Soft Cinema is a (potentially unlimited) collection of short movies in different styles. Some are in the tradition of “film essays,” some are fictional narratives, some are non-narrative ambient music videos. While the sound track of each movie is fixed by the author, the visual track is constructed by software. This includes the layout of a screen, the selection of media elements and their temporal order. In other words, software decides what appears on the screen, where, and in what sequence. The decisions made by software are partly based on a system of rules defined by the author and partly random. In short, Soft Cinema can be thought of as an automatic VJs - or more precisely, a FJ (Film Jockey).
Soft Cinema software selects video clips and animations from a database. Each video clip in the database follows Dogma 95 rules: it is shot using hand-held camera, it contains no edits, etc. Most of the clips have been recorded by the author in Berlin, Tokyo, Moscow, San Paolo and other locations between 1999 and 2002; the rest of the clips are simulated (i.e. a still image was animated to look like a video shot on location).
The length of each movie
corresponds to the typical length of a music track (3 - 7 minutes).
Soft Cinema software is written in Director and it runs on a standard PC or Mac. In order to assure the correct playback speed, a computer with a relatively fast CPU should be used (P4/1800 or G4). No other additional hardware or software is required.
Soft cinema can be exhibited in a number of different ways, for instance:
- as a viewing station consisting from a computer and a monitor or a projector – presented stand alone or included in a media lounge;
- as a video tape or DVD included in a screening program (a linear version);
- as an installation that consists from a few computers with projectors or monitors, each running a different version of the system (installation version).
In the case of an installation, it
is desirable to show at least three different versions in order to demonstrate
the range of the system. One uses fixed audio track with voice over which provides
narrative continuity. Another version also uses voice over but a different
system of editing rules that results in a very different visual track. In the
third version software assembles both video and audio. Therefore, while in the
first and second version software loops through the same set of movies (each
set runs for about one hour), in the third version it runs forever creating
different combinations of video and audio.
Referencing brandscaping (the three-dimensional design of brand settings), early algorithmic computer art, and the logic of modernist art movements (where paintings, graphic design, archiecture, and industrial design were typically driven by a single aesthetic system), we use the same algorithm to generate the screen layouts, the layout of Soft Cinema book, and the 3D layouts of Soft Cinema Installation. The proportions between different walls in the installation are the same as the proportions of the difirent windows in Soft Cinema Movies. If Modular was based on the dimensions of a human body, our system of proportions takes as its origin the dimensions of DV NTSC image: 720 x 480 pixels. In addition, the contrast between diffirent types of images (video, 2D animation, etc.) in Soft Cinema movies is translated into the contrast between diffirent materials in the installation.
Soft Cinema includes a number of movie sets each presenting a fictional short story. The stories come from a collection of stories written by the author between 1998 and 2002 and entitled Global User Interface [GUI]. Each story takes place in a diffirent location: Texas, Hamburg, Kiev, Mongolia, etc. (In writing the short stories, I followed the principle that they can only take place in the locations I never visited.)
Typically a story has been divided into a number of sequential parts, each part becoming a short movie. At the beginning of each part software generates a new screen layout, which can consist from two to six different windows. Software also selects which video clips and animations will play in these windows and in what order. This process is repeated for each of the part of the narrative. Following the same modular logic, diffirent voices are used for the diffirent parts of a story.
The small window that always appears in the bottom left corner identifies the part of a stories currently playing (for instance, texas_01.txt, texas_02.txt, etc.) A narrow horizontal window presents scrolling sentences selected from the same story part.
While the stories refer to the processes of globalisation and their effects of subjectivity, the visual track makes similar references in diffirent ways. Since most clips show the typical urban activities, Soft Cinema at first can be thought of as belonging to the genre of "city films" defined by such classics such as "A Man with a Movie Camera" and a "Symphony of a City." However, in contrast to these earlier films which included the expressive shots of various form of industrial labor, Soft Cinema repeatedly returns to the same bland image of information labor: a person in front of a computer. In addition, since we often see the clips shot on diffirent continents side by side, Soft Cinema more properly can be thought as a “global city film.”
Put differently, the subject of Soft Cinema is a new “global style” or a “global layer” of globalization: the hotel lobbies, the airport waiting lounges, shopping, the info-workers staring at computer terminals, the reflecting and transparent surfaces of office buildings, people waiting for the next train on a new shiny station, etc.
Soft Cinema is based on three ideas.
The first is algorithmic editing of media materials. Each video clips used in Soft Cinema is assigned keywords which describe both the "content" of a clip (geographical location, presence of people in the scene, etc.) and its "formal" properties (dominant color, dominant line orientation, contrast, camera movement, etc.). Some of the keywords are generated automatically using image processing software while others are input by hand. The program (written in LINGO) assembles the video track by selecting clips one after another using a system rules (i.e. an algorithm). Diffirent systems of rules are possible: for instance, selecting a clip which is closest in color or type of motion to the previous one; selecting a clip which matches the previous one party in content and party in color, replacing only every other clip to create a kind of parallel montage sequence, and on on.
The current version of Soft Cinema software lets the author define such systems of rules which it then uses to put together a sequences of video clips which best satisfy these rules. However, it is also possible to create other versions of the software which would give the author a tighet control over the sequence. For instance, one version may involve a video track completly edited by the human author beforehand. Some shots are designated as “replaceable” while others remain without modification (to keep narrative continuity.) Another version may contain a variable which is set by the author and which tells the program the probability that any shot will be replaced. In sumary, instead of opposing complete randomness to the complete control of a human author, Soft Cinema investigates a diffirent paradigm: using a computer as an "association machine" that complements / reacts to images selected by the user with other images.
(Interestingly, CD and MP3 players and well as software for music playback, such as iTunes, all include an option to play songs in random order. Can this be another example of electronic music culture being ahead of other parts of culture in using new computer logic?)
While a dominant recent tendency in audiovisual computer culture (VJs, Flash and Shockwave audiovisual pieces) is to synchronize image and sound (using video output to control/generate the sound, or, conversely, using audio to control video), Soft Cinema adopts another model, influenced by Eisenstein's theory of audio-visual montage based on musical contrapunct. In Soft Cinema movies, visuals create their own flow which mostly runs independently in parallel to the flow of narrative, but "syncs up" with it in the key moments. That is, periodically a particular video clip is selected to “anchor” the narrative events.
The second idea is what I call macro-cinema. While filmmakers such as Peter Greenaway and Mike Figgis have already used a multi-screen format for fiction films, thinking about the visual conventions of Graphical User Interface as used in computer culture gives us a diffirent way to do macro-cinema. If a computer user employs windows of diffirent proportions and sizes, why not adopt the similar aesthetics for cinema? In Soft Cinema, the generation of each movie begins with the computer program semi-ranomly breaking the screen into a number of square regions of diffirent dimensions. During the playback diffirent clips are assigned to diffirent regions. In this way, software determines both temporal and spatial organisation of a work, i.e. both sequencing of clips and their composition.
Another inspiration for macro-cinema comes from contemporary cultural sites which already adopted a multi-frame format. One example is news and financial broadcasts which combine a video of an annoncer, a looping text, charts of stocks, etc. Another example is the use of multiple frames in many computer games where each frame may present the environment as scene by a diffirent character. Importantly, in both examples, the information presented in diffirent frame is related to each other but it also has a semantic autonomy (in contast to the traditional cinema montage): for instance, an annoncer would still make complete sense if all additional graphics are taken away. This gives us some directions of how to use multiple frames in macro-cinema.
Finally, yet another inspiration for macro-cinema comes from the evolution of video production and distribution technologies. While NTSC/PAL resolution of analog video and television was hardly sufficient to present even a single scene, HDTV standards (1920 x 1080 and the like) makes possible to devide the screen into a few frames. In fact HDTV television specefications allow broadcasters to break the total bandwidth of a signal (approximately 19 GB/sec in the US) in diffirent ways: for instance, transmitting one high-res image and a few medium images, or a larger number of low resolution images, etc. In short, "fixed resolution - single image" convention of both 20th cntury cinema and television has already become technologicaly and conceptually obsolete.
While at present (2002) HDTV equipment is quite expensive for artistic use, it is possible to use Quicktime at DV resolution (480 x 720) to experiment with how multi-frame high-res cinema and television may look like in the future. This is the strategy used in Soft Cinema 2002 version. The original DV material is scaled down to 320x240, 240x180 and similar resolutions and encoded in QuickTime using Sorenson codec. This allows the Director program to play up to 6 clips next to each other within one DV NTSC resolution frame (720 x 480).
In the installation and the PC versions a Director program assembles movies in real-time. It is also possible to have linear versions of the project available on DVD and video tape. To create this linear version, we (1) choose the movies we want; (2) connect a video camera to a computer; (3) run Soft Cinema software. The camera records what appears on a computer monitor. (The linear version is available on DVD and all standard digital and analog video formats.)
The third idea is to create a true
multi-media cinema. In Soft Cinema
video is used as only one type of representation among others: 2D animation,
motion graphics (i.e. animated text), stills, 3D scenes (as in computer games),
diagrams, etc. In addition, Soft Cinema supplements a "normal" video
image with other types of lens-based imagiory commonly used today by industry,
science, medecine and military: the low res web cam image, an infrared image,
edge-detected image as employed in computer vision, etc. While some music
videos and artist videos already mix some of these diffirent types of imagery
in one work, Soft Cinema assigns each type of imagery to a separate window in
order to dramatize the new status of “normal” video, photographic
and film image today – no longer the dominant but just one source of
visual information about reality among many others. The additional inspiration
for using diffirent representation of the same scene next to each other comes from
the display setups used in medecine, aviation and other contemporary
workplaces. Finally, rather than simply using these diffirent types of
representation for a purely visual effect, Soft Cinema investigates the
possibilities of using them together for fictional narration.
[ Lev Manovich, Berlin, July-September 2002 ]
Soft Cinema: Project History
Soft Cinema incorporates the earlier project macro-cinema (1997-2000) which was developed as a counterpoint to the earlier project little movies (1994-1997).
Soft Cinema: Additional Texts
The text from The Language of New Media on macro-cinema (1999):
www.manovich.net/macrocinema.doc
Soft Cinema is an example of a "database narrative" proposed in my article Database as a Symbolic Form (1998)
Soft Cinema: Credits
Lev Manovich:
media database (videography, 2D and 3D animations, narratives), keywords, editing rules, spatial composition principles, image processing software, sound design, installation design concept.
Andreas Kratky (Berlin):
implementation logistics, edit list generation software, display software, screen layout.
DJ Spooky (New York):
music database.
Gloria Sutton (Los Angeles):
voice over.
Francesca Ferguson (Berlin):
voice over.
Rachel Stevens (New York):
voice over.
Ruth Lorenz (Berlin):
architect (installation version at ZKM).
Jason Danziger (Berlin):
Architectural consultant (installation design).
Andreas Angelidakis (New York):
architectural design (unrealised).
ZKM:
installation construction,
hardware.
Christine Bokelmann (Berlin):
graphic design (Soft Cinema book).