tag:blogger.com,1999:blog-86041962011-06-08T02:27:39.870-04:00christinenoreply@blogger.comBlogger21100tag:blogger.com,1999:blog-8604196.post-1097474274074420072004-10-11T01:50:00.000-04:002004-10-11T01:57:54.073-04:00<a href="http://www.graphics.csail.mit.edu/~sand/hand-edit.avi">test video</a> (10mb)<div class="blogger-post-footer"><img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/8604196-109747427407442007?l=cmliutangible.blogspot.com' alt='' /></div>Peter Sandhttp://www.blogger.com/profile/06635909613923411034noreply@blogger.com3tag:blogger.com,1999:blog-8604196.post-1097029897330504042004-10-05T22:30:00.000-04:002004-10-05T22:57:14.146-04:00met with dick and peter this afternoon [over ramen, coffee, and strong magnets] and discussed ideas for the first tangible project. the assignment: to design a TUI using something related to graspable objects and hands-on materiality. we hope to use computational power to augment a task in a meaningful, expressive way. pushing for a video demo for next week's class. <br /> <br />initial inspiration sparked from idea about how to manage high numbers of objects [akin to sorting hundreds of lego blocks by color, etc]. most projects we've seen have dealt with objects of few number [on the order of less than 10, even 100], and hadn't fully tackled the idea of manipulating manifold pieces of data. we wanted to harness the power of two-handed quick finger actions on a multi-item scale of intricate organization. <br /> <br />conversation veered to theater / film / arts applications, and honed onto film editing. for a small-scale project with a couple clips to work with, a regular GUI seems sufficient, but software tends to get unwieldy with larger-scale projects with a sheer amount of small clips to rearrange, edit, and preview. therefore, our TUI will use small handy two-headed blocks to represent each clip; projected onto the block are the first and the last frames of the clip. <br /> <br />[insert model here] <br /> <br />we formed a quick mockup using steel nuts and tiny magnets to hold them together. the magnets had a large enough field that two pairs of 'clips' could be joined by a 'transition nut' without an additional magnet. <br /> <br />to view frames within a clip, you could stretch the two ends out, and the contents would be displayed within the endpoints. <br /> <br />[insert picture here] <br /> <br />because each block is small, easily identifiable, and snap together [oh so satisfyingly], a tangible interface is ideal for quickly sorting out and manipulating a film sequence. <br /> <br />more ideas: you could layer [in the other dimension] audio tracks, or overlay multiple clips. to compare parallel clips, you could have a nut signifying 'choice', in which you could quickly preview and decide which clip to choose from the parallel contenders. <br /> <br />[insert model here] <br /> <br />on the side of the table, there would be 'function' areas where one could delete, duplicate, or archive a clip. the task would be completed by merely scanning the clip over the area. the internal data would be updated accordingly. <br /> <br />on the surface or on a screen, there could be a preview window where, if the user tapped the table or signified through some input, the movie could play. more nuts signifying <b>start</b> and <b>end</B> placed on the sequence of clips could define what frames the user would want to preview. <br /> <br />to implement the demo video, the current approach is to spraypaint the nuts with matte black paint for the tracking software. materials: camera, projector, table, and lots of nuts [or thick washers] with magnets.<div class="blogger-post-footer"><img width='1' height='1' src='https://blogger.googleusercontent.com/tracker/8604196-109702989733050404?l=cmliutangible.blogspot.com' alt='' /></div>christinenoreply@blogger.com0