Important Notes
You need to select all the languages present in your audio/video for Sonix to accurately identify and transcribe them.
While Sonix can transcribe multiple languages, the accuracy may vary depending on the clarity of the audio and the distinctness of the languages spoken.
Step-by-step Instructions
1. Go to Sonix login page and use your registered email and password to access your account.
2. Upload your file: Click on the “Upload” button or drag and drop your audio/video file into the Sonix interface.
3. Select multiple languages: In Step 2, under the question “What languages were spoken?”, select Multiple Languages.
4. Choose all languages spoken: Select all the languages spoken in your audio or video file from the drop-down list.
5. Start transcription: Click the “Transcribe in Multiple Languages” button to begin the transcription process.
Common Issues and Troubleshooting
What's the file length limit for multi-language transcriptions?
Sonix supports multi-language transcription for files up to 2 hours and 15 minutes. If your file is longer than this and has multiple languages, it won't process correctly. However, if your file contains only one language, longer durations are supported without a problem. If you run into issues with a long multi-language file, try splitting your audio into shorter segments, each under 2 hours and 15 minutes.
Why is my transcription accuracy poor, and what can I do?
Poor transcription accuracy can happen for a couple of reasons:
You might not have selected all the languages spoken in the file.
The audio quality might be low, for instance, due to a lot of background noise.
To improve accuracy, first verify that you've selected all spoken languages. If you can, improve the audio quality by reducing background noise. For really challenging audio, you might need to do some manual correction after the initial transcription.
Why aren't languages being accurately identified?
Languages might not be accurately identified if you haven't selected all the languages spoken in the file from the language dropdown during Step 2 of the upload process. Sonix can also struggle to differentiate between languages if the audio quality is poor or if the languages are very similar. For the best results, always aim for clear audio and make sure you've made accurate language selections.
Why are some of my multi-language transcriptions in red, and what does that mean?
In the multi-language feature, you might see parts of your transcription appear in red. This indicates uncertainty, and it's because this feature is still in beta, so the accuracy of the results can vary.
Frequently Asked Questions
Can Sonix automatically detect all languages in my file?
While Sonix can identify multiple languages in a file, you’ll need to select “Multiple Languages” and specify all the languages spoken during the upload process for the most accurate results.
Does transcribing in multiple languages cost extra?
No, transcribing files with multiple languages does not incur any additional cost. You only pay the usual transcription rate regardless of whether your audio contains one or multiple languages.
How accurate is multi-language transcription?
Accuracy depends on several factors, including audio clarity, the distinctness of the languages spoken, and whether all languages were correctly selected. Generally, Sonix aims for high accuracy but may require some post-editing for complex multi-language files.
For further assistance, please contact Sonix support at [email protected] or through the chatbox located in the bottom right corner of our website.