PRINT

GENERAL RULES

---

The General Rules list condenses the segmenting guidelines for easy access. For more detailed information about segmenting, please read the corresponding sections in the Segmenting Guide.

Timing: 

• The segment beginning boundary line must be within 0.1s before speech starts:

​Length: 

• Always try to divide your segments by person and by sentence, avoiding placing two people in a segment at all costs. If it is not possible to include an entire sentence in a segment, try to split the segment according to breaks in speech.
• Avoid creating segments that are less than 1 second long.
• Segments should generally not be more than 6 seconds long; however, there are exceptions. Always take into consideration how much, how fast, and what is being said.
• Combine shorter segments by the SAME speaker.
• Combine segments by DIFFERENT speakers when their speech is significantly overlapping or a speaker's segment is less than 1 second.
• Meaningless interjections/expressions (e.g., um, uh, hm) should generally not be segmented. However, if a interjection/expression is followed by a speaker's line, then it can be combined in the same segment (e.g., "Hm... I understand now.") in order to give the segment additional reading time or to prevent placing two people in a segment.

Extensions: 

• Segments must extend 1-1.2s before an empty section:

Gaps: 

• Close gaps that are 0.5s or less:

​Text: 

• Segment meaningful on-screen text
• The segment must begin when text appears or is focused on by the camera
• End the segment when text disappears, or, if there is not enough time to read, add an extension

Songs: 

• Segment songs that are audible and can be subtitled without making any speech/text subtitles difficult to read
• The segments for songs can generally be longer than 6 seconds because the words tend to be lengthened

​

ERROR TYPES AND HOW TO FIX THEM

---

The following list includes common errors and instructions on how to fix them.

0.Xs Late
To fix: CTRL/CMD + Left Arrow X times

BEFORE CORRECTION:A segment is late when the voice begins before the segment appears. When paused at where the first hint of voice is heard, the green line is to overlapping or to left of the segment starting line.

AFTER CORRECTION: The segment is adjusted deciseconds earlier to resynchronize to the voice. Now when paused at the first hint of voice, the green line is just right of the yellow line. The segment is synchronized.

0.Xs Early
To fix: CTRL/CMD + Right Arrow X times

BEFORE CORRECTION: A segment is early when the beginning is trimmed too far to the left of the base of the wave. The subtitle appears before voice is heard. When paused at where the first hint of voice is heard, the distance between the yellow line and the green line is greater than 0.1 seconds.

AFTER CORRECTION: The segment is adjusted deciseconds later to resynchronize to the voice. Now when paused at the first hint of voice, the green line is just right of the yellow line. The segment is synchronized.

Ending Too Short
To fix: CTRL/CMD + Up Arrow

BEFORE CORRECTION: Here, there is not enough extra length added after voice ends. The segment will be too short for the audience to read the subtitle well.

AFTER CORRECTION: The segment is lengthened so there is a 1.0 to 1.2 second extension after voice ends.

Ending Too Long
To fix: CTRL/CMD + Down Arrow

BEFORE CORRECTION: Here, there is too much extra length added after voice ends. The segment will be too long for the audience and will bore/annoy them.

AFTER CORRECTION: The segment is decreased so there is a 1.0 to 1.2 second extension after voice ends.

Split Segments
To fix: Spacebar to cut

BEFORE CORRECTION: Split long segments, in general those that are longer than 5-6 seconds.

AFTER CORRECTION: Split the segments according to breaks in speech or between the speaker’s sentences.

Combine Segments
To fix: Delete (Del) the second segment and extend the first (CTRL/CMD + Up Arrow)

BEFORE CORRECTION: Combine shorter segments by the same speaker so that they are, in general, no greater than 6 seconds in length. You may also need to combine shorter segments by different speakers when their lines are too short to read separately.

AFTER CORRECTION: Here, the segments are combined.

Add a Segment
To fix: Press Spacebar to start and end the segment

BEFORE CORRECTION: Segments that are missing for meaningful text or dialogue/speech should be added. For instance, in the sample above, there is a missing segment for the episode number.

AFTER CORRECTION: We must segment to the slight appearance of the hardsub and end the segment when the hardsub disappears for perfect on-screen appearance.

Delete a Segment
To fix: Select (S) segment and press delete (Del)

BEFORE CORRECTION: Segments without meaningful speech or text should be deleted. For example, the segment above didn’t have any words but only noises made by a speaker after being injured.

AFTER CORRECTION: The segment is removed.

Close Gap
To fix: Select the first segment, then click CTRL/CMD + Up Arrow until the gap is closed

BEFORE CORRECTION: Close gaps that are 0.5 seconds or less. Here, there is a mini-gap.

AFTER CORRECTION: Close these gaps to reduce “flashing.”

Remove Microgap
To fix: Use CTRL/CMD + Left and Right Arrow to glue the two segments’ ends together

BEFORE CORRECTION: When the green line passes the beginning of the segment, there is a flash.

BEFORE CORRECTION: The subtitle box disappears for a very short amount of time in between the segments.

AFTER CORRECTION: Once you move the end of the segment once (0.1s) each way, the segments will be glued together and there will no longer be a flash.

Add a Gap
To fix: Click Spacebar one second after voice ends then delete (Del) the empty segment.

BEFORE CORRECTION: Segments are in a continuous sequence, but the extension of the first segment is too long. Therefore, this subtitle will remain on screen for longer than necessary.

AFTER CORRECTION: One second after voice ends, press spacebar to cut and then remove the empty segment. The gap is larger than 0.5s so it is acceptable.