A transcript must be created before the closed captioning process can begin for pre-recorded material. While listening to a video, a typist can easily transcribe the video using any standard word processor. Since closed captioning software is designed to format and timestamp the transcribed text for TV broadcast, the transcription process is done faster outside the captioning software.
The transcription of the video can be done by listening to a small video or audio file such as .MP3. This article describes several ways to do transcription.
V-Pedal
The V-Pedal is a device that controls the playback of video and audio files on a computer using a push-down foot pedal. This device is a time saver because it allows the transcriber to easily pause and rewind the video without having to take their hands off the keyboard. V-Pedal devices can connect via a USB port to either Mac or Windows software.
Speech to Text Technology
There are several ways that speech-to-text software can speed up the transcription process. The first is a fully automated solution such as MacSpeech Scribe which is software that generates a transcript from an audio file. The software can read the audio waveform and try to match the sound to words in a dictionary. The results vary and typically require the user to proofread and correct mistakes.
Another option is to train the user’s voice with Dragon Dictate, and repeat what is being said in the video. The result is a plain text document. This is a good solution because it is very accurate. Punctuation still will need to be corrected.
Recently, mobile phones have been equipped with speech to text technology that can be used to create a transcript for captioning. Some users find it practical to simply speak into their mobile phones while watching a video to create the transcript for captioning.
Italics and Bold Text
When creating a transcript it is best to avoid using italics and bold. Italics in closed captioning generally mean that the speaker off-screen. Bold is not supported in TV closed captioning, and could result in text blinking on and off. TV sets and cable boxes do not have the ability to bold text, so we recommend avoiding bold text.
Changing Speakers
When importing a transcript into captioning software, the captioner looks for a change in speakers to start a new caption block. A good captioning project should have an indicator for the viewer to show a change in the speaker. The transcript can help this effort by placing a symbol before the speaker change. The symbol “>>” typically shows this change.
Tab-delimited
If line breaks or timecodes are critical for the captioning project, a simple tab character can be placed in between the text. In the same way, spreadsheet software can separate cells, captioning software can do so when it detects a tab character. For example, if a sentence is long, a tab character will put a line break in the text when available.
File Formats for Text Documents
Although captioning software can accept a variety of text files such as .DOC, RTF, and .TXT, the recommended file format is a plain text document. Many word processors may include extraneous data in native file formats such as margins, footnotes, and other non-text related code. Most word processors can save the document as a plain text file.