Understand how to amend or delete caption and chapter files in your WorkCast on-demand events
Quick Links
Create chapters for your on-demand media
How to edit or update your captions or chapters
How to edit or update your captions or chapters
Although we use the most advanced automatic speech recognition software, on occasion it can generate some words or a Company name wrong, you can edit the generated captions/chapters at any time. You can choose to:
- Replace the existing captions VTT - Upload a new or edited captions/chapters file.
- Delete the existing captions VTT - This will remove captions/chapters from the on demand event. You will need to delete the current captions if you wish to regenerate captions/chapters via the WorkCast platform.
- Restore the backup captions VTT - Revert to the original WorkCast-created captions/chapters.
You will need some information from the WorkCast platform to identify the media and to verify your identity to secure the request. Follow the steps below:
Step 1 - Find your captions URL & security key
While logged into the platform, navigate to the Media tab and find the media your wish to update the captions/chapters for.
- Use the
icon to copy the url of the media. All caption/chapters files are associated with a media file, so this is used to identify the media that the captions/chapters relate to - You will also need the security key shown. If you don't see a security key, you may need to generate the initial captions link by clicking the GENERATE CAPTIONS text under ACTIONS on the left side of the media page.
When the captions URL appears, there will be a security code generated. You copy this by clicking the
icon. This code verifies your identify for the captions management form below. You should paste it into the Security Key field in the form below.
Step 2 - Editing your file
Captions
Download the captions file by clicking the long file link shown. This ends in .vtt as above.
When the file is downloaded you can edit it in a simple text editor on your PC.
The VTT caption file is a simple format for captions that provides captions/chapters text and timings that are understood by the WorkCast player. There is a blog explaining the VTT file format here that outlines how to edit these files: https://info.workcast.com/blog/how-to-manually-edit-a-captions-vtt-file-on-the-workcast-platorm
Chapters
You can use a simple text editor to create a new chapters file.
The VTT chapters file is a simple format for captions that provides captions text and timings that are understood by the WorkCast player.
- It must be saved as a VTT file
- It must show WEBVTT at the top of the file
- All timings must be in this format for the length of each chapter:
00:00:13.989 --> 00:00:18.450 - Chapters titles to be underneath the timings
An example is below each
WEBVTT
00:00:13.989 --> 00:00:18.450
Introduction to Feature Frenzy webinar
2
00:01:04.830 --> 00:01:07.110
Starting with Cloudshare feature overview
3
00:02:10.679 --> 00:02:25.358
Benefits of Cloudshare for seamless transitions
4
00:03:22.589 --> 00:03:25.589
Sharing MP4 videos via Cloudshare
5
00:05:10.040 --> 00:05:13.470
Audience engagement pod improvements explained
6
00:06:38.839 --> 00:07:01.804
AI functionality for generating seed questions
7
00:09:24.979 --> 00:09:27.739
Introduction to live captions feature
8
00:13:00.090 --> 00:13:04.669
Real life example: Cloudshare improves slide quality
9
00:17:21.219 --> 00:17:26.390
Handy tips for event delivery success
10
00:21:31.619 --> 00:21:36.050
Q&A session: challenges and feedback on updates
Step 3 - Complete the update form
The form fields are described here:
VTT Management URL - As outlined above, this is the url on the WorkCast platform for the media file. All captions/chapters relate to the media, so it is required for all captions management requests.
Language version to update - The platform can support translations as well as captions. Captions are generally created in English as standard. The language you choose here is the language version of the captions that your request relates to. If you are uploading a replacement version of the English captions then choose English. If you have created a translated version of captions then you can upload that translation here and select the appropriate language.
VTT Management Option - This is where you select the action that you want performed. There are 2;
- Replace the existing captions (VTT) with an uploaded file - This option will request a VTT file be uploaded. The file will replace the existing captions/chapters file but a backup is also taken of the current live version. The Restore the backup captions/chapters file option can be used if you need to get the last version back. Note: Only 1 historic version is kept.
- Delete the existing captions (VTT) - Will remove the captions/chapters file associated with the media. The language selection is used to determine which version to remove if more than just English captions are present. A backup will be made of the live version.
WorkCast Security Key - As outlined above, this is a security key that identifies you as a valid WorkCast user with permissions to update the captions or chapters.
Captions and chapters management form
This form can be used to update both captions and chapters. Ensure you click the correct option for what you wish to update.
After you have submitted the form
The form will trigger the request to be actioned immediately. It can take a couple of minutes for the request to be completed. You will receive an email outlining the tasks that have been completed. After you receive the email, you will be able to test the captions/chapters in your WorkCast event.
Caption Management FAQs
Location of the servers processing / running the AI captioning
These are located in Ireland
Details about the AI / LLM system, what is it, how is it taught?
We are using Amazon Transcribe for the caption generation and it is taught via either custom vocabularies or custom language models. These will be mostly on an account basis to keep training specific to the needs of individual Customers
How accurate is the system?
Accuracy of Amazon Transcribe, with no training, is improving in every iteration - in their latest updates they are reporting improvements of between 20% and 50% across most languages
Can the language model be trained?
Yes, The WorkCast team can create specific customer trained AI models that will provide better captions for key domain and brand information. We can also access specific medical models to provide more accurate captions in these areas? note: There is an extra fee for training and using customer AI models.