Jump in the discussion.

No email address required.

The quality is dogshit though.

Jump in the discussion.

No email address required.

Just upsample it before playback then

Jump in the discussion.

No email address required.

Smart.

Jump in the discussion.

No email address required.

:pe#pojam:

Jump in the discussion.

No email address required.

for the compression rate it offers the quality is great. the first few iterations of mp3 weren't that great either. the quality of this approach will keep improving. but encoding/decoding requires a lot of resources and the model is ~300MB. whether it makes sense depends on the application. e.g. if you're storing a a million hours of audio recordings from court sessions or interviews, the quality is good enough at 6 kb/s (that's bits, not bytes. 64kb/s for comparable m4a) you now only need 3TB storage instead of 30TB.

Jump in the discussion.

No email address required.

That's fair but it's not gonna save space on my phone right this second so I'm going to discard it snobbishly regardless :!marseyindignant:.

Jump in the discussion.

No email address required.

It sounds the same

Jump in the discussion.

No email address required.

You might have ear-aids.

Jump in the discussion.

No email address required.

WHAT

Jump in the discussion.

No email address required.

AAAAAAAAAAIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIDDDDDDDDDDDDDDDDDDDDSSSSSSSSSSSSSS!

Jump in the discussion.

No email address required.

He doesn't change the sampling rate so that's irrelevant. Also testing 96kHz to 44kHz is not interesting since human hearing sucks at high frequencies, you could just use a low pass filter and not any compression to get rid of those. I can tell you without a study that 99% (excluding the deaf) can tell the difference between 22kHz and 44kHz, and at 8kHz you can't understand speech anymore, and that's still a smaller range than 96 to 44.

Here he uses a transformer to compress 44kHz audio to a lower kbps, the frequency features below and up to 44kHz are going to be affected in unpredictable ways regardless of the output sample rate, and you can hear it clear as day with shitty earphones if you compare the originals vs the compressed.

Jump in the discussion.

No email address required.

You can use ffmpeg and test yourself. Not top secret :marseyglow: cowtools or anything.

Jump in the discussion.

No email address required.

Stfu nerd with you gay butt audiophile "uhhhh you can totally hear 495fps" as if we need anything more than 30

Kill ykursy


:#marsey:

Jump in the discussion.

No email address required.

:marseyconfused:

he was wondering if he could hear 96KHz - you can't and it's easy to test with ffmpeg.

Jump in the discussion.

No email address required.

Link copied to clipboard
Action successful!
Error, please refresh the page and try again.