hpr2489 :: CONTEXT IS EVERYTHING
Lostnbronx creates a tiny video, from a large audio file.
Hosted by lostnbronx on Thursday 2018-02-15 is flagged as Explicit and is released under a CC-BY-SA license.
Listen in ogg, spx, or mp3 format. | Comments (1)
CONTEXT IS EVERYTHING
hosted by Lostnbronx
Turning a large audio file into a tiny video file for a very specific use case.
This solution probably won't work for you, because none of the ones I found on the Internet worked for me. Trial and error led me here, and it's probably the only guide you'll have as well.
Start with as high a quality audio file as you can manage. .wav is good, but it's big. Let's convert it to .flac. If you already have a .flac file, skip this part.
1.) Convert .wav to .flac.
sox INPUT.wav OUTPUT.flac
2.) Convert the .flac to a very small mono .opus. Bitrate can be even smaller. I went down to 14.
opusenc --bitrate 18 --downmix-mono INPUT.flac OUTPUT.opus
3.) Combine the .opus file with a single static image, and output to a .webm video. This should not be very much bigger in file size than the .opus and .jpg combined. The smaller the image file, the better. (I tried using a .gif, but it was actually bigger than the .jpg I ended up with.)
ffmpeg -i INPUT.opus -r 1 -loop 1 -i INPUT.jpg -c:v libvpx -tune stillimage -shortest -y -c:a copy OUTPUT.webm
The final file. It doesn't sound great, but it's listenable, which is all that was desired.
Here's a better quality version of the audiobook.
The process and final result can be improved upon by people smarter than I, without doubt, but this works for now.
SPECIAL THANKS to the Urandom guys (X1101, Thaj, and Pokey), Monsterjavaguns (Jason van Gumster), and the ever-fabulous Klaatu, for their suggestions and encouragement. I would not have found a solution to this, nor even thought to do an episode of HPR, without them!
Editor's Note 2018-02-15: The wrong audio was accidentally released with this show. It has been corrected and should be re-uploaded by your podcatcher.
Automatically generated using whisper
<< First, < Previous, Next >, Latest >>
whisper --model tiny --language en hpr2489.wav