Scripta is the latest in-house transcription tool at Ai-Media and is available for Transcription and Timed Transcription tasks on Community Portal.
Scripta has been created with over two decades of transcription knowledge in mind. Its features have been designed to make working as simple as possible so that you can focus on what’s really important – accurate transcription!
Claim a Transcription or Timed Transcription task and click ‘Launch Scripta‘. Scripta will open in a new tab so make sure you have pop-ups enabled.
Scripta will launch with the automatically generated transcript (ASR) pre-loaded and split into segments ready for you to start editing. If someone has worked on the task before you claimed it, their edits will be visible. You can revert to the original ASR by clicking Reset at the top of the editor.
Scripta has a variety of controls that can be triggered by using your keyboard. These controls are also known as hotkeys. Find a full list of Scripta controls here or see below for the most commonly used hotkeys.
|Tab||Play and pause the media|
|Ctrl + Alt||Skip back in the media|
|Shift + Alt||Skip forward in the media|
|Esc||Set start timestamp|
|F2||Set end timestamp|
– At end of segment
– At start of segment
– In blank segment
– In middle of segment
– Create new segment below
– Create new segment above
– Rotate segment speaker label
– Split segment at cursor
|Backspace||Merge segment up |
(with cursor at beginning of segment)
|Delete||Merge segment down|
(with cursor at end of segment)
|Shift + Backspace||Delete segment|
(with cursor inside segment)
|Alt + S||Save work|
Editing segments in Scripta is as simple as clicking into a segment and starting to type! Segment breaks are required every 8-10 lines and should be broken at a natural point in the speech. To indicate a segment break, please use > as the speaker label. See an example here.
- Ctrl+Z will undo text edits
- Ctrl+J will jump the media to the start timestamp of the segment your cursor is in
- Be careful not to press Shift+Backspace as this will delete the entire segment your cursor is in
The Text Tracker is an awesome feature exclusive to Timed Transcription tasks. It works by using automatically generated timecodes to follow along with each word in a segment as it is spoken. You can double-click on a word to jump to that part in the media. Pretty handy!
Editing speaker labels
Speaker labels can be found to the right of the media player in Scripta. Speaker labels in Scripta are generated automatically by our Recorded Speech Processing Service. It assigns the speaker as either a male or female and numbers them sequentially based on how many speakers it identifies.
Speaker labels must be edited to meet Ai-Media standards.
You can click the small x to clear a speaker label and type your own. Once you have edited a speaker label, it will be available to use for subsequent segments. Just click the small arrow to choose the speaker label from the list.
To edit all instances of a speaker label, simply click into the box without clicking the small x, type your edit and press Enter. This will change all occurrences of that speaker label. This is useful if you need to change the spelling of a name or if you wish to change all the autogenerated speaker labels to LECTURER, for example.
You do not need to add colons to speaker tags in Scripta. This is because they are added automatically when a task is submitted by the processing service. For example, SPEAKER will become SPEAKER: once the task is submitted.
If you’re working on a Transcription task, the timecodes for the segments will be 00:00:00 and you don’t need to edit them.
If you’re working on a Timed Transcription task, segments will have automatically generated timecodes. If you create a new segment, it will have timecodes of 00:00:00 that need to be timestamped. These must be set within two seconds of the audible content.
To set a timestamp, navigate to the point in the media where the audible content begins or ends. Press Esc on your keyboard to set start timestamps. Press F2 on your keyboard to set end timestamps.
- Keep a cheat sheet of controls next to your computer (print off our page or create your own handwritten one!)
- Double-click on a word to jump to that point in the media (please note this only works for Timed Transcription tasks with automatically generated timecodes)
- Set the end timestamp on the current segment before setting the start timestamp on the next segment to avoid overlaps
- Always edit 00:00:00 timestamps for Timed Transcription tasks
Practice makes perfect! If you want to practice using Scripta, follow the instructions in this article. Don’t forget to unclaim the task when you’re finished practising so someone else can work on it.