Your Transcript
Your Transcript
When your audio is uploaded, Kaituhi will ‘slice’ that audio into manageable chunks, or what we will continue to refer to as captions, based on the average length of a spoken te reo Māori sentence and ‘silence’ or breaks in speech. This is what you will see when you first open your file.
Auto Transcribe
If the auto transcribe failed when you originally uploaded your file, you can use the Robot icon in the bottom toolbar and watch in real-time as the machine automatically transcribes the audio. The length of the audio determines how long this will take and once complete, your transcript is ready for editing.
Editing your Transcript
Once your file has been transcribed you can edit and analyse your audio transcript according to your purpose and workflow. You can find all of the shortcuts in the lower toolbar to help make transcribing quick and easy.
Each caption has the timestamps displayed at either end of the audio wave and the total seconds for the audio in that caption. Use the Zoom in your browser to change the number of captions you’d like to display at once and increase the size of text.
TIP: Displaying at least three captions at once is helpful to see what comes before and after. This is helpful later when we you begin splitting and merging captions.
Edit Transcription Name
Changing the name of your file is as easy as selecting the text on the top panel and renaming your file. Initially, it will automatically be set to the name of the file you uploaded.
Playing Audio
You can use the Play icon at the bottom left of your screen or the shortcut key Shift+Space to play and pause your audio. Once it has played out, it will return to the beginning of the caption. Return/Enter will also play and take you back 2 seconds. It will play each caption individually, alternatively you can click the Playback Link icon in the lower toolbar to continue playing through the entire audio.
The Playback speed icon in the lower toolbar allows you to speed up or slow down the audio.
TIP: This can be really helpful if you are struggling to decipher what someone is saying or moving through the audio quickly to pick up key messages or find a specific part you want to work on.
Editing Speaker
If you have multiple speakers throughout your transcript, you may want to assign a speaker to each caption. This can be done by selecting Enter Speaker, below the audio wave and typing in the name of the speaker for that caption of audio. Kaituhi will remember those names and create a dropdown list for you to select from as the speakers switch.
TIP: We recommend, where possible, having one speaker per caption. If one speaker interjects or talks over the other, you can determine the most dominant voice, or who is saying the most important thing and assign the speaker to that caption. This may depend on the purpose of your work. This will create a tidy transcript once the file is exported.
Editing Text
This is easy, select the caption you would like to edit and start typing!
TIP: You can determine the standard orthography you want to use to suit your purpose. If you are doing verbatim transcription, think about how you will deal with stutters or where the speaker changes direction mid sentence, also, whether or not you will transcribe ‘ums’ and ‘ahs’. These are helpful to think about when establishing standards across the team, and before getting into the work of transcription. You may decide that the transcript you develop in Kaituhi will include all of the unique features of speech and upon exporting the transcript for publication, you will edit the exported file, keeping the Kaituhi file as the master copy.
Splitting and Merging Captions
While our algorithm does it’s best to split the audio into meaningful captions. There will be cases where you will need to bring together captions if it misses the end of the sentence or split them if another speaker begins talking mid-caption or if you’d like the next caption to start at the beginning of a new sentence.
TIP: If you are using Kaituhi to create subtitles/captions for a video, you may want to think about how many words will appear on screen and split the captions accordingly.
To Split caption double click your mouse on the audio wave where you want to split it. This will create a new caption underneath with the audio and transcribed words you split off.
To Merge captions together, you use the arrows to the left or right of the audio wave in each caption, depending on which way you would like to merge. The Left arrow will merge to the previous caption and the Right arrow will merge to the following caption.
TIP: Listening ahead three or four captions can make the process of splitting and merging faster. Assigning one complete sentence per caption is a useful way of being able to follow along with the audio and makes reviewing a lot simpler.
Syncing Your Data
If you have agreed to allow your data to support the improvement of the Papa Reo ASR machine, you will see a green cloud indicating that the caption and its associated audio will be uploaded and synced to the Papa Reo ASR machine. The privacy policy and explanation of how we will care for your data will be outlined in the terms and conditions of your license agreement.
If, however, a portion of your data is highly sensitive and you would prefer that it wasn’t used, you can deselect the syncing icon turning it grey.
Exporting Your Transcription
When you have completed your transcript you are able to export your work using the Export icon in the lower toolbar.
Currently the platform supports exporting the text as a .txt file, the audio as an .mp3, and the subtitles/captions as .vtt.
Support
If you have any questions or issues while you are using Kaituhi, you can contact us at: info@tehiku.co.nz
We would also love to hear from you around ideas for improvements that would make using Kaituhi a better experience for you.