Dialog subtitles for vignettes, movies, and conversations

Subtitles for dialog requires more precise timing than for narration. You will need to devote more time and effort to creating subtitles for conversations or group dialog exchanges.

The first difference is to prepare your transcribed script differently for dialog than for narration. For narration (including documentaries and corporate), a blob of transcribed text is fine. This software will divide narration into small chunks for easy subtitling. With subtitles over narration, the subtitle text can remain on screen as long as required for the audience to read the text, because narration is more free-flowing and b-roll covers the gap giving a second or two leeway in timing with the on-screen image.

But for dialog, there is an increased need for exact timing. If what a character says appears on screen 2 seconds after it was spoken, another character may be speaking and the audience loses the thread of the story.

In order to achieve better subtitle timing for dialog, the transcription (original script) often needs to be split differently than the speech of a single narrator. The original director’s script may be an excellent start for this purpose. Although there will be notes and action descriptives, in a director’s script the dialog is generally well divided to form the basics of subtitling. However, if no original script is available, typing a transcription of a video should not be just a “blob” of text, but should be divided by character.

Dialog script transcription

Below are suggestions for pre-subtitling dialog transcription that include either designating a character name or simply designating “new speaker”. The character name will need to be removed before subtitling, so is optional during the transcription phase of the project. Transcription examples:

Example 1:
Person A: Hi, Joe how are you today?
Person B: I guess I’m okay.
Person A: Have you by any chance seen Hank? He was supposed to come by today.

Because designating the exact character (Person A and B) is wholly optional from a subtitling standpoint, during transcription it may be easier to use “>” to indicate a new speaker, as follows:

Example 2:
> Hi, Joe how are you?
> I guess I’m okay.
> Have you by any chance seen Hank? He was supposed to come by today.

Preparing dialog for subtitles

Once the script has been transcribed in the above manner, come back to combine very short phrases into a single line, unless you know that there is a significant pause. So, assuming that the dialog is non-stop, when you revise your script, it may look like this:

> Hi, Joe how are you? > I guess I’m okay.
> Have you by any chance seen Hank? He was supposed to come by today.

In this way, when you click in the software to assign timecodes to each subtitle, the two short ones will appear as 2 lines of the same subtitle:

> Hi, Joe how are you?
> I guess I’m okay.

The software will split the longer sentence, then remove the “>” because there is no need to designate “new speaker” within a single subtitle:

Have you by any chance seen Hank?
He was supposed to come by today.

Note that this software will auto-split long lines and paragraphs, so concentrate on short phrases and changes in speaker.

Designating a pause in dialog

If during the transcription phase the transcriber notes that there is, indeed, a pause, insert “>>”. In this way, a blank subtitle will be inserted by the software, so that during the timecode-setting phase, the subtitle will disappear from the screen. When you go back to combine 2 short phrases into one, then this same >> will also indicate to you that these particular phrases will not play well on one subtitle.

Example 3:
> Hi, Joe how are you?
>> I guess I’m okay.
> Have you by any chance seen Hank? He was supposed to come by today.

The software will treat the above as follows.   When you click, each subtitle group will be properly divided:

Hi, Joe how are you?
(blank)
I guess I’m okay.
Have you by any chance seen Hank? He was supposed to come by today.

3 phases to subtitle creation

All subtitles go through 3 phases:   transcription, timecode assignment to each sentence, and lastly an auto-split by the software to create short subtitles.   The last auto-split is where:

Have you by any chance seen Hank? He was supposed to come by today.”

will become:

Have you by any chance seen Hank?
He was supposed to come by today.