Adobe Soundbooth CS5 Speech Analysis

Written by seamus islwyn
  • Share
  • Tweet
  • Share
  • Pin
  • Email
Adobe Soundbooth CS5 Speech Analysis
Adobe Premiere also includes speech analysis support. (Justin Sullivan/Getty Images News/Getty Images)

Adobe Soundbooth CS5's speech analysis feature converts a voice recording into text, then embeds the text into the recording's metadata. You can use Soundbooth's speech analysis to create a transcription of a lecture, for example, or to transcribe the dialogue from a radio play recording. Soundbooth also lets you edit speech analysis metadata for accuracy, and offers downloadable add-ons for several non-English languages.

How it Works

Adobe Soundbooth CS5's speech-to-text feature analyses the waveform of a recorded voice, then marks each individual word in the recording. It then compares the word waveforms in the recording to a database file containing matching pairs of waveforms and text words. Soundbooth embeds the text words into the audio file's metadata field using timecodes to match the speech metadata entries with their corresponding waveforms. If you have a reference script that's either an exact transcription or a transcription of a similar type of voice recording, Soundbooth can use this script to create more accurate results.

Running Speech Analysis

Open the recorded vocal file in Soundbooth CS5, then click the "Metadata" tab in the left-hand pane of the Soundbooth window. Click the "Analyze Speech" button at the bottom of the pane. Select a reference script if you're using one, then choose either "High" or "Medium" from the "Quality" drop-down menu. Check the box next to "Identify Speakers" if you want Soundbooth to create a separate metadata entry for each individual voice on the recording. Click "OK" to analyse the speech and add metadata text to the file.

Editing Speech Metadata

Although Soundbooth CS5's speech analysis is fairly accurate, transcription errors may appear in the speech metadata. If Soundbooth incorrectly identifies a word, you can correct it by editing the speech metadata. Click and drag the top of the "Speech Analysis" section of the left-hand pane upward to expand it and view the transcription. Double-click an incorrectly identified word, then type the correct word into the box that appears. Right-click on a word, then select "Delete," "Merge With Next Word" or "Merge With Previous Word" to perform these editing functions.

Additional Languages

The default Soundbooth CS5 installation only includes speech analysis support for American English. If you need to transcribe audio in another language, download its language pack from Adobe's website. Install the "International English" speech analysis pack to transcribe recordings of non-American English speakers. Adobe also offers language packs for recordings in French, German, Italian, Japanese, Korean and Spanish. Language packs are version- and platform-specific: be sure to download the correct pack for your operating system. If you've upgraded from Soundbooth CS4 to Soundbooth CS5, you'll need to upgrade the language packs as well.

Don't Miss

  • All types
  • Articles
  • Slideshows
  • Videos
  • Most relevant
  • Most popular
  • Most recent

No articles available

No slideshows available

No videos available

By using the site, you consent to the use of cookies. For more information, please see our Cookie policy.