jeremyforest / nmc-videos

Pipeline to create videos with Neuromatch conference

update 2021

Below still applied. Env is out of date. Process is to query youtube videos from airtable get_yt_links_from_airtable.py, download them locally,process them (sound, and adding splashscreen), then reupload to youtube and add to the right playlist. Creates a data.csv file to summarized failed/success at each step.

Defintely some progress to make in the code to get it more robust.

pipeline.py is still a work in progress. Was never really used yet. get_yt_links_from_airtable allow to query the video links from airtable database. dl_yt_video.py download youtube videos locally to be processed.

Minor updates to others code files to make it compatible with updated flow.

nmc-videos

Pipeline to create videos with Neuromatch conference.

Create the right conda environment with conda env create -f environment.yml

conda activate videos

python cut_videos.py shows how to cut a video (with splash screen).

python cut_airtable.py shows how to cut a video based off of timings stored in Airtable.

python update_description.py shows how to update a video's description based off of airtable data.

Secret tokens should be stored in .env:

• GOOGLE_CLIENT_ID
• GOOGLE_CLIENT_SECRET
• GOOGLE_API_KEY
• AT_APP_KEY_ORANGE
• AT_APP_KEY_GRAY
• AT_API_KEY

AT are the airtable keys, Google the YouTube keys.

Only tested on Ubuntu.