podcast.phor.net

This is a reusable project template to self-publish your YouTube channel or other audio files as a podcast. Output fully complies with Apple Podcasts for Creators RSS feed requirements.

We use Jekyll to generate the RSS feed and HTML pages, and GitHub Pages to host the site. And you would go around registering your podcast with iTunes, Google Play, and other podcast directories.

The data (_data, _episodes) in this repository are for the NFT/Web3 Community Service hour.

Subscribe to Community Service Hour at:

Apple Podcasts: https://podcasts.apple.com/us/podcast/community-service-hour/id1662422217
Google Podcasts: https://podcasts.google.com/feed/aHR0cHM6Ly9wb2RjYXN0LnBob3IubmV0L2ZlZWQueG1s?sa=X&ved=0CAMQ4aUDahcKEwignIL6m7f8AhUAAAAAHQAAAAAQBA
Spotify: https://open.spotify.com/show/3k4PnmjfLiuNo9HpXemCdJ
Stitcher // NEED LINK
Breaker // NEED LINK
Pocket Casts // NEED LINK
RadioPublic // NEED LINK
Overcast // NEED LINK
RSS: https://podcast.phor.net/feed.xml

Syndicate your podcast

Audio encoding

We follow all Apple audio requirements and, when appropriate, their recommendations.

Format:

For RSS feeds, Apple Podcasts accepts MP3 or AAC formats.
For RSS feeds, we strongly recommend using AAC instead of MP3.
When choosing AAC, we recommend using the MP4 format over the ADTS format because MP4 allows for the most-efficient streaming usage and accurate seeking.

Bit rate:

... recommended bit rate... | Number of channels | 22.05/24 kHz | 44.1/48 kHz | | :--------------------- | :--------------- | :-------------- | | 1 (mono) | 40–80 kbps | 64–128 kbps | | 2 (stereo) | 80–160 kbps | 128–256 kbps |

Levels:

... we recommend that the audio signals are preconditioned so the overall loudness remains around -16 dB LKFS, with a +/- 1 dB tolerance, and that the true-peak value doesn’t exceed -1 dB FS

Our selections:

AAC/MP4
Stereo, 44.1kHz, 160kbps
Overall loudness of -16 dB LKFS with +/- 1 dB tolerance, true peak of -1 dBTP
Encode loudness information in the header of the MP4 file

We achieve this encoding using:

ffmpeg -i IN.m4v -vn -acodec aac -ac 1 -ar 44100 -b:a 160k -af loudnorm=I=-16:TP=-1:LRA=11:print_format=json -f mp4 -movflags +faststart OUT-WITHOUT-CHAPTERS.m4a

Chapter markers

Specification: https://podcasters.apple.com/support/2482-using-chapters-on-apple-podcasts

This site generates the ffmetadata files needed by ffmpeg to add chapter titles into an episode. See the output /ffmetadata folder. Also, a separate text-to-ffmetadata.js script is provided for convenience.

We add this chapter metadata using:

ffmpeg -i Episode\ 14\ FULL.m4a -i ~/Sites/podcast.phor.net/_site/ffmetadata/2022-03-08-episode-14.txt -map_metadata 1 -codec copy 2022-03-08-episode-14.m4a

Production notes

Save chapter markers in this format to chapters.txt (use HH:MM:SS or MM:SS):

00:00:00 Intro
00:00:45 The agenda
00:02:27 Area contract intro
00:07:22 Push and pull Ether sending
00:17:49 Ethereum should be illegal
00:33:49 Free tool: batch transfer NFTs
00:44:19 Adidas drop gamed by robot?
01:13:15 Composing the Tweet

Then convert to ffmetadata format using:

node text-to-ffmetadata.js < chapters.txt > chapters.ffmetadata

Use ffmpeg to extract audio from the video episode, use required format and add chapter markers from the file chapters.ffmetadata.

ffmpeg -i ../Episode\ 3\ FULL.m4v -i chapters.ffmetadata -map_metadata 1 -vn -acodec aac -ac 2 -ar 44100 -b:a 160k -af loudnorm=I=-16:TP=-1:LRA=11:print_format=json -f mp4 -movflags +faststart episode.m4a

check:

ffprobe -i episode.m4a -print_format json -show_chapters -loglevel error

Metadata

Some tricks to update some metadata.

UUID=$(uuidgen); sed -i -e "s/guid: .*/guid: \"$UUID\"/" 2022-03-08-episode-14.md

for a in *.md; do UUID=$(uuidgen); sed -i '' -e "s/guid: .*/guid: \"$UUID\"/" $a; done

Get an episode length in seconds, itunes-duration:

ffprobe -v error -show_entries format=duration -of default=noprint_wrappers=1:nokey=1 2022-03-08-episode-14.m4a

Get an episode file size in bytes, enclosure-length:

wc -c 2022-03-08-episode-14.m4a

How to add an episode

Translate chapter markers to FFMpeg format (above)
Upload media to media.phor.net
Create a _episodes/ file
Fill in all metadata
Add show notes in markdown, can use GPT to translate any random format you have into paragraph format:
```
# Below is a paragraph-based description of the episode and the above topics:
```

TIPS AND HACKS:

ADD notes from audio transcrbe to show notes

<!-- TODO: add summary here

❯ ffmpeg -i 2022-12-21-episode-3.m4a -ss 00:07:22 -to 00:17:49 -c copy out.m4a
❯ ./products/hear -d -i ~/Desktop/out.m4a > pushpull.txt   
And use a text model for:

Topic: Push and pull Ether sending

Garbled text:

INPUT HERE <<< >>>

Cleaned up, intelligible text:

THIS IS YOUR OUTPUT <<< >>>
 -->

dtedesco1 / podcast.phor.net