What is Docsoft?

  • Docsoft: AV, accessed via http://docsoft.ua.edu, automates speech-to-text conversion and exports transcripts/captions.
  • It is available for UA faculty, staff, and students. Currently, it is being used by campus stakeholders who already have been producing captions and transcripts. Students who need captioning services should contact the Office of Disability Services.
  • Available transcript types include Encore Subtitle Script, ESX File, Native transcript (required for use with Docsoft Transcript Editor software), Quicktime text, RealText, SAMI, Scenarist Closed Caption, SubRip, Text, or Timed Text (DFXP). These are connected to the media file using an external tool.
  • Users create profile and vocabulary files for individual speakers to improve accuracy of the automated transcription.
  • Docsoft must be accessed from on campus location or via UA’s VPN.

How do I get a Docsoft Account?

  1. At docsoft.ua.edu, click ‘Register Account’.
  2. Enter an account name. It is preferred for this to be your myBama username. However, if a departmental or generic account is needed, please use an account name that identifies the department or office.
  3. Enter and confirm a password that is at least 7 characters long and includes at least 1 non-alphanumeric character.
  4. Enter a valid email address, preferably a University of Alabama address.
  5. Enter a security question and answer.
  6. Click ‘Create User’. At this point, the Docsoft administrator will be notified that an account is ready for activation. You will be contacted when the account is ready to use.

How do I create a speaker profile?

You will need a 3-5 minute recording of the speaker and a text file containing an exact transcription of the recording. You may want to record the speaker reading text as a quick way to produce these crucial pieces. Sample reading.

Video: Create and train a speaker profile (opens in new window)

  1. Log in to docsoft.ua.edu.
  2. Click on ‘Profiles’.
  3. Click ‘Create New…’
  4. Enter the speaker’s name and a short description.
  5. Choose a base acoustic model. The ‘Base Acoustic Model’ selections allow for various language and quality settings. Clean, clear, and well-recorded audio will always mine more successfully. Change this setting if your audio better matches a different selection.
  6. Decide whether others will be allowed to use the profile.
  7. Click ‘Next’.
  8. Enter a name for the speaker’s vocabulary and choose a base vocabulary that is most appropriate.
  9. Click ‘Finish’.
  10. Click ‘Train’ to add media and transcript files.
  11. Select a media file and a 100% accurate (or as close as possible to 100%) transcript then upload both files to Docsoft. You can do this multiple times to speed up the training process. After adding all desired the training, select ‘Finish’ and either ‘Monitor events for completion’ or ‘View all profiles’.

How do I submit media files for transcription?

Video: Process a media file (opens in new window)

  1. Log in to docsoft.ua.edu.
  2. Click on ‘Uploads’.
  3. Choose the appropriate speaker profile and vocabulary.
  4. Browse to the AV file.
  5. Add any additional or related text (.doc, .txt, .rtf, .htm, .html, or ANSI text).
  6. Click ‘Submit’.
  7. Monitor the status of your transcription process from ‘View Files’.

How do I access my transcript or caption file?

  1. Log in to docsoft.ua.edu.
  2. Click on ‘View Files’.
  3. Click on ‘Export Transcripts’.
  4. Choose your preferred transcript type and style.
  5. Click ‘Finish’ to save the file locally.

How do I edit my transcript or caption file?

There are many tools that can be used to edit transcripts, depending on where and how the media file will be accessed.

  • Text editors (Mac- TextEdit, PC- NotePad): Only some transcript types can be opened and edited with a text editor. Within the text editor, you can correct words, capitalization, and punctuation. Do not alter line numbers or time stamps. Use outside media player (eg. iTunes) to listen to the audio while editing.
  • Docsoft TE: PC Only. Requires Docsoft ‘native transcript’ type and media file. Allows for line break changes, automatic and editable time stamps, word edits, export into multiple caption types. Includes embedded media player. Contact rsthompson2@ua.edu for more information.
  • Other tools: Most commonly-used video editing tools. MovieCaptioner is used by several other schools.

Captioning Resources (links open in new window)

Captioning Tegrity Recordings via Docsoft

Video: Caption a Tegrity recording (opens in new window)

  1. Create a recording in Tegrity and upload it for processing.
  2. Save Tegrity podcast MP3: Navigate to the course containing the recording in Tegrity. On the Course Tasks menu, choose Podcast. This opens a list of podcast MP3s. Save the MP3 to your computer. Note: The MP3 will have a non-intuitive name; Consider renaming it when you save locally.
  3. In Docsoft, upload the MP3 file for speech-to-text processing: Choose Upload AV files. Select the speaker profile. Browse to the MP3 file you saved from Tegrity. Also add any related text files you may have, such as speaker notes or summary. Choose Submit. The file will be scheduled for processing and its status will update in the View Files list.
  4. Export SubRip transcript file: Once Docsoft has processed the MP3 file, check the box next to the file name and click Export Transcript. Choose the transcript type SubRip and click Finish to save the .SRT file where you want on your computer.
  5. Edit the transcript (.SRT file): Play/pause the MP3 file via iTunes or another tool on your computer. Open the .SRT transcript file in a plain text editor, such as TextEdit (Mac) or NotePad (PC). Do not use Microsoft Word or WordPad. As you listen to your MP3 file, correct the transcript text. Do not alter the time stamps or line numbers. Save the edited .SRT file.
  6. Add Closed Captions to Tegrity recording: Navigate to the recording in Tegrity. Check the box next to the recording. Under Recording Tasks, choose Edit Recording. Choose Add Closed Captioning. Select Closed Caption File option appears. Select the edited .SRT file. If the file is accepted, a green dot appears. Choose Add Closed Captions. A confirmation message appears when the captions are added successfully.