First step: laboratory start the identification. From a copy of the film with a time-code (or TC) and a transcript of the dialogue, a technician cut the dialogue in subtitles. Using subtitling software, it defines for each subtitle entry point (TC in) indicating the start of subtitle, and an exit point (TC out) indicating the end of the subtitle. The difference between these two values defines the duration of the subtitle and therefore the maximum number of characters available to the translator.
Eg: TC in 01:03:27:22 / TC out 01:03:29:15
duration 1 second 18 images number of characters: 25
It takes between one and two days to locate a movie based on the number of subtitles. To give an idea, a film normally “talkative” contains about 1000 subtitles. An action movie 800, and a Woody Allen film 1500.
The identification must facilitate maximum reading of the viewer and, to do this, it is based on the plans and their rhythm. Following the development of subtitling software, translators are becoming more likely to make themselves the identification. It is a technical work which must of course be paid separately and paid in wages, while the translation work is paid in royalties.
The translator then receives from the laboratory: a MPEG-1 time-encoded file, the numbered dialogue and the computer file containing all the tracking data.
Then comes the stage of drafting subtitles or the actual translation. The common problems to all translators are additional technical constraints described above, which often requires the translator to show some of brevity and clarity, due to the subtitle frame rate. The set must be faithful to the work (language levels, plot subtleties) while being consistent (discussion level between characters, to give just one example).
This phase in principle lead to exchanges with the sponsor’s representatives (distributor, producer for films that do not yet have a distributor, sometimes the director).
The file back to the lab to the third step, the simulation.
In still video, the translator views subtitles as they will appear on the screen, with a technician (simulator, or more often a technician) and possibly the sponsor or its representatives. The simulator is the first spectator, his opinion on the subtitles is important for the quality of work. Although this is not necessarily the case, everything can still be changed: final corrections of the text, the typographical arrangement, spotting (extend a subtitle that we have no time to read, shorten if it runs over a hesitation, the split, etc.). Although work began on work items, before the final cut of the film, the simulation must be made on a video taken of the final element that will be subtitled.
Engraving, inlaying or packaging
Engraving (silver film), inlay (video) or packaging (digital cinema, DVD) is the last step.
- In silver film with few copies, subtitles are engraved on film by a laser system, according to the final post-time simulation codes. The laser has replaced the old chemical process still used in some countries where subtitles pierce a layer of paraffin, and where the emulsion is burned by acid.
- In silver film with a high number of copies, laser process is too expensive and is used the optical subtitling, where subtitles are engraved on a black band that is over-impressed with the copy in the draw.
- In digital film (Digital Cinema Package, DCP), digital copies are packages of computer files, some of which contain subtitles. These subtitle files are loaded when projecting into the cab of the theater by the D-Cinema player that interprets and generates display on the image that is transmitted to the digital projector.
- In DVD-Video or Blu-ray Disc, no mechanical keying operation is required. During a DVD manufacturing phase called authoring, the subtitle tracks are added (up to 32).
- In television, one option is to embed subtitles on the master (ready to broadcast, PAD), which is used during diffusion.
- In television with teletext subtitling, standards are stricter (40 characters per line, some characters are not allowed). The subtitles are broadcast, as appropriate in different languages separately from the image and sound. There you need a teletext decoder in the TV to view subtitles (the inlay is done at the reception).
- In digital TV with DVB subtitling, the channel broadcasts the subtitles coded into the hidden lines of the digital video signal (vertical blanking interval). This information is read by the operator of the DVB encoding platform (operator of digital television headend) at an equipment called DVB encoder that compresses the video and audio signal to the MPEG standard and generates images subtitles that are transmitted in a component of the DVB resulting stream. These images are received and displayed (or not) by the terminal equipment of the viewer (digital TV receiver DVB).
The future of the broadcast subtitles in the field of television will require the gradual disappearance of the teletext broadcast or overlay in the image on the broadcast. The spread is now almost entirely digital, simply adding as many subtitles components in the data flow of the chain as there are versions of subtitles (like a DVD). The viewer are able to choose its version (for deaf and hearing impaired, or a specific language) and DVB terminal superimposed on the images of the video images of subtitles. We thus find the flexibility of teletext, with the added quality of the subtitles embedded in the distribution (each chain remains master of typography, colors, etc …)
The live broadcast
In some cases and for different reasons (cost, availability of copies, etc.), the desired subtitles are not present on the broadcast media. Generally referred to surtitrage to describe the fact of projecting additional subtitles live. This is a common practice in the festival for very recent movies and some cinema rooms accustomed to the distribution of rare or old films, difficult to access subtitled version in the national language.