Genderfucked schizoqueer autistic plural plushie eevee mystics


one of those annoying vegans your unexamined ethical system warned you about


Seer of Blood (Wyatt) and Witch of Void (fractal)


not a person, Ξ˜Ξ”, tma

this user likes the homestuck epilogues

name-color: #ff007b


🎀 psychosisposting 🎀
cohost.org/feedbackmicrophone

artemis
@artemis
breaking the serenity of dawn - [ffmpeg builtin] breaking the serenity of dawn
[ffmpeg builtin] breaking the serenity of dawn
breaking the serenity of dawn
00:00
breaking the serenity of dawn - [libfdk_aac] breaking the serenity of dawn
[libfdk_aac] breaking the serenity of dawn
breaking the serenity of dawn
00:00

ffmpeg's builtin AAC encoder has trouble sometimes with certain audio. libfdk_aac is basically always really good. Compare these two audio files. These are the same bitrate (96k). But listen to all that artifacting on the ffmpeg builtin encoder! (and fsr it thinks the song is longer? idk...). I used 96k to make it easy to hear even if you're not sensitive to this stuff, but it persists surprisingly high up the bitrate!

Not all songs artifact like this through the builtin encoder, but I've had it bite me a bunch and I'd rather just not use it. If you're transcoding for a platform that supports opus, well, just use that instead. But if you're not... here's how you get the cool encoder I used for the second one the easiest way possible.


natatorialremnants
@natatorialremnants
This page's posts are visible only to users who are logged in.

You must log in to comment.

in reply to @artemis's post:

I found this last night looking at the ffmpeg tag and realize this had been posted basically the day before I sat down and compiled ffmpeg "the hard way" on a machine for the exact same reason, having found that libfdk_aac was way way way more efficient.

Looking forward to playing around with nix for the occasional edge case like this. Thank you!