rimple

Thank you @smcleod, accepted.

Thanks for trying it out. You’re right, the audio length varies depending on the chosen voice. What I do first is break the video into n parts and extract the text. Then, when updating the audio, I piece the parts back together by adjusting the video speed to match.

It takes sometime to get the transciption ready. Can you share the which video did you try, I will check and get back.

The website got blocked by the SURBL list. Here's alternate link: papertube.live. All the kindle emails were based on user actions, not sure why the site got flaged. I have the SPF/DKIM setup as well.

To get the ebook on kindle, please add @papertube.site to your approved kindle email list. The previously provided address was incorrect, my bad.

You can leave and come back. The processing should get done in few minutes(<5 mins for 3 hours video).

Updated.

Not handling video content here. The focus is on cases where no visual content is needed, such as podcasts, TED talks, and conversations.

Good idea, thanks. I'm currently downloading all the videos, and this should help reduce costs. I can use the video description/title to determine if there's only one person speaking by using LLM.

Thanks for the feedback. I agree the current web view flow is awkward; I'm working on fixing it. The mobile UI is a bit weird too, but it should be better on Kindle.

That's cool. I've created a website(https://papertube.site) that essentially transcribes video conversations for reading on Kindle. Right now, I'm relying on third-party APIs, but I was thinking about self-hosting to reduce costs.