After recording the voiceover for my ergonomic keyboards blog post, I was frustrated at how hands-on the editing was. Why was I manually going in, removing bad takes, and adding timestamps to the post? I was determined to find a better way.
Using the Web Speech Recognition API, Narration.studio detects which sentence you are saying, allowing you to retake previous lines, pause and take a drink, or chat to someone mid-recording. It will all get cut automatically.
Fine-tune the cuts in the editor, adjusting the start and end point for each clip, or the amount of time between sentences. When you're happy, download your edited audio as a high-quality WAV file and grab the script - now annotated with a timestamp for each paragraph.
Using the Web Speech Recognition API, Narration.studio detects which sentence you are saying, allowing you to retake previous lines, pause and take a drink, or chat to someone mid-recording. It will all get cut automatically.
Fine-tune the cuts in the editor, adjusting the start and end point for each clip, or the amount of time between sentences. When you're happy, download your edited audio as a high-quality WAV file and grab the script - now annotated with a timestamp for each paragraph.
It's all open source on GitHub, and the readme has more detail about the implementation, so go take a look and submit a PR! https://github.com/stevenwaterman/narration.studio/