http://opus-codec.org/release/stabl e/2017/06/20/libopus-1_2

>48kHz

File: trashman7.gif (2MB, 580x433px) Image search: [Google]

2MB, 580x433px

>no 44.1KHz

But does it beat 80Kbit/s AAC-HE v2 files or 40Kbit/s AAC-HE v2 streams in terms of quality?

Where can /gentooman/ find opus streams?

File: Opus_quality_comparison_colorblind_compatible.svg.png (102KB, 1280x950px) Image search: [Google]

102KB, 1280x950px

>>60997123
https://en.wikipedia.org/wiki/Opus_(audio_format)

Opus destroys AAC in every metric possible

>>60997123
Compile and encode yourself for testing. Use latest source code to get best quality available at m.

>>60997163
>2015

>AV1 still not out yet
Like it matters though, has Mook even enabled VP9 on webms?

File: Untitled.png (192KB, 1298x1058px) Image search: [Google]

192KB, 1298x1058px

I made this.

>>60997394
Not allowed by his Google overlords. Same for sound. If they are added, Google will pull free captcha service and gook will lose his golden pass money.

>>60997163
except adoption rate and embedded support.
Companies rarely implement free/open standards as they're not profitable.

>>60997163
Very nice. So all in all we might agree that it is the best free/open source audio codec.
However: It seems to be 48KHz only!? Why? CDs are 44.1KHz, converting will make quality decrease.
Also:
>Currently, Microsoft Windows and macOS do not provide native support for the playback of Opus files. [...]

>Google added limited native support for the Opus audio codec in Android version 5.0 ("Lollipop"). However, it is unable to playback local Opus files in Ogg containers [...]

>Android version 6.0 ("Marshmallow") is able to playback local Opus files encapsulated in Ogg containers, but the filename extension .opus is not recognized by built-in players. [...]

Those are big problems. They might vanish, but nothing's sure.

>>60997538
Actually it's the opposite, the said companies have invested R&D money in other standards and are grabbing every penny they can, supporting competing standards isn't part of the business.

>>60997538
wire uses ffmpeg and opus

>>60997053
MP3 is more than enough also most supported.

>>60997644
MP3 should have at least been outmoded by Vorbis. Theres plenty of devices with Vorbis hardware support.

>>60997489
what the fuck? why

>>60997053
Seems to encode much quicker on my server now.

>>60997644
Not enough for sharp sounds because of its format design.

>>60997489
multiple boards already have sound enabled

>>60999808
Really gave me an idea...

>select all images showing a tablet computer
>one of those tablets literally says "GILFST" on it

I seek asylum from Google

But... I did, OP, many years ago, all my music is in Opus format now (with FLAC backups of the original CDs, of course).

>>60997110
>>60997068
44.1 should not even exist, the correct one was always 48.

>>60997558
The vast majority of audio hardware runs at a minimum of 48KHz internally anyway and resamples 44.1 to 48 on the fly.
You might as well resample it at encoding time where you have control over it. The opus resampler has be found to be very good.

>>60997538
Chrome supports Opus.

If you watch HD video on YouTube with Chrome, you are listening to an Opus-encoded audio track.

>>60997163
That certainly wasn't true until 1.2. AAC was clearly better for music at low bitrates.
I'll have to try 1.2, though, I heard they made some improvement for music.
Also AAC decodes faster.

>>61004227
zis

ever heard of the shannon sampling theorem you faggots? your adult eardrums dont even move to frequencies above 16khZ

>>61004593
Explain.

>>61004623

it basically means it's physically impossible for almost all adults to hear the difference between 32kHz or 100kHz sampling

>>61004652
What does it say about psychological effects?

>>61004681

shannon said a signal that is saved on your computer at a 32kHz sampling rate PERFECTLY describes a real life signal that has up to 16kHz frequencies.

and humans cant hear the difference between 16kHz or 100kHz signals.

>>61004745

ok this is a very inaccurate explanation though because real life signals have infinite frequencies

Okay, so I just found this website:
http://www.radio-browser.info/gui/#/bycodec/OGG
Plenty of ogg streams. Is there a preferred method of seeing what codec is being used? Ogg is just the container, right?

>>60997538
how does that make sense?
it isn't true first of all and using proprietary standards that aren't theirs is more costly

unless they implement their own, but who does that for everything? free software fills in the gaps, look at the legal section for any software and find a large section dedicated to specifying licenses for the large amount of free/open shit they probably used

>opus
HAHAHAHAHAHAHAHAHA
/thread

>>61004497
It's clear that Opus has been superior to the best AAC encoder(Apple) at 64 ~ 96 kbps, though.

http://listening-tests.hydrogenaud.io/igorc/results.html

>>60997163
If they equal out at 128kbps what's the difference? No one these days is using less than that.

File: scores_by_tracks_en.png (18KB, 720x450px) Image search: [Google]

18KB, 720x450px

>>61005423
Never heard of streaming?

>>61005427
But apple, google and spotify all stream at well over 128kbps

>>61005441
They are giants. Small companies would prefer lower bitrate.

>>61005441
First of all that's a wrong generalization and secondly 40 kbit/s aac he sounds perfect on a bluetooth speaker when being outside.

Literally nobody needs hifi quality for that (I'm 100% certain there will be people that demand it though). What counts is how long your data plan will last. So in that case: low bitrate > high bitrate

>>61005508
True that, I guess I don't even think about it because I have 30gb a month of data.

opus ist amazing for audio book collections.

32kb/s or even less is perfectly fine.
for mono I can use 24 kb/s.
A complete book is usually around 100 mb.

>>61004497
>>61005416
I just tried 1.2. AAC is still better for very low bitrate music. Try them both at 16kbps, AAC still sounds kinda decent, Opus does not.

>>61005729
I agree with that, for mono/speech/audiobooks Opus is great.

>>61006279
What AAC encoder did you use? Also, it might be better to prepare 24000Hz source for Opus at such very low bitrate.

>>61006502
>What AAC encoder did you use?
FDK, because it's the only one that actually encodes to 16kbps (with Foobar and manual parameters).

>Also, it might be better to prepare 24000Hz source for Opus at such very low bitrate.
I tried that, didn't really help.

This is identical to libopus 1.2-rc1, right? I'm a major autismo about having my mobile music library encoded with the latest encoder version.

>>61006611
Okay, then. I heard that good old TwinVQ became a MPEG-4 codec for very low bitrate. I wonder if it's better than HE-AAC at 16kbps.

I did a double blind listening test on myself with Opus. Opus' quality caps at around 256kbps. A 320kbps Opus stream stands out between a 320kbps MP3 or Vorbis file, where both the MP3 and Vorbis files are indistinguishable from raw uncompressed PCM. It's subtle, but it's there.

It also takes reasonably more CPU to decode an Opus stream, so listening to Opus can drain your battery more than MP3 or Vorbis. This is in part due to the extremely tiny buffer lengths that Opus works with, which was chosen to provide lowest possible latency for VOIP streaming, but inefficient for general playback listening purposes.

I don't think I hear a difference starting at 64 kbps

https://people.xiph.org/~jm/opus/opus-1.2/

>>61006616
I assume so as there is no mention of bugs found in rc1 and if there was there would have been an rc2.

>>60997068
>>60997110

Please stop this meme.It's been explained over and over again: https://xiph.org/video/

Educate yourselves before you start posting illiterate crap.

>>60997558
>However: It seems to be 48KHz only!? Why? CDs are 44.1KHz, converting will make quality decrease.

No, the loss is imperceptible and virtually non-existent when compared to the sheer fact we're dealing with lossy compression anyway.

I'm gonna have re-encode by library for my mobile. The sound-quality is acceptable even down at 48kbps

Is opus supposed to sound good even at 32kbs? I'm having a hard time telling it apart from the uncompressed reference, MP3 sounds like you're listening to it through a long hallway

>>61006910
this
Opus at 48 kb/s behaves quite interestingly in my case. I've encoded some flacs to such 48 kb/s opuses and when ABX-ing on any of my headphones and whether on my pc or my phone, I can't tell the difference for shit
but then when I play these on my phone's loudspeaker, the difference is instantly audible
I really wonder what the hell is going on here, but yeah, for music on headphones Opus @ 48 kb/s is really all you need

>>60997053
>Why have you not joined the Opus master race, /g/?
I use it for music on my low capacity DAP, generally stick to ALAC though.

>>61007201
>I really wonder what the hell is going on here
Contrary to the popular /g/ meme, better equipment won't necessarily expose the differences more. Quite the contrary, worse equipment with more irregularities will often do that, because the irregularities in the source and the speakers accumulate.

>>61007437
i'm of the opinion the the thing that matters the most, is the source material.

a poorly recorded/mastered album, will still sound like shit, no matter how expensive your gear is

shit tier tagging and no ReplayGain

>>61009326
Vorbiscomment is fine and RG works in foobar, etc.