Editing a Transcript

Once a transcript is completely created, you can edit it. The transcript portions already shown for EWC clips or very long clips stay in read-only mode until the entire transcribe operation successfully finishes.

Avid Ada Transcribe uses an Artificial Intelligence (AI) / Machine Learning (ML) model to produce transcripts and speaker labels from assets that include non-musical, spoken word. In terms of a newsroom workflow, this might be identified as a VO or "talking head". As with all AI/ML models, results might be incorrect as the model is predicting the output from the provided input. Therefore, the Transcript tab provides options to correct any of these errors.

For more information on these editing options, see the following topics:

c All edits you make to a transcript will be overwritten if a re-transcribe job for the same audio track is triggered.

Renaming Speakers

When a transcript is created, each speaker is identified by an ID and a name "Speaker <number>". You can edit each speaker name to make it a person's name. In the following example, Speaker 2 will be renamed to John.

To rename a speaker:

  1. Right-click a speaker name in the Speaker column.

  2. In the context menu that appears, select Rename.

    The speaker name field becomes a editable field. The speaker name is highlighted.

  3. Type the new speaker name.

  4. Click outside the field.

    All instances of the speaker name in the transcript are updated. This applies also to the speaker name in the Change Speaker control.

    If you deleted the entire speaker name while editing, the previous name is restored when you click outside the field.

Creating New Speaker

When you inspect a transcript and identify a speaker that is not included in the speakers list, you can add the missing speaker name.

To create a new speaker:

  1. Do one of the following:

    • Right-click a segment or a speaker name.

    • Select word(s) in a Transcript segment and right-click the selection.

  2. In the context menu that appears, select Add New Speaker.

    The Add New Speaker dialog box opens.

  3. Type the name of the speaker to be created in the Speaker field.

  4. Click the Add button.

    If a speaker with the same name already exists in the transcript, a corresponding message opens. Click outside the message, provide a unique name in the Add New Speaker dialog box and click the Add button again.

    The Add New Speaker dialog box is closed. Depending on the source on which the Add New Speaker action was started (entire segment or selection of words), the following happens:

    • The segment is assigned to the new speaker.

    • A new segment with the selected word(s) is created for the new speaker.

    The new name is shown in the Speakers column.

Assigning Transcript Segment to Another Speaker

If you find out that a portion of the transcript is not assigned to the right speaker, you can correct this. You have the following options:

  • Assign an entire segment to another speaker.

  • Assign one or more words of a segment to another speaker.

To assign a segment to another speaker:

  1. Right-click a segment that you want to assign to another speaker.

  2. In the context menu that appears, select Change Speaker > <speaker name>.

    The segment is assigned to the new speaker; the name is updated in the Speakers column. If the segment above or below are already assigned to the same speaker, the segments are merged; the start timecode of the first segment is kept.

To assign individuals words of a segment to another speaker:

  1. Scroll to the segment that contains words to be assigned to another speaker.

  2. Select the word(s) to be re-assigned:

    1. Click on a word to be re-assigned.

      The selected word is highlighted.

    2. To create a selection of words, press the Shift key and do one of the following:

      • Click on a word to the left side (the initially selected word will be the last word of the selection)

      • Click on a word to the right (the initially selected word will be the first word of the selection).

  3. Right-click the selection.

  4. In the context menu that appears, select Change Speaker > <speaker name>.

    The selected words are assigned to the new speaker.

    Depending on the time stamp of the words, the following happens:

    • The words are added to a previous segment of the speaker.

    • The words are added to a following segment of the speaker.

    • A new segment with the words is created for the speaker.

Editing the Transcript Text

You can edit the transcript text, for example, to correct wrong names.

Note the following limitations:

  • You can edit only one word at a time.

  • You cannot delete a word.

  • You cannot add and remove blank space between characters or words.

    As a side effect, you cannot modify a word containing an apostrophe when the change would introduce a blank space. For example, changing "she's" to "she is" would be denied.

When you enter the Edit transcript mode while the asset is played back in the Media Viewer, playback will be stopped; the Synchronize Playhead feature is ignored even if the toggle is enabled.

To correct words in the transcript:

  1. Click the Edit Mode button so it turns blue.

  2. Scroll to the segment that contains text you want to modify.

  3. Double-click the word you want to modify.

    The word is highlighted and underlined and enabled for editing.

  4. Type the new text.

  5. Press Enter or click outside the field.

    Your change is applied.

    If your changes cannot be applied – because, for example, you added a blank space or deleted the word – the following message is shown on the right side of the Fast Bar. You can click the Try again link to repeat the update request. The message disappears after a few seconds.