I've cloned the PC20 feed and called it "Podcasting 2.0 - Sandbox Feed". You can use this feed to test against the newest tags.

It is feed id 1318635 for devs using the API. The feed url is:

The feed API endpoints now returns "locked" status.

The episode API endpoints now return a "chaptersUrl" for the new chapters tag file.

Going forward I will update this feed when tags change or when new episodes are released.

· · Web · 2 · 0 · 3

@dave you should start using the content:encoded tag, and put only a short summary to itunes:summary...

@dave Can we also get the other attributes in the sandbox that has been finalized, like funding and transcript?

Not asking to have it in all podcasts, but maybe you could make an if/else and return some dummy data in the sandbox so that we could prepare our apps to support it, when it goes "live"?

@martin Got transcripts in the aggregator last night. And working on funding tonight. Then I’ll expose in the api endpoints. In need a good transcription service. I usually create them through AWS in the past. But they don’t put time codes in.

@dave @martin which AWS transcript service are you using? We did a proof of concept with them last year and they gave us timestamps for each word.

@tomrossi7 @martin I used the Amazon Transcribe service when it was first released. I didn't remember it having timestamps. Let me look again.

@tomrossi7 @martin Ok, I ran episode 7 through Amazon Transcribe and it produced this JSON file:

I see what you mean. I think I remember it doing that now.

@dave yeah, its super verbose, but I love the high fidelity! When we worked with it, we stripped it down to the JSON format we now support.

@dave I need to move the formats over to podcastindex github! I've updated the JSON to match camelcase too...

@tomrossi7 Ah! Thanks. I'll write a converter from AWS to your format. Will the camel-case be a version 1.1?

@dave nah, no one has implemented it yet, so I just kept it at 1.0. You can see an examples on this feed:

@tomrossi7 I wonder if an AWS to PODCAST json format converter would fit into HyperCatcher Studio somewhere.

What do you think @hypercatcher ?

@dave @hypercatcher we've had to write converters for Descript, AWS, Temi, and SRT. I'm happy to share the code, but we use Ruby.

@tomrossi7 @hypercatcher I can make my way through Ruby if you have time to share. I'll make a utils folder in the repo and post my version there.

@dave @tomrossi7 Yeah, I've definitely thought about adding transcripts to HyperCatcher Studio. I've already written a feature that will take a podcast RSS feed send it to a Speech to Text API and then parse it into podcast Chapters using OpenAI's new GPT-3 beta api.

The only problem is that the speech to text part does cost money so I need to work in some sort of payment solution before enabling that feature or I'm afraid I'm going to lose my shirt :)

@dave @tomrossi7 Just translating from one format to another should be relatively simple and free though, I'll definitely look into that for sure

Sign in to participate in the conversation
PodcastIndex Social

Intended for all stake holders of podcasting who are interested in improving the eco system