danceart.

doc

Multi Media in the teaching of Dance Performance. (From the Dance Disc to current research) Ref: 199202 By Jim Schofield Director of Information Technology Goldsmiths' College UNIVERSITY of LONDON
Background The appropriateness of technology Interactive Video has been around for many years. Indeed it is generally considered to be a bit long in the tooth, and indeed that the blossoming of computer controlled and accessed videoimages into "new age" methods of teaching is long overdue. There has of course been a great deal of material produced in the area of management training, based mainly on video sequences of problem situations, with the trainee asked at various points for judgements as to appropriate actions in the given contexts. There has also been several very influential forays into technical training with a richer level of interaction and the use of facilities such as graphics and animation, but these have not yet led the expected explosion of growth. The extended wait for CDI has perhaps been a major factor in delaying wider use, for only after this technical watershed could the very much larger consumer marketplace greatly increase the possible production runs and hence reduce unit costs. The wealth of resources available now to address problems of instruction within multi media is positively breathtaking, and, maybe just because of this, there prevails an attitude of being tooled up with the latest in high technology, and actively looking for an appropriate application. The Dance Disc did not start this way. The Problem came first and the techniques developed were not applied from a well-used tool set, but were expressly developed to address difficulties that were inherent in the teaching of dance performance. The crucial principle in the development of the teaching resources was that they had to be appropriate. If it was better to do a certain part of work in the form of a small booklet, then this was done rather than to squeeze it into the multi media presentation. The technology that was entirely appropriate to the effective presentation of movement itself however, was, on the other hand, milked to the nth degree. The problems to be solved The problem was in the performance part of the GCSE in dance. Four dances were specially choreographed, one of which had to be chosen, learnt and then performed for assessment. Each dance was notated and this notation (on paper), along with the selected music (on tape), was sent to the dance teacher at the school. In addition, it was arranged that these teachers could attend special sessions where the dances were shown, and some help was given in organising the teaching of the pieces. The major difficulty lay in the inadequacy of this transmission mechanism. As you may know, a large number of teachers of dance in schools are not specialists, and only a very small number can read notation. This led to inaccuracy in the dance movements which was of course to the pupils' detriment at assessment. Also, the minimal contact with the performed work meant that teachers were not able to develop a feel for the content and expressiveness of the pieces, which, once more, hindered the effectiveness of their pupils' performance. Another area of difficulty lay in the teacher having to demonstrate the dance in front of the pupils, who would follow her from the rear. This meant that the teacher had to be

visible by all and therefore some distance from individual pupils, and also showing her back was obviously not the most advantageous position for observing errors in her pupils performance. Of course, many of these problems were tackled as best they could by the teachers involved but even among very dedicated practitioners, the continuing difficulties were
admitted.

Previous steps towards a solution Of course, these major difficulties were quickly appreciated by the examining authorities, and attempts were made to put them right. The first, and most obvious move was to video the dance, and make this available to schools. This was done as two complete single viewpoint run throughs of each piece: one from the front, and one from the rear. The teacher could play them through OK, but it could be very messy cycling round a particular bit, using the pause, rewind and play buttons, and was particularly difficult jumping between two or more occurrences of the same motif in different parts of the dance. In addition, the limitation on views and the lack of detail blow ups led to inaccuracies in movement, even with the video in front of the users, and even more frequently to inaccuracies in position and the movement across the dancing space. Frequently, the dance would be performed "on a sixpence". The Philosophy and Form of the Dance Disc This Interactive Video work is unique in that it does not present a structured teaching sequence of the material with a small amount of interaction at particular points. The disc is better considered as a "teaching resource" than as a "learning package". Access to all sequences and treatments of the material is available at all times, so that any particular arrangement of the material, or the amount of time spent on any part or method of presentation, is entirely down to the teacher who is using it. That is not to say, however, that the material is unstructured, for the variety of treatments to the various sections of the dance have been determined by the structure of the dance itself, and the known difficulties in its teaching derived from the choreographer herself, and from trial work on teaching the material to its target audience. Neither was it considered that the Video, textual and graphical material presented onscreen was in any way complete in itself. Differences in the background of the students and experience and interests of the teacher, would of course mean that the material available in this package could be used in a great variety of ways and embedded within a wide variety of tuition content. A major gain of the research was the realisation that quality resource material could be used very differently in many diverse contexts. For example, because of the supreme professionalism of the dancer on this disc and the synchronised images of both the dance and the notation symbols on screen together, it would, and has been, quite possible to invert the use of the material and use it to teach Labanotation. It has also been possible to use this material in the teaching of choreography. The video material was a great deal more extensive than would be available in a simple linear video tape of the dance. The full dance was shot from the front and the rear, and many detail shots were taken of appropriate phrases of the dance. These included close-ups of the feet or upper torso, overhead or other angled sequences, and even posed stills of jump positions. The whole screen can be used for a single front or rear view, or most of it can be filled with a detail, while a reduced synchronised reference sequence of the full view can be inserted in the top corner. In addition, it is possible to share the screen between the back view and the corresponding symbols of the Labanotation score synchronised with the video movement. At the touch of a button, the dance can be restarted from the beginning, switched between front or rear views, rerun over the last shown detail sequence, or paused, restarted, switched to slow motion or even run backwards. A second audio track with voiced instructions can be turned on an off by a single key depression and a complete menu of the phrases of the dance can be brought on screen to allow access to any part of the work at any time.
Some points about interface

The layout of the screen and the means of access were considered at great length and tried out on typical users, and this led to a unique user interface. A hierarchy of menus was considered inappropriate for the kind of user involved here, and a set of transparent icon type indicators was developed, constantly in view, but with minimal interference with the quality of the moving video in the background. If the user is concentrating on the moving video, the icon overlays melt

away, whereas when an interaction becomes necessary, the user can easily switch his/her attention to the icons. The effect is very similar to head-up displays in aircraft, where attention can be switched between what action is going on outside the window, and what transparent information is projected upon it. Some of the icons represent the interactive options available at any given moment and change at different stages of the dance. Others indicate the various modes of access to detail video, such as "loop back", "stop at end of phrase" and "run through into full view & continue". Finally there are, an animated spatial position map synchronised to the underlying video ,a time elapsed "thermometer", and a bar count indicating the current point in time. The interaction of the user is by a limited set of keys. In fact, only function keys are used, which are programmed to perform meaningful (to a teacher of dance) tasks at a touch. No entries of more than a single touch or number are required. Typing of words or commands are unnecessary. In later implementations a single shot button has been added which can be used to move one frame at a time forward or backward, or, by keeping the button depressed a form of slow motion can be achieved. What is particularly useful about this addition is that the slow motion is perfectly under control in that removing the finger causes an instant freezing of the action on screen. Another button acts as a toggle between forward and reverse motion, so that a cycling round of a particular phrase can be achieved simply by repeated pressing of the same button. The option of the same functionalities by mouse has been developed for later versions. The development and production team -analogy with a film No small part of the success of this work was in the team that was put together to undertake it, and the apportionment of responsibilities both overall and specialist between these team members. The first principle in building the team and deciding on these responsibilities was that "the technological tail must not wag the educational dog". It was not a technologically led project. The real leader of the project was the discipline expert, though there was a computer expert and a very able television director, the decisions as to what should be done was always in the end determined from the standpoint of the discipline aims and targets. Nowhere was some particularly clever or newly released technological trick allowed to intrude to give "tech cred", but, on the contrary, if it got in the way of the educational purposes, such a trick would be rejected. For example, it was put forward at one point that the Labanotation should be scrolled by, in synch with the video of the dance in the other half of the screen, but though very jazzy looking, the effect would have been a very busy screen, with no easy transition for the user between the moving image and the notation, so this method was not used. In the end, the best way of handling the notation was to flick from one page to another with a half bar lead in and out. This meant that the notation was at all times stationary, and the user, having glanced at it once, knew where to find her place when required, until the page was completed. At the same time the role of the other members of the team, should not be underestimated. Often the discipline expert would not know what could be achieved in computing, or in the editing suite, and informed advice led her to make the very best of these different areas. A very close relationship between the team members in planning the work was essential from an early stage. It would not have worked as effectively if, for example, the video director had been brought in merely to do the filming, or the computer director employed when everything was in the can to write the software. Some technical aspects An extremely relevant part of the technical solutions to the teaching problems mentioned earlier, was in the synchronisation of underlying video with animated over-lay graphics. These were used, for example, to give an animated map of where the dancer was at any time, even when the video was rolling. This was possible because the floor of the dance studio was marked out with a clear grid and a graphic representation of this on screen was furnished with an animated dot showing the precise position of the dancer, at each moment in time. Knowing where you were in the dance, could not simply be recognised by the current on-screen movements, because such moves and motifs could be repeated at various times within the dance,

so a synchronised animated thermometer indicated exactly how far through the dance it was at any time. These were achieved, rather surprisingly, by real-time animation effects generated as the user was viewing the material, though, of course, such images could have been generated beforehand and edited onto the video image before mastering. This method was not chosen, however, because it was clear that, at a later stage, real time effects would be essential in the further development of the project. It was also necessary that the consequences of the resultant housekeeping overheads were clear at a very early stage, or the feasibility of such further developments would be unknown. Such effects include the ability of the user to "draw" upon the image as they do in television sport, to add value, and to make it possible that these marks be whatever is demanded at the time, and not frozen onto the Laser disc as a limited set of possibilities. One of the more remarkable achievements arose out of the fact that there was not sufficient time in each cycle of housekeeping to keep the animations up to schedule, particularly when the user was pressing keys. The solution was to run the housekeeping cycle on a variable speed clock, that would run fast when it had time to do things, and slowly when it was swamped. The solution was effective because the user interaction was only a small fraction of the total time, and the system very quickly caught up with the action. In fact, the delays at slow times were so small that the time it took to look from key being depressed to the screen overlays, was sufficient for the animation to be back in accurate synch. The Result The research described above obtained funding from Digital Equipment Company, the Calouste Gulbenkian Foundation, the Arts Council and the Television Fund, and won a British Interactive Video Award in 1989 in the education section for Jim Schofield, the Computer Director (now of Goldsmiths' College) and Jacquie Smith, the Choreographer and Research Director (of Bedford College of Higher Education). Future developments for IV in dance Since the original work was completed, the team involved has been keen to develop the ideas in many new directions. First the techniques needed to be ported to other computer platforms, the chosen ones were IBM compatible and the Apple Mac. In addition it was necessary to develop user friendly software to enable discipline experts such as teachers and choreographers to themselves edit the video material for division into teaching phrases, and also to be able, themselves, to add the graphics overlays to the underlying video precisely where it was needed. The first phase of this has been successfully completed in the development of a Interactive Video Editor. In addition, it was necessary to test out and develop the techniques in further examples of dance performance, so in 1991 further videos were shot, 2 of Contemporary Dance (one male dancer and one female), and in a separate piece of research, a beautiful Indian dance and an exciting example of Nigerian dance have been videoed for analysis. Perhaps the most original current work is in new research by Jacqueline Smith, who with colleagues in mathematics and music is using dance to teach primary mathematical concepts and is communicating the movements, symbols and ideas through multi media. Wider areas of application (The parametered figure) The techniques developed for the dance disc were quickly realised to be entirely appropriate for other forms of movement analysis such as sport's coaching and gait analysis. The method of overlaying animated graphics onto an underlying video could add value to the original image. The plane of swing in a golfer's stroke, the movement of the shoulder in a tennis volley and many other similar areas, could be overlaid as alternatives to an underlying video case, without the enormous data loading of doing them all in video. Close-up shots from various angles could be juxtaposed with a synchronised, full-view, reference image of the player, and run together at

full speed or slow motion, or even paused, and advanced in single shots, to accurately study a particular movement. Sequences of regularly timed stills can be shown against the same sequence as a moving but controllable image, again revealing the changes of rate and orientation within a particular movement. A further development of the graphic overlay is in the area of a parametered figure. Such a driveable wire frame figure could be superimposed upon the underlying video of an athlete in action to show alternative, animated movements synchronised with filmed images. The heart of this work, which is still in development, is to constrain the parametered figure to move with "realistic" motion. This is achieved by first "fitting" a figure to large amounts of underlying video to generate a rich database of actual movements, then processing this information to develop a set of constraints or limits on articulations, extensions, ranges of movement, and speeds. Then, when such a figure is driven to shown "new" movements, the "derived laws" can be made to constrain movements in accordance with those distilled out as real human motion. The main centres for both the development and use of this work are still the Bedford and Goldsmiths' Colleges, though the applications to biomechanics and athletics are beginning to be taken up at other centres. The more the merrier. Multi media is entirely appropriate to the study of human movement whether aesthetic or athletic.