Dive has two types of reading features. Each also supports multiple languages (the narration voices support 65 languages; other languages are read aloud in English. For the list of supported languages, see Languages Supported in Automatic Translation).
① Read Text
When you open a step, the text in the Description field is read aloud. In addition to the main text, Work tips, Quality warnings, and Important safety information are also read aloud (which items to include can be switched in "Items to read aloud" below). For non-video steps, this reading is on by default.
Characteristic: Not synchronized with video, but requires no timing adjustment so editing is easy. Can be used even for non-video content.
If the reading takes longer than the video in the step, the video will repeat until the reading is finished. To eliminate repetition, see The same video in the exported MP4 repeats multiple times.
② Read Subtitles
Subtitles are read aloud in sync with their display timing. Only subtitles are read; Work tips, Quality warnings, and Important safety information are not read. When creating an SOP from a video, select the narration target in the "Step conversion settings" during creation. The default is "Subtitles" (the previously selected option is retained).
Characteristic: Synchronized with video, so understanding the narration is smooth
Selecting Narration Target
For each step, select one narration target from "None", "Description", or "Subtitles". You cannot read both subtitles and description in a single step. For steps where you want to read tips and warnings, select "Description". In this case, subtitles are displayed on screen but not read; instead, the main text and tips/warnings are read aloud.
In the SOP edit screen, configure as follows:
- Video step: Click the speaker button at the top that appears when you select a video step (showing the current setting as "Subtitles", "Description", or "None") to open "Common audio settings", where you can change the narration target for all steps created from the same video at once.
- Change only one video step: Turn on "Define individually" in "Audio settings" for that step, then select the narration target.
- Non-video step: Turn on/off "Read text" in "Audio settings" for that step.
- Change multiple steps at once: In the step list, hold down Ctrl (Cmd on Mac) or Shift and select steps, then click "Bulk edit" on the right. Select the narration target in the "Audio" tab and apply (opening from "Bulk edit" in the "Other" menu will target all steps).
Note: Subtitle Reading Speed
In subtitle reading, the narration speed (tempo) is automatically adjusted based on the subtitle display length (frame) during editing. When the audio does not fit in the frame, the speed increases and speech becomes rapid.
To minimize rapid speech, space is secured in the following order:
- If there is space between a subtitle and the next subtitle, it reads until just before the next subtitle begins
- If it still does not fit, the audio is created at a slightly faster pace when the voice is generated (the sound is clearer than speeding up the audio afterward)
- If there is a static graphic in a step, reading continues during the static period, which then restores the speed to normal
You can check which subtitles are rapid in the subtitle list in the video editor. Subtitles that do not fit in the frame display "Rapid speech ○x" (the value after speed is restored by the static graphic). If you want natural speech speed, shorten the sentences or add space between the next subtitle (the last subtitle in the step will have its end extended backward). For details, see When the step duration is short, subtitle narration is too rapid to understand.
Note: Narration Volume
Subtitle narration creates voice one sentence at a time and connects them. Before connecting, the volume is leveled sentence by sentence, so the volume does not vary by sentence within the same SOP. Even across SOPs, or when switching to a translated language, the volume remains nearly the same.
Note: Items to Include in Text Reading
In text reading, in addition to the main text, items such as tips, quality warnings, and safety are also read. If the reading becomes long, you can toggle which items to include in the "Items to read aloud" setting under Playback & display settings > Advanced display settings in the Basic information section of the SOP edit screen.
The main text is always read. Excluded items are removed from the audio, but screen display does not change. Switching changes the narration audio and regenerates it.
This setting only applies to steps where the narration target is "Description". For steps where the narration target is "Subtitles", only subtitles are read, so changing this setting has no effect. For steps where you want to read tips or warnings, use the "Selecting Narration Target" method above to change to "Description".
Note: Reference Language for Reading
Translation and reading are based on the Original language setting on the SOP. The default is Japanese.
For SOPs created in a language other than Japanese, set the original language in Playback & display settings in the Basic information section of the SOP edit screen. If you play without setting this and without translating, the narration voice may not be created correctly.
The default for the entire team can be set in "Language of SOPs you create from now on" in the Terminology Glossary (Team Administrator or higher). It does not affect already created SOPs.
Note: When Narration Audio Is Regenerated
Narration audio is regenerated by the following operations. When the narration specifications change, existing SOPs reflect the changes from the time they are regenerated.
- When you edit and save subtitles or text
- When you change the original language
- When you register in the Pronunciation Dictionary (changes to existing audio may take a few minutes)
- When you perform bulk text replacement
- When you switch items to read aloud