensively used for websites. GIF files contain an internal colour palette with up to 256 colours. GIFs are best for diagrams and line drawings, whereas JPEGs can have millions of colours and are best for colour photographs. JPEGs "scale" better than GIFs when the size of the image is altered and JPEGs can also be saved in a degraded form which loses some of the fine detail but economises greatly on storage space.
It is impossible in this brief tutorial to exercise all the functions that are available from this drawing program. Instead we will demonstrate a small number of routine tasks, and encourage you to explore the other facilities in your own time. A comprehensive set of on-line manuals is available through the "Help" menu. If you need on-line help with specific operations click first the query icon at the top of the screen, then click the icon that you are trying to use.
The general approach in PSP and many similar drawing programs is to select the appropriate tool from a toolbar alongside the image and then apply it to the relevant part of the picture. It is easy to adjust the effects produced by many of these tools (for example, changing the brush width) and to control their area of application to the image. The main tools available in PSP are shown on the drawing below:
We will examine the following techniques:
The magnifying glass tool adjusts the size of the image. First select the tool: then left click in the image to enlarge it, right click to shrink. The redrawn image is centered on the place where you clicked. In addition you can control the image size from the "View" menu. The "Normal View" command is particularly useful for returning an image quickly to its natural size. You can also use the "View" menu to switch the grid on and off. This is useful for aligning text and graphics accurately. The grid controls are well hidden, but you can adjust the grid spacing and appearance from the "File: Preferences: General Preferences: Rulers & Units" tab.
The finished size of a drawing depends on the number of pixels, and on the pixel spacing, which varies with the display medium. Most office printers currently work around 300 dots per inch, but the normal resolution for internet graphics is only 75 dots per inch, which explains why many website images look disappointing when printed. You can change the orientation and the overall dimensions of your image from the "Image" menu. Flip, mirror and right angle rotations map one pixel to another without loss of image quality, but the free rotation command involves recalculating new pixel values from the previous set, and this will slowly degrade the image if applied a large number of times.
The "add borders" and "canvas size" commands add extra pixels to the edges of the image, without changing the original resolution. "Resize" adds or removes pixels within the image, recalculating the new values from the adjacent pixels to achieve a smooth result. Note that re-sizing does not add any new information to the image, so it will not improve the ultimate sharpness or resolution unless you insert additional details at a later stage.
Many scanned images have unsatisfactory backgrounds and contrast and you can often improve these by using the facilities provided on the "Colors: Adjust: Brightness/Contrast" menu. If you adjust the brightness and contrast carefully it is possible to convert a dirty off-colour background into maximum white without losing any picture detail. There are many other adjustments available on the "Colors" menu. It is quicker to experiment with these controls on actual images than to describe their effects. For most purposes it is sufficient to adjust all three colours together, but if necessary you can process the red, green and blue components separately to remove a colour cast.
The "stretch" command on the "Colors: Histogram Functions" tab will adjust the image contrast to exploit the full dynamic range of the medium. It can be applied more than once if you want to disregard a few pixels at each end of the range, but the effects become progressively smaller on repeated use.
The last few commands on the "Colors" menu allow you to count the number of unique colours in your image, and to increase or decrease the "colour depth". Detailed colour information occupies a lot of storage space, but it is often possible to use fewer colours without significant effects on image quality. Website authors in particular are always trying to reduce the size of image files, because the smaller files will load much more quickly. Reducing the number of colours might even make an improvement if the scanner introduced random variations into an area of uniform colour fill.
Careful adjustment of brightness, contrast and colour depth may allow you to restore a scanned line drawing to just two colours: black and white, like the original. Having done this you can subsequently increase the colour depth to (say) sixteen colours which will allow you to add new information and colour it in. Always adjust the contrast first to produce a harsh, over-contrasty image before decreasing the colour depth. When you subsequently increase the number of colours, all your new colours are initially set to black. You must edit the palette (see below) before you can work with them.
JPEG images are always stored in full colour even when the subject is black and white, but other formats can economise on storage space by selecting colours from a restricted palette which is stored with the image. This usually means that the colour information requires only 4 or 8 bits per pixel instead of 24 or 32. You can change or edit these palettes from the "Colors" menu, in which case all the pixels with the amended colour value will change together, while their neighbours are unaffected. Note that it is impossible to introduce new colours into an 8-bit (256 level) greyscale image: all the palette positions are occupied and you must first of all change the colour depth (using the "Colors" menu) to obtain a new palette.
There are several ways to select new colours for subsequent drawing operations: (1) The easiest way is to position the mouse pointer over the multicoloured panel towards the upper right of the screen. It will change into the colour dropper symbol, and you can load new foreground or background colours with the dropper by clicking with the left or right mouse buttons. (2) You can manually select the colour dropper from the toolbar and load an existing colour from the image itself. (3) Alternatively, click the appropriate coloured square in the "active colors" panel, below the multicoloured box to the right of your screen:
The subsequent dialogues are slightly different, depending on whether you are editing a full colour JPEG image, or a paletted GIF file. In the JPEG case you can either choose one of the pre-formatted colours from the "Basic colors" matrix on the upper left of the dialog box, or you could select (with full control of hue and saturation) from the colour wheel on the right. GIF files are restricted to the colours in the palette. You can alter these by cancelling the "Load color from palette" dialog box and instead clicking "Colors: Edit palette" at the top of the screen. When the new dialog box is displayed, double click one of the existing colour squares to alter it. The above instructions apply to Paint Shop Pro 6: other versions of this software may not be completely identical.
There are various methods to cut and paste image information from one picture area to another, or even between different images. The simplest method is to mark the required area by dragging over it with the selection tool and then use "edit: copy" or Ctrl-C to make a copy on the clipboard. The selection tool is the sixth one down on the left of your screen.
There are at least five different ways to paste your data back again, and all of them are useful. The two methods you will need most often are to paste as a new selection, or as a transparent selection. Their effects are subtly different: a new selection over-writes the original image, but a transparent selection allows parts of the original image to show through. In either case you can slide the selection around with the mouse, until you fix its position with a right or left click.
It is also possible to move a selection one pixel at a time by holding down the "Ctrl" or "shift" keys while using the cursor movement keys. This can achieve much greater precision than is possible with the mouse. You can use both techniques on the same selection.
You may sometimes encounter a situation where the computer seems to have gone on strike. There are no error messages, but most of the drawing tools are "greyed-out" and the machine refuses to obey your instructions. There are two likely explanations for this irritating development: EITHER you have accidentally selected a vector layer (see below) most probably while entering text, OR you have left a tiny, hardly visible selection box somewhere on the image, and the computer is waiting for your instructions about this box. It is easy to create these boxes by accidently clicking the mouse in the image when going for your next tool. In the first case you should check the "Layers" menu to see which layer you are in, and in the second case you should pick the selection tool and right click in the image to cancel any open selection box.
At any one time the current "foreground" and "background" colours are displayed in the active colours panel on the right of your screen. If you select any of the subsequent drawing tools, your left mouse button applies the foreground colour and the right mouse button applies the background colour. You can swop these colours over, or select new colours from the palette, by clicking the active colours panel.
Before drawing you must select the appropriate tool, and adjust its width. The paintbrush allows freehand application of individual pixels (or much larger areas) while the line and shape tools allow you to insert straight or curved lines, or geometric figures. Eraser air-brush and flood-fill effects are also available. The size (and other properties) of each tool are controlled through the "tool options" window, using an icon at the top of your screen.
You can toggle the tool options icon on and off. The window automatically rolls up when not in use, but if you select your tool and point at the icon it will expand to show the services available.
The quickest way to master drawing is to do it. Try selecting each drawing tool in turn, changing the tool options, and apply some colour to the screen. Use the magnifying glass to expand the image so that you can see the individual pixels, then shrink it down to judge the overall effect. Draw some straight and curved lines, filled and empty geometric figures and freehand constructions in various colours to see what the program can do.
The colour replacer tool allows you to systematically replace one colour by another in a defined area of the screen. The affected area is controlled by the tool dimensions and may be further restricted by pre-selecting an area with the selection tool. This can be used to omit an area, as well as to include it, depending on which side of the boundary you click with the mouse. You can adjust the tolerance for the replacer tool, so that it will replace close matches but ignore more distant ones. The retouching tool (immediately below the colour replacer tool) can be used in similar, but more sophisticated ways. See the program "Help" menu for detailed instructions on this.
The use of the spray can is fairly obvious, but flood-fill requires more explanation. This tool automatically fills a screen area with either the foreground or background colour (left and right mouse buttons again) until it encounters a boundary. Image defects on scanned images can result in some unexpected effects, where the colour spills through a broken line to cover a much larger area than you intended, or the line you are colouring has an unintended bridge to another part of the picture. If this happens then you must inspect the image at high resolution to find the offending defects, and then zap them with the paintbrush. An alternative method is to draw a line across the image to make an artificial boundary for the flood fill effect.
The text tool allows you to superimpose text on your drawing in a wide variety of colours, sizes and fonts. Although it possible to add lettering very neatly in PSP, the results may not scale very well if the image dimensions are subsequently altered. The effect can look amateurish unless the picture is always displayed at its original size. For this reason many authors prefer to add their lettering as their final step, using transparent text boxes after pasting the image into PowerPoint or Word.
Recent versions of Paint Shop have overcome some of these difficulties by adding text to a separate vector layer, which overlays the background bit-map image. Such text is freely mobile and re-sizable, in the way that may already be familiar from drawings in Microsoft Word. Although you can easily save and recover these images in the native Paint Shop Pro format, other file formats may not support multiple image layers. You must then collapse your image to a single layer before saving it, after which the old re-sizing problems will re-appear.
It is inevitable that you will make numerous mistakes. You can undo your most recent action using the edit menu, and you can also review you command history and backtrack to an earlier point. You cannot remove individual items from the command history without also deleting the subsequent events, which might depend on what you previously have done. You can preserve the current state of your drawing by saving it to a file.
A simple "Save" keeps the image in its existing format, but "Save As" allows you to change the format of the file. Numerous alternatives are possible from a pull down menu. The most useful options are .gif and .jpg for use in websites. When saving for the first time in the JPEG format, make sure to click the "options" button in the dialog box so that you can set the compression factor explicitly. It can be very irritating to accidently save your delicate artwork in a highly compressed format, because once you have closed down your editing session you will never get the details back again.
The GIF standard has always included animations, and your browser automatically knows how to run them. Other proprietary formats such as Macromedia "Flash" may be easier to use, but require users to download and install a plug-in. Alternatively, you can create animations using Microsoft PowerPoint.
There is an extension of Paint Shop Pro called "Animation Shop" which is also available on the Leeds University desktop. It is easy to exchange information between these applications to create your own animated files.
This program forms part of the MS Office suite and is used to create both manual and automatic slide presentations, with associated speaker's notes and handouts. The package includes a rich variety of pre-designed slide templates that greatly simplify the task of creating a complete presentation. As was the case with PSP, is not possible in this brief tutorial to exercise the full range of functions that are available from this program. We will demonstrate merely how to paste an image onto blank slide and add some lettering to it, but we encourage you to explore the other facilities in your own time. A comprehensive set of on-line manuals is available through the PowerPoint "Help" menu.
Fire up PowerPoint from the Windows NT desktop. The program starts with a dialogue box, and in this case you should select "Blank Presentation". You must now choose a slide layout appropriate to the image you have already created with PSP. Scroll through the alternatives to see what is available. We suggest perhaps using "object plus title" to start with, but you might prefer a different layout, perhaps a slide title plus an image below with bullet points alongside.
Assuming you picked the last alternative, you can edit each of the three boxes in turn by clicking in them, and re-sizing if necessary. Type a title into the title bar, add your bullet points to the text box, and insert your previously created image file into the picture area. You can align text-boxes and graphic objects using the guide lines, found on the "View" menu. Make sure the drawing toolbar (also found on the "View" menu) is visible at the bottom of the PowerPoint screen. You can add more lettering to the image by adding text boxes and/or arrows and pointers using the tools available on the drawing toolbar. Here is an example of slide creation in progress:
You can insert drawings into a Word file using a very similar technique, and you can add arrows and lettering to these drawings after they have been inserted. PowerPoint and Word make a good job of text, and the large fonts in PowerPoint will cause you to make good slides with minimal lettering that can be seen from the back of the room. Text positioning can be a bit tricky: try toggling "snap to grid" in the drawing menu if it won't do what you want. If all else fails and you absolutely must have your text in a pre-defined position then you must add it to the bit-map image instead.
If you have comments, queries or suggestions, email me at J.A.Illingworth@leeds.ac.uk