Swell AI Transcription Tutorial
The transcript is the first piece of content Swell AI creates for you. This serves as the building block for all content Swell AI can do for you.
This section aims to give you an overview of how Swell AI's transcription system functions and it's features.
Automatic Filler Word Removal: Filler words in your podcast are automatically removed for a cleaner transcript.
Word-by-word Transcription: Our AI transcribes word by word, meaning every word is timestamped and is linked a specific time stamp in the transcript file
Word-by-word Highlighting: Words being spoken are highlighted for easy tracking.
Listen to the Transcript: All words are timestamped, so you can click on any word in the transcript and start listening.
Real-time Editing: Make updates to the transcript text and save changes easily.
Scroll and Navigate: Scroll through the transcript and jump to a specific section in the audio file.
Word Highlighting: Whatever word is currently being spoken is highlighted in the transcript.
Speaker Labels: The AI is trained to identify multiple speakers in the audio. The labels serves as the name of each speaker.
Media Player: Play your podcast to follow through while editing your transcript.
Alright, let's talk about transcription accuracy. So, our transcript hits a 99% accuracy level, which is pretty much on par with what a human can do. Here's how we came up with that number: we took the same file and ran it through Swell and Rev.com and then compared the two. That's where the 99% accuracy comes from.
Now, the 1% inaccuracy mostly comes from dealing with people and business names. So to address this we created a universal find and replace tool that makes it super easy to swap out those names, more on that below.
The Find and Replace is a simple tool that lets you search for specific keywords within a transcript and replace them with something else in Swell. It's super useful when you need to fix typos or update names.
You can use it on just one transcript or apply changes to all your documents at once. If you find yourself constantly fixing the same mistake, you can create a find and replace rule at the project level, and it'll automatically correct it every time in the future.
That's pretty much the gist of it – a quick way to make bulk changes to your content with just a few clicks.
How you can edit the names of your speakers on your Transcript.
Save: With every change or edit within the transcript, always hit save to preserve them.
Just like a document, you can make the necessary edits on your Transcript by clicking and typing on it. Don't forget to always press "Save" when applying changes in your transcript. You can also adjust the format of your transcript: How to Customize Export Transcript Format
To edit a line in your transcript simply highlight the line in your transcript you want edited. From there the option to edit that line should appear. Once done making the edits, click "Save".
After making the necessary edits you'll have to press 2 "Save" buttons. First will be on the line edited, then the "Save" button on the top of the corner of your transcript.
How to Add more Segments and Delete Segments
To add in more segments or dialogues in a transcript, simply click the box with a "+" sign on it.
To delete a segment simply hover your cursor to the side of the segment and a "Trash Can" button should appear.
On your documents, this feature allows you to add timestamps at specified intervals. You have the option to the timestamp's content showing either a title or a summary.
You can copy the entire Transcript and paste it to your clipboard for blank text documents.
Grab the URL of your transcript in order to start collaborating with team members. Once clicked, you'll be given the option to sha
You can export your transcript as a .srt and .vtt and you can even set up a distribution point from syncing in any of your Destinations with Swell.
This section aims to give you an overview of how Swell AI's transcription system functions and it's features.
Transcript Features Overview
Automatic Filler Word Removal: Filler words in your podcast are automatically removed for a cleaner transcript.
Word-by-word Transcription: Our AI transcribes word by word, meaning every word is timestamped and is linked a specific time stamp in the transcript file
Word-by-word Highlighting: Words being spoken are highlighted for easy tracking.
Listen to the Transcript: All words are timestamped, so you can click on any word in the transcript and start listening.
Real-time Editing: Make updates to the transcript text and save changes easily.
Scroll and Navigate: Scroll through the transcript and jump to a specific section in the audio file.
Word Highlighting: Whatever word is currently being spoken is highlighted in the transcript.
Speaker Labels: The AI is trained to identify multiple speakers in the audio. The labels serves as the name of each speaker.
Media Player: Play your podcast to follow through while editing your transcript.
Transcription Accuracy
Alright, let's talk about transcription accuracy. So, our transcript hits a 99% accuracy level, which is pretty much on par with what a human can do. Here's how we came up with that number: we took the same file and ran it through Swell and Rev.com and then compared the two. That's where the 99% accuracy comes from.
Now, the 1% inaccuracy mostly comes from dealing with people and business names. So to address this we created a universal find and replace tool that makes it super easy to swap out those names, more on that below.
Find and Replace
The Find and Replace is a simple tool that lets you search for specific keywords within a transcript and replace them with something else in Swell. It's super useful when you need to fix typos or update names.
You can use it on just one transcript or apply changes to all your documents at once. If you find yourself constantly fixing the same mistake, you can create a find and replace rule at the project level, and it'll automatically correct it every time in the future.
That's pretty much the gist of it – a quick way to make bulk changes to your content with just a few clicks.
Speaker Identification and Editing Speaker Labels
How you can edit the names of your speakers on your Transcript.
Save: With every change or edit within the transcript, always hit save to preserve them.
Transcript Editing and Saving
Just like a document, you can make the necessary edits on your Transcript by clicking and typing on it. Don't forget to always press "Save" when applying changes in your transcript. You can also adjust the format of your transcript: How to Customize Export Transcript Format
To edit a line in your transcript simply highlight the line in your transcript you want edited. From there the option to edit that line should appear. Once done making the edits, click "Save".
After making the necessary edits you'll have to press 2 "Save" buttons. First will be on the line edited, then the "Save" button on the top of the corner of your transcript.
How to Add more Segments and Delete Segments
To add in more segments or dialogues in a transcript, simply click the box with a "+" sign on it.
To delete a segment simply hover your cursor to the side of the segment and a "Trash Can" button should appear.
Timestamp Formatting
On your documents, this feature allows you to add timestamps at specified intervals. You have the option to the timestamp's content showing either a title or a summary.
Copy and Paste
You can copy the entire Transcript and paste it to your clipboard for blank text documents.
Shareable URL
Grab the URL of your transcript in order to start collaborating with team members. Once clicked, you'll be given the option to sha
Export Options
You can export your transcript as a .srt and .vtt and you can even set up a distribution point from syncing in any of your Destinations with Swell.
Updated on: 09/19/2024