Multi-Language Video Subtitling is a type of audiovisual translation often referred to as "subordinate translation." This is because the translation process is constrained by specific technical parameters of the project, including but not limited to the number of characters allowed per line and per subtitle, minimum and maximum on-screen durations, and reading speed, which refers to the average number of characters per second that the audience is able to read. These technical parameters are time-and-space restrictions that affect the final result on different levels.

Additionally, the translation of audio-visual files is influenced by two sources of information that add meaning and context to dialogue -- content that is heard (audio) and content that is seen (video). This can be both beneficial and problematic to the process. On the one hand, audio and video content provides extra-linguistic pieces of information that are often crucial for understanding some dialogues, references, concepts, puns, and jokes. On the other hand, due to the project’s parameters, there are cases in which it is difficult to include everything that is heard in the audio or seen in the video. The subtitling team may encounter spoken content that is very fast paced in the video or long passages of text, notes, headlines, etc., appearing on screen in a short period of time. When this happens, it is almost impossible for the team to include all of the information that was conveyed without violating the time-and-space restrictions mentioned above. For these reasons, subtitling is a daunting task.

Below, we will discuss important elements of a multi-language video subtitling project.

Project Evaluation:

In order to evaluate your project, you must check all variables. These include duration, video quality, number of speakers, intelligibility, number of files, target language and on-screen specifications. Durations of previewed videos in a file sharing service such as Dropbox can be different (usually shorter) than actual durations when you download the files and open them. Therefore, make sure you download, open and check the duration of each file so there are no surprises when you are wrapping up the project.

Team Assembly/Connection:

Before you consider how your team will look, you must have an idea of what it will take to complete the work. This means breaking down the project into stages and considering particularities such as deadline, budget, specifications, etc. It is advisable to work with experienced providers. You will also need to connect team members so that they may share terminology, and scheduling information such as delivery timelines, etc. The team will also need to be in touch with each other in order to standardize parameters as well.

Proposal:

When you prepare your proposal, you must consider the output of the video. A transcript corresponding to one hour of video will yield approximately 10,000 words. You will need to estimate the time spent doing each portion of the project.

Steps:

1. Preparation of the Transcript
This step comes first because the audio content in each video needs to be recorded in print before anything else can be done with the project.

2. Segmenting the Transcript
The segmenting process consists of breaking down the transcript and any relevant existing on-screen text into smaller sections or segments, following syntax and grammar rules to add line breaks at logical places in order to ensure and improve readability. Depending on the project and the purpose of the videos, these segments can have a maximum of two lines and their length may vary from 29 characters to 42 characters per line, including spaces, for a total amount of 58 to 84 characters per segment or subtitle.

3. Spotting
After the segmenting process is complete, all of the segments are introduced into a specialized computer software application where the time-codes are created and formatted so that the final spotted file can be used as a subtitling file. These time-codes indicate when the subtitles should appear and disappear, as well as other technical elements, such as their position if there is on-screen text overlapping. Time-codes are manually adjusted to ensure the best possible accuracy. At this point, the segments are ready to be translated.

4. Translation of Segmented Transcript
In the translation step, it is important to make sure that the text flows naturally and that it respects the style and linguistic features of the original text, such as register and language conventions. Also, due to technical requirements and specific parameters such as number of characters per line and reading speed, the translator always has to try to summarize the spoken dialogue in the most efficient way to create concise and easy-to-read subtitles. Some say that subtitles represent only two thirds of the spoken dialogue. However, at times – for example, when subtitling a training video that might contain very long nouns, omissions might lead to misunderstandings. The subtitling team must manipulate the time-codes and try to include as much information as possible (except for redundancies and hesitations). Nonetheless, as a general rule, the translator needs to summarize, depending on the specific parameters of the project, for the audience to be able to read and have access to all of the meanings rather than include every word.

5. Simulation
The simulation phase consists of ensuring that the translated subtitles in the subtitled video file meet all of the project’s requirements. In this step, the final deliverable is reviewed. Time-coding and synchronization are tweaked, as necessary, during the proofreading and editing of the text.

6. Burning Subtitles
Finally, after the translated subtitles have been reviewed, proofread and edited, it is time to insert them into the video. When it comes to subtitle output, there are two possible methods. The first is the hardcoding or hard burn method, which writes the subtitles on top of the image so they will be permanently displayed over the video. The second method is the soft burn, which allows you to turn the subtitles on or off as required, as they will be a separate selectable track in the output file. This second method allows you to have multiple tracks available for subtitles in different languages.

7. Uploading/Sharing
The final translated videos can be shared through Dropbox, Google Drive, OneDrive, WeTransfer or any other cloud-based file transfer service, or via an encrypted file-transfer service, such as Firefox Send.

In closing, it must be said that certain characteristics must be present for the successful execution of a subtitling project. They are commitment to quality, respect for deadlines, constant communication and flexibility. Add these to a clear working knowledge and understanding of each respective piece of the project, and success will be inevitable.



Co-authored by Andrea F. González Ocando and Deborah Avril Lockhart. Copy edited by Carol A. Huie.