Speech Analysis

Search
Adobe Premiere Pro User Guide

On this page

Applies to: Adobe Premiere Pro

某些 Creative Cloud 应用程序、服务和功能在中国不可用。

Removal of Speech Analysis | Premiere Pro CC 8.2

Starting with the December 2014 release of Premiere Pro CC (8.2), the Speech-to-Text feature or Speech Analysis functionality is removed from Premiere Pro.

The Speech Analysis functionality in Premiere Pro analyzes speech and converts spoken words into text-based, searchable metadata.

At the time of implementing Speech Analysis, it was anticipated that the underlying engine for this feature would improve at a faster rate than it did. Over time, user feedback and research confirm that this feature is not feasible to a degree of accuracy that is useful to users.

Given this focus, Speech Analysis is removed in Premiere Pro. That is, you can no longer analyze any new media and generate speech-to-text metadata in Premiere Pro.

You can, however, continue to view the speech-to-text metadata for any clips that were previously analyzed using Premiere Pro CC 8.1 or earlier versions. The generated speech analysis metadata is embedded in the media files, so the text continues to be visible and usable when the analyzed clips are imported and used in the latest version of Premiere Pro.

FAQ

The last version of Premiere Pro with the Speech Analysis functionality was the October 2014 release of Premiere Pro CC 8.1.

Adobe Media Encoder and Adobe Prelude support the Speech Analysis functionality. Starting with the December 2014 release, Speech Analysis will no longer be supported in Adobe Media Encoder CC (8.2) and Adobe Prelude CC (3.2).

In Adobe Media Encoder CC 8.2, you can export speech XMP metadata for any clips that were previously analyzed using Premiere Pro CC 8.1 or earlier. Select Export Master Speech Track And Sequence Markers in the Preferences dialog to include the speech XMP metadata while exporting.

In Prelude CC 3.2, Speech Transcription markers are removed from the Marker Type panel. However, when you open a clip that has Speech Transcription markers created from a previous version of Prelude, the markers reappear.   

Adobe is currently exploring potential alternatives, however there are no plans or guidance to share at this time.

Using Speech Analysis | Premiere Pro CC 8.1 and earlier

The Speech Analysis functionality can analyze spoken words and generate text metadata. You can edit and search the generated text metadata like any other metadata properties. You can also navigate to the times at which specific words are spoken, to better align edits, advertising, and subtitles.

In Premiere Pro, Speech Analysis is available for clips when there is at least one component audio clip in the merged clip. You can choose to analyze all audio tracks in a merged clip, or a single one.

  1. (Optional) Before you begin analyzing your media, switch your workspace to Metalogging (Window > Workspace > Metalogging).

    The Metalogging workspace layout is designed to help you easily see and work with the metadata associated with your audio and video files.

  2. Select a file or clip. You can analyze multiple merged clips, multiple clips, or a combination of both.

  3. To analyze a single track for speech, select the clip that you want to analyze and then click Analyze in the Speech Analysis section of the Metadata panel.

    Note:

    The Speech Analysis section is at the bottom of the Metadata panel.

    To analyze multiple or merged clips, select the clips, and then choose Clip > Analyze Content.

  4. In the Analyze Content dialog, set the following options.

    Language

    Select the language in which you want to create the text metadata.

    The speech analysis feature can use any of several language-specific and dialect-specific libraries, such as libraries for Spanish and UK English.

    To download more language-specific and dialect-specific libraries, see this link on the Adobe website.

    Quality

    Choose either High or Medium quality.

    Script

    You can improve the quality of speech analysis by using a reference script or an embedded Adobe Story script.  

    Identify Speakers

    Select this option if you want Premiere Pro to create separate speech metadata for each speaker.

  5. Click OK. The spoken words appear as speech metadata in the Speech Analysis section.

  6. To retain the text in the speech metadata, save the project.

  7. You can copy text from speech metadata to the clipboard for use in a text editor. Right-click the transcript, and choose Copy All

Note:

If you import files with a speech metadata into After Effects, each word appears as a layer marker on layers based on these footage items.

In the Speech Analysis section, do any of the following:

  • To correct a word, click it, and type.
  • To insert, delete, merge, cut, or copy words, right-click an existing word, and choose a command from the context menu.

In the Speech Analysis section, select the word. Timecode In and Duration indicate the precise location and length of your selection.

To hear the selection, click either Play or Loop. The Loop playback option repeatedly plays the selected word, with some preroll and postroll.

Accuracy of speech-to-text conversion depends on the clarity of the spoken words and the quality of the recorded dialog. Dialog recorded in a noisy environment or with poor microphone placement cannot produce highly accurate results even with a reference script. You can nevertheless use a reference script to improve speech analysis.

A reference script is a text document containing dialog similar to the dialog recorded in your assets. There are two types of reference scripts:

  • A script that contains similar dialog, but was not necessarily written for the current project. For example, a series of medical training scripts for different products can be combined into a single text document. You can use this text document as a reference script. With this type of reference script, speech analysis produces results more accurate than it does when using only the default language models.
  • A script that matches the recorded dialog. This type of reference script provides the highest accuracy possible. For example, you can use the script that the talent read during the shoot as a reference script. Alternatively, you could use a transcript typed from the assets for the purposes of close captioning.

Attach a reference script

Speech Analysis supports reference scripts only in the UTF-8 encoded text format, including Adobe Story scripts, which have the .astx filename extension.

Note:

The closeness of the match between the embedded script text and the recorded dialog determines the accuracy of matched-script text. If 100% accuracy is important, edit and revise the script text first. Ensure that the script matches the recorded dialog before using it as a reference script.

  1. In the Analyze Content dialog box, from the Reference Script menu, select Add.
  2. Browse to the reference script text or .astx file, select it, and click Open.
  3. In the Import Script dialog box, type a name for the reference script, and select the language of the script.

Note: You can view the text of the file in a scrolling window.

  1. Select Script Text Matches Dialog only if the imported script covers the recorded dialog verbatim. For example, if the reference script is the script from which the talent read their lines, select Script Text Matches Dialog. Select this option even if the recorded dialog is shorter than what the script file covers.
  2. Click OK.The Import Script dialog closes, and the reference script is selected in the Reference Script menu.
  3. Click OK in the Analyze Content dialog.

Speech analysis is more accurate if Adobe Story script data is associated with a clip. Then, Adobe Premiere Pro automatically uses the Adobe Story script as a reference script.

When Adobe Premiere Pro finds enough matches with the embedded script, it replaces the analyzed speech text with the embedded script text. Adobe Premiere Pro carries over the correct spelling, proper names, and punctuation from the reference script, benefits that standard speech analysis cannot provide.

Note:

You need to match the scene number of the clips to the scene numbers in the Adobe Story script. Adobe Premiere Pro needs the information so that it can match the clips to the correct scenes in the script. You can assign scene numbers to clips in the Project panel or the Metadata panel.

  1. Select one or more clips in the Project panel, right-click, and choose Attach Script File or select File > Adobe Story > Attach Script File.

Note: An Adobe Story script cannot be attached to a Merged Clip. If, however, a Story script was attached to a component clip before the merge, the merged clip may be analyzed using the previously attached Story script.

  1. In the Analyze Content dialog, under the Script section, select Use Embedded Adobe Story Script.

When a clip has a Story script attached to it, the script is displayed in the Embedded Adobe Story Script view in the Speech Analysis section of the Metadata panel. Compare the script displayed in this view to the results of the speech analysis displayed in the Analysis Text view below it.

Note: The Embedded Adobe Story Script view is read only. You can't perform editing operations in it.

You can add In and Out points to speech analysis text to select a portion of a master clip. You can insert or overwrite the selected portion directly from the Speech Analysis section of the Metadata panel.

  1. In the Speech Analysis section of the Metadata panel, select a word.

  2. Do one of the following:

    • To set the word as an In point, click Mark In.
    • To set the word as an Out point, click Mark Out.

    Premiere Pro sets In points at the beginnings of words, and Out points at the ends of words. Premiere Pro highlights the range between the In point and Out point in the Speech Analysis section of the Metadata panel.