AGOCG logo
Graphics Multimedia VR Visualisation Contents
Training Reports Workshops Briefings Index
Next Back

A comparative review of HyperCard and Director as tools for time-based expressive work

Ian Phillips
Coventry University


I have pointed out elsewhere ["Supporting Postgraduate Work In Digital Media] that individual artists have become much more interested in producing time-based expressive work in recent years than ever before, partly because relatively cheap software tools have made the complex and expensive processes involved more accessible.

Most of the work produced falls within the definition of "multimedia", in that it incorporates some or all of the following media components:

I have also pointed out that the artist producing time-based expressive work may adopt one of three basic approaches:

This study compares two software tools for creating such work, HyperCard and Director. The headings under which the tools will be compared are:

The study does not attempt to provide instructions for using either tool, though detailed description of some aspects of operation is included where necessary for purposes of comparison.

Software tools for time-based expressive work

Computer-based tools for time-based expressive work must have certain capabilities and characteristics. The main ones are that they be:

Certain other characteristics, for instance availability across common computer platforms, are much less important to artists than to designers. For the artist, the fact that a work can only be seen or used on one particular computer system is not likely to be a deterrent to the use of a tool. Furthermore, interactivity has not yet proved as interesting to visual artists as some predicted. For most, it will therefore be the way software tools handle the creation of pieces that are essentially cinematic that will be the main basis for choice.

Both pieces of software under consideration in this study have all the capabilities and characteristics listed. They handle all media components. They have been available for some time and run on widely-used types of personal computer, so they are accessible to and usable by visual artists. They are in use at many sites, and hence are fairly well understood and supported in both academic and commercial life. They can be programmed (and extended) to perform complex operations. The next section looks at their history to provide a context for more detailed comparison.

A context for tite comparison a brief history of the software tools

HyperCard (and an earlier innovation, Guide) were the first generally available software ernbodiments of the concept of "hypermedia": media which enabled associative links to be made between entities, on the lines propounded by Vannevar Bush in the 1940s. Both were produced for the Macintosh, though an MS DOS version of Guide soon appeared.

The purpose of Guide was to allow an author to hide detail and provide navigational aids in electronic documents. It provided means of designating text or pictures as links to other rnaterial. A pointer moving across a link changed shape to indicate the nature of that link: expansion button, pop-up note, or cross-reference. A companion prograrn, Guide Envelope, allowed developers to distribute copies of their products to people who did not possess Guide because it built a "stand-alone" or "run-time" version. A recipient could thus view and interact with a Guide document but not build one of their own.

HyperCard starbed from a different set of priorities, aiming to provide users with a "software tool kit" with which to build their own applications. HyperCard's originator, Bill Atkinson, listed being "a cassette player for information" as one of these applications. Information had to fit within HyperCard's model of the world, where entities called cards were arranged in the manner of a Rolodex. These cards were at fist limited to the 9" diagonal size of the original Macintosh screen, which was monochrome. A collection of cards was called a sbck, though the arrangement was in fact a loop. Cards could contain entities known as buttons and text fields, along with black and white bit-mapped graphics. Any button or field could be linked to any other card or stack to make an associative link and there were huilt in visual effects such as dissolves to mark the activation of a link.

There was no means of extending Guide's inbuilt capabilities but HyperCard included a fully-featured programming language, HyperTalk. Simple interaction could be created using only menus and mouse clicks to activate small HyperTalk programs, or "handlers", supplied with HyperCard. Users could also write their own HyperTalk programs. These were known as scripts and could be used to define new procedures, as well as make use of those built in. Procedures which were too complex for efficient coding in HyperTalk could be written in lower-level languages like Pascal or C and called from within a script. HyperCard could therefore be extended aLrnost indefinitely.

This approach was inferior to Guide from a commercial point of view in that there was no provision to build a stand-alone product: users needed HyperCard itself. Apart from incurring extra expense, stack users could therefore pull a product apart, given sufficient time and interest.

Neither of these software tools supported sound or animation originally, though in HyperCard simple "flip-card" animation was possible and bitrnaps could be moved under script control. A new wave of products, touted as "multimedia" software, soon appeared which did accommodate these elements.

The first of these was for the Apple Macintosh because it offered better support for graphics and sound in standard form than either DOS or Unix rnachines. The program was called VideoWorks and was the forerunner of Director. It was aimed at the business presentations market but went beyond such products as Persuasion and PowerPoint in allowing the incorporation of sound and animation into a piece of work. It also introduced the novel idea of orchestrating elements on a "stage" by means of a "score", thus providing a direct visual representation of both the construction and the construction process along with the instant playback facility also found in HyperCard.

However, there was no means of providing interactive capabilities as in Guide or HyperCard. When VideoWorks metamorphosed into Director, some elementary programming capabilities were eventually bolted on in the form of the Lingo scripting language. This was initially neither as fully- featured, nor as well integrated with the host, as in the case of HyperTaLk and HyperCard but it has been developed substantially with each new release. It has taken until 1994 for a DOS/Windows version of Director to appear, so other programs have had the market for "authoring" software on that platform to themselves. However, there has been a "movie player" for DOS for some years.

Over time, the two tools have become much more alike in terms of their capabilities but there are still some striking differences. The tools will be compared in the next section, under the headings listed in the Introduction. The detailed comparison will be followed by a brief resume of user experience at Coventry School of Art & Design. This will include a short description of the making of the same sample piece with both tools.

Comparison of the software tools

Concepts and terminology
HyperCard still uses the card metaphor as the basic building block, where Director uses the stage. HyperCard projects are built up from stacks, Director projects from movies. In HyperCard the user navigates from place to place on a card, from card to card, or from stack to stack: in Director all action takes place on one stage and navigation is from frame to frame or movie to movie.

HyperCard uses several add-ons (ADDmotion, ColorTools, QuickTime Tools) to provide facilities for time-based work whereas Director has always included these facilities, apart from a sound editor, in the main program.

ADDmotion uses actors and props where Director uses castmembers, sprites, and backgrounds. ADDmotion actors consist of one or many cels and they can be animated along a path. There is a high degree of control over paths and they can be stored in a library for re-use or modification, as can actors and props. Props cannot be animated, but they can enter or leave the action with various visual effects. Actors and props can be created or edited in the ADDmotion paint system or imported as PICT or PICS files.

A Director sprite is an "instance" of a castmernber and it inherits its properties. However, it can be given different properties, or made to behave differendy, if required. Any sprite can be animated in Director, whedler it forms a backdrop to the action or not.

The equivalent of ADDmotion's multi-cel actor is dhe filmloop: any sequence of frames can be made into a filmloop and dlen used as a single entity. Path creation and control facilities are not as sophisticated as in ADDmotion and there is no library facility, dhough dhe fact dlat castmernbers are templates for sprites is an equivalent It is also possible to set up a "Shared Cast" movie containing any elemenb common to several movies: all movies in the same folder as the Shared Cast dhen have access to its castmembers. This makes it easy to split large pieces into smaller, more responsive, movies and make changes with relative ease.

ADDmotion uses a media controller widh two modes and a timelines window to create and control animations; Director uses a cast window and a score window. The Director controller sets tempo and background colour, as well as controlling dle position of the play head. Bodh approaches allow many events to be incorporated and for dhe play head to be seen in motion as the movie runs. There are many additional windows in Director, and dialogue boxes in ADDmotion, that may obscure the screen but which are used much less frequently.

Both tools support all the media components likely to be required in time-based expressive work, that is:

and both support the control of external devices such as videodisc players, VCRs, and CD-ROM drives. The next section looks at how dhey handle dhe various media components: dhere has not been time to investigate dhe control of external devices.

Setup costs

HyperCard is now in version 2.2. It has had a chequered pricing history. It was initially developed by Bill Addnson (one of dhe original Macintosh programming team) and, at his insistence, given away free widh every computer sold. It was later taken over by Claris, Apple's application software spin-off, who supplied a basic version free but made a charge for dle full development toolkit It is now back in dhe Apple fold and, much extended, retails for under 150.00 UKP complete widh development tools.

Hardware requirements are relatively modest, dhough not compared widl dhe original which ran happily on a flopw disk Macintosh widh lMb RAM. The current version requires over lMb RAM for the program alone. To run the animation add-on, at least 2Mb are required. HyperCard itself, the ColorTools, and ADDmotion will together occupy 2.5Mb of disk space. A complete installation occupies 10Mb but that includes 4Mb of ready-made stacks; HyperCard and HyperTalk reference stacks, and ADDmotion tutorial material. Running with both add-ons enabled and a relatively simple stack open, HyperCard occupies between a third and a half of a 4Mb block of RAM.

Large and complex stacks with extensive use of colour and animation will of course consume more resources and require more powerful hardware to perform properly. Allocating 6Mb to 8Mb of RAM to HyperCard is probably more realistic for serious time-based work: this implies a machine with at least 12Mb of RAM, since the operating system will need around 4Mb in most situations. All in all, a minimal hardware platform will cost around 1,000.00 UKP at October 1994 prices.

Simple sound editing tools are provided, so minimum setup costs to use HyperCard for time-based work will be around 1,200.00 UKP including hardware but only 150.00 UKP if a suitable machine is already available.

Director costs are in a different league. Macromedia, the developers, now offer a stable of programs for the creation of time-based work. There are three packages in the line-up apart from Director: for accelerating movies where no interaction is required (Accelerator); for threedimensional modelling and rendering (MaaoMedia 3D); and for sound recording and editing (SoundEdit Pro, SoundEdit 16, or SoundEdit 8). Director itself is available for both Macintosh and DOS/Windows machines.

Director 4.0 for the Macintosh costs 999.00 UKP; Accelerator 175.00 UKP; MacroMedia 3D 1,195.00 UKP and SoundEdit between 149.00 UKP and 279.00 UKP. As a minimum, one of the SoundEdit packages will be required in addition to Director itself. Software alone will therefore add up to about the same as a complete HyperCard and machine setup. Hardware requirements are similar at the bottom end: a complete installation, with 4Mb of extras including tutorial movies, occupies around 8Mb of disk space and Director requires at least 8Mb of RAM. However, movies eat up disk space and Director will require lots of memory for serious work: consider 20Mb RAM and a 500Mb hard disk the desirable minimum.

There is no sound editing facility built in, so initial setup costs for Director work will be around 1.200.00 UKP: double that if no suitable machine is available.

Neither HyperCard nor Director offer a one-stop shop for generating time-based work: it is likely that at least one two-dimensional graphics package will be required in addition to the paint systerns built in.3D software might also be required for use with either. Prices range from around 200.00 UKP for a simple 2D package to well over 1,000.00 UKP for 3D modelling/animation (e.g. MacroMedia 3D at 1,195.00 UKP).

Productivity can be inaeased with second monitos to hold the various construction windows. Standard 14" colour monitors cost less than 300.00 UKP and extra display cards range from less than 100.00 UKP to almost 3,000.00,UKP depending on the colour depth and speed required. Extra memory will enable larger projects to be undertaken: prices fluctuate but are currently around 30.00 UKP per megabyte. Faster machines are of course desirable if the money is available but second monitors, extra memory, and some means of taking back-ups are probably more cost-effective.

Handling of the various media components

A. Still pictures and animation frames

B. Sound

C. Quicktime digital video

D. Text

Integration with other sofware tools

The user will need both prograrns to cooperate with other software tools from time to time. Director integrates far better, with its support for common file formats and simple import and export capabilities.

There is little to choose between the two tools in this area. Although HyperCard had the edge until recently with HyperTalk, it was pointed out by a contributor to the Internet forum on Director that "Lingo, if it was wetl documented and supported, would be an excellent first language for budding programmers ... Lingo is truly object-oriented programrning, whereas HyperTalk is only "object-oriented liken (Simon Biggs). Both scripting languages can be extended with external code written in lower- level languages, so their programrnability is lirnited more by the skills of the programmer than by inherent qualities.

Performance of finished pieces
There is more information on this in the next section, "User experience at Covently School of Art & Design".

Learning to use the tools
Documentation and tutorial software are important aids to learning how to use any program. Poor or non-existent documentation can hold a product back as there is now fairly widespread resistance to buying software only to have it sit on a shelf because it is hard to learn to use. Supplier after-sales support, in the form of courses or help lines, is also important. Finally, knowledge within the user community represents an invaluable resource and there is increasing use of networked forurns to give and receive advice and technical support.

The original documentation for HyperCard was rather meagre and a large number of third-party books appeared to supplement it Apple now provides much better documentation for HyperTalk, together with good help software. Ready to use sample stacks have always been provided, making the product useful straight out of the box, and these can also be used for tutorial purposes. There is also good on-line documentation of most aspects of the software and the add-ons. Two third-party books are still available in the UK, one specialising in programming in HyperTalk.

There are still problems with the documentation: the ADDmotion software is bought in and has its own manual. This harks back to an earlier generation of Macintosh software documentation, being one slim perfect- bound volume short on detail. However, there is a reasonable tutorial stack and an excellent (though graphically crude) introduction to the principles of anirnation. All in all, the new user has a reasonable amount of help in getting started.

Director documentation was also widely perceived as deficient but only one third-parly book appeared and that did not deal with Lingo. The Version 4.0 manuals are much irnproved and one or two new books are on the way. The tutorial examples are not as good as those for HyperCard but there is still a large amount of material to work through which will help new and more experienced users alike.

Supplier and user community support
As far as I know, there is no supplier after-sales support of HyperCard. Macromedia do support Director, mainly in network forums. They also provide courses but I do not know of any in the UIC. Both products have good help systems: Director's is slow but easy to use; HyperCard's is quicker, slightly opaque but more detailed. There are active conferences and supplies of public domain and cheap software on-line for both products: networks are certainly the best source of after-sales support.

User experience at Coventry School of Art & Design

Everyone who has to learn about software for a living eventually develops a strategy for coming to terms with a new product. Mine is to look at all supplied documentation to get a picture of what the software is supposed to do; work through some or all of any tutorials supplied, and then embark on some work of my own. For this study, I therefore set out to create the same piece with both tools after completing the preparatory work.

It took some days of part-time work to complete the tutorials, making notes all the time. I then decided to make a short movie based on some old studio work which used simple step and repeat transformations. A square element containing an alphabetic character would appear on screen, move to the right, and come to rest in row one, column one, of an invisible four by four grid. A second element, rotated ninety degrees, would appear in the same place as the first and move to occupy column two. A third element, rotated a further ninety degrees, would then appear and move to column three. Finally, a fourth element, rotated a further ninety degrees, would appear and move into place. The next row would then be started, with the element for column one rotated ninety degrees from the one in the previous row. These systematic transformations would continue until the grid of sixteen elements was complete. A simple looping sound accompaniment would then be created, with punctuation as each element appeared for the first time. Both movies would be compared as to size and performance and then exported: first as QuickTime movies and then as standalone applications.

Using HyperCard
I started with HyperCard/ADDmotion. The first problem was creating the elements. I had already made one original animation after completing the tutorial example but neither movie used precise geometric shapes. ADDmotion's paint system, incorporating facilities for creating multi-cel "actorsn and viewing the previous cel while working on another, had no obvious way of ensuring precise registration for rotations. It proved impossible in the time I had to make a square and rotate it about its centre with sufficient accuracy to fit it with others in a grid. I resorted to Photoshop and created the four elements I needed.

The process of orchestrating the visuals was fairly simple, though laborious, once the elements were created. The use of libraries, dialogue boxes, and the timeline was easy to understand but involved a lot of repetitive actions. Director definitely has the edge in this sort of work, with good facilities for replicating and then modifying frames of a movie, although its score can be harder to read than ADDmotion's timelines.

The sound required was simple enough in this case but neither the HyperCard nor the ADDmotion sound editing hcilities were sufficient on their own. Again, I resorted to another tool: SoundEdit Pro. Orchestrating the sounds created was very easy as it is possible to have many soundtracks in a timeline. Director is limited to two sound channels, so mixing percussive events with continuous sound is more difficult and any multi-layered soundtrack must be built up off-line.

The finished movie exhibited very rapid fall-off in performance as the elements slid into place. By row four, the frame rate had fallen to less than a quarter of its original value. The manual offered no clue as to why this might be and playing about with the nurnber of actors and sounds made no difference to performance.

Using Director
I was more familiar with Director, though notwith Version 4.0, when I started to recreate my sample movie. I built the movie in about the same time, discounting that taken to re-orientate after using ADDmotion. However, I was no more successful in creating the elements. Director's paint system rounded the 66 pixel square elements to 66 x 65, 65 x 65, 65 x 66, and 66 x 66! Again, I had to resort to the elements created in Photoshop.

Although complex movements cause Director movies to slow down, I had no trouble with mine and it ran appreciably faster than the HyperCard version. The Director documentation is full of tips on optimising performance, whereas the HyperCard documentation does not even recognise the existence of ADDmotion, but in neither case could I find any explanation for the difference in performance. Perhaps it is due to coding efficiency: the finished HyperCard movie was 1,092Kb, whereas the Director movie was only 400Kb. The difference was even larger in the case of exports as QuickTime digital video: the HyperCard export was 1,008Kb, the Director export was only 280Kb. The two standalone applications were much closer in size, with the HyperCard one at 340Kb compared to 688Kb for the Director movie.

My own experience is similar to that of colleagues and students at CSAD. Prior to the release of version 2.2, HyperCard was used wherever interaction was the main feature of a project and Director wherever integration of media components and/or output to video was the most important aspect. HyperCard was not used for expressive work after Director became established in the late 1980s.

This year a colleague, who was also a part-time student on the postgraduate course I teach, produced the first exploratory expressive piece using HyperCard seen for some time. Interestingly, he did not use ADDmotion, though he did make use of the sound and colour tools. All other expressive pieces (both interactive and cinematic) produced since I took up my post in 1991 have been executed with Director. This tool has also been the one chosen most often for inforrnation, education, or entertainment products. A colleague used Director to produce a CD-ROM exploring the sources of the world's major religions, despite the fact that HyperCard had superior control of interaction, because of its handling of the media components. Students and former students employed in multimedia development are using Director without exception, though it is sometimes used only for prototyping.


Time-based expressive work can be produced with either of these products. Neither product copes particularly well with pieces employing precise geometry. Pieces employing cartoon imagery or multi-layered collage work much better.

Director offes better integration of the media components and superior performance; HyperCard provides a lot of functionality at a very low price. Both have similar hardware requirements. Learning to use either product is difficult and time-consuming; partly because there are still many flaws in the tutorial documentation, partly because both are complex pieces of software for performing inherently complex operations.

Although HyperCard in its original form was an elegant design and offered better interactive capabilities than Director, much of the elegance has now been lost. The interactive and time-based aspects are not well integrated, exposing the fact that the architecture of HyperCard and its add- ons is rather unsound. As well as leading to clumsy operation, it discourages the use of other tools. It is also likely to limit future development.

The final choice will depend on personal circumstances as much as the nature of the work undertaken. The artist working in an educational establishment is likely to have access to both: the individual working alone is less likely to have access to Director. If HyperCard does what is required as it stands, then it offers a cost-effective solution for creating time-based expressive work. However, Director is likely to be the most rewarding choice in the long mn, provided enough money is available. Integration of the various functions and co-operation with other tools is better and there is more scope for development than in the case of HyperCard. In the end, I suspect that Director will continue to dominate the market thanks to its cohesiveness and smooth handling of various media types.

Next Back