Transcriptions is a feature available on paid plans that automatically transcribes recordings done on Zencastr in English, French, German, Portuguese, and Spanish.
In this article:
What Are Transcriptions And Why Use Them?
When you record on Zencastr in English, French, German, Portuguese, and Spanish (based on your session’s selected language), a transcription of the entire conversation is automatically generated after you finish recording and all participants’ files upload to the recording page.
These transcriptions are timestamped, speaker-labeled, and can be downloaded in CSV or TXT formats, or copied to your clipboard.
These automated transcriptions save time and effort compared to manual transcription, and for podcasters can:
Boost your podcast’s discoverability with SEO.
Make your content more accessible to a wider audience.
Simplify your editing process by offering a clear reference for your content.
Additionally, with Zencastr Editing, you can edit your transcript to cut and trim your audio and video.
Getting Transcripts For New Recordings
Transcriptions for Zencastr recordings are automatically generated when you select English, French, German, Portuguese, or Spanish as your session’s language.
After you stop the recording, the transcription starts processing. It’s important for all participants to keep the page open until their files finish uploading. Once everyone’s files are uploaded, the transcription will continue processing in the background, even if the page is closed.
When it completes, CSV, TXT, and COPY buttons will appear on the transcription track on the Recording Page, indicating that the transcript is available to download.
To download the transcript:
From the Sessions tab of you Dashboard, click the name of the session in the ‘My Sessions’ list.
Click the Join Session button.
Locate the Transcription track on the Recording Page.
Click the CSV button or TXT button and the transcription file downloads to your computer.
Getting Transcripts For Old Recordings
If you recorded on Zencastr in English before October 7, 2021 or in French, German, Portuguese and Spanish before January 17, 2024, transcripts were not automatically generated. However, you can still run them now.
From the Sessions tab of you Dashboard, click the name of the session in the ‘My Sessions’ list.
On the Recording Page, locate the transcription track, and click the Transcribe Recording button.
Result: The Transcribe Recording button will change to ‘Processing’. When the transcript is ready, solid grey CSV, TXT, and COPY buttons will appear on the track, indicating that the transcript is available to download.
Related Questions
Do I need to stay on the Recording Page for my transcript to finish processing?
For the transcript to finish processing, the host and each guest need to keep the Recording Page open until their files finish uploading. Once a participant's files have uploaded, they can close their page. After all participants' files are uploaded, the host can also close the page, and the transcript will continue processing in the background.
What languages are supported for transcripts?
At this time, transcripts are automatically generated for recordings done on Zencastr in English, French, German, Portuguese, and Spanish. We hope to offer support for additional languages in the future.
My transcription isn't completing. What can I do?
Can I upload an audio or video file that I recorded or edited elsewhere to transcribe?
Yes, you can upload a file that wasn’t recorded on Zencastr to use with Zencastr Editing and get a transcript. However, the transcript won’t include individual speaker labels like those from Zencastr recordings since the platform can’t read source file information, such as participant data, from content recorded elsewhere.
What is the transcription word error rate/accuracy?
Zencastr Transcriptions are produced by our own language modeling AI and proprietary machine learning tools. The Word Error Rate (WER) of our technology is currently comparable to that of transcription industry leaders such as Google, Descript, and Otter.ai. And we only expect the WER to improve in the future!