Image Processing Reference
then reconstructed and mixed locally. Taking the speech-synthesis and MIDI information
into account, very rich audio experiences can be created.
22.214.171.124 Phonetic Component Track With the advent of speech-recognition and
search systems such as those developed by FastTalk, UI Speech, and DreMedia, it is likely
that the audio signal will be processed and stored in other forms. FastTalk used a phonetic
search algorithm, so to search the video file for phonetic matches, a track containing
phonetic fragments decoded from the audio and synchronized with the time code at
which that they are spoken might be searchable very rapidly.
126.96.36.199 Subtitle Track In video that is being prepared for mastering onto a DVD,
there could be eight or more subtitle tracks. Although the delivery platform you deploy to
is likely to have a strict limit on the maximum number of supported subtitle tracks, the
master video file may be stored in a format that allows an unlimited number of tracks to
be maintained. The rendering process for the target platform then merely selects those
subtitle caption tracks that it requires. This is a specific variant of the simple text track.
188.8.131.52 URL Track This is another variant of the simple text track and contains a
URL that is passed to a Web browser by the desktop video player when the video is clicked
on. This has been underused so much that many people don't even realize it is available.
Real Networks media producers can create a list of URLs with time codes and then process
the file through an rmmerge utility to create a URL-tagged video file. Interestingly, the
allow for very reactive synchronized content to be deployed.
184.108.40.206 Hotspot Track Sometimes hotspot maps are created and associated with
the URL tracks in order to provide clickable regions on the screen. This is certainly used to
good effect in QuickTime VR movies, where the hotspot map can be applied before or after
the de-warping of the image. The hotspots can appear as positionally relative to the scene
when applied before the de-warping, or as an overlay on the control surface if they are
added afterward. This is a pixmap with the pixel values selecting one of several available
links. Those links may be URLs for web pages, or they may be navigable links to other
scenes or panoramas within the movie. Hotspots are also useful in linear-viewed video, and
you could have a product placement that was clickable to a “buy me now” web page, for
example. Hotspot tracks might be quite bulky, possibly similar in bulk to an alpha channel.
220.127.116.11 Sprite Track This is a track containing an animated object that is overlaid
on top of the video. It is likely that there will be multiple sprite tracks.
18.104.22.168 Metadata Track Information about the video, such as copyrights, date
of shooting, location, and camera settings, is called metadata. There is a wide variety of
information that is useful to the production process and still more that assists the end user
when consuming the final product. Adding this information to the video content is some-
times referred to as tagging. There is also some metadata that is outside the video file and