1
$\begingroup$

Assumed we have those three audio streams available for listening to:

  • MP3 with a bitrate of 128 kbit/s
  • AAC with a bitrate of 192 kbit/s
  • OPUS with a bitrate of 64 kbit/s

I assume that OPUS offers the best compression ratio and therefore would deliver the best audio quality in case of all bitrates would be the same.

However, as seen above, the provided bitrates are different.


I would like to know if 64 kbit/s OPUS offers a similar quality in comparison to 192 kbit/s AAC?

I guess the bitrate difference between 64 kbit/s and 192 kbit/s is just too much, so even with its better compression, OPUS won't reach the AAC quality in this case.

On the other hand, MP3 is very much outdated nowadays, so I would not consider that further (or am I wrong?).

$\endgroup$
2
  • 2
    $\begingroup$ Depends on what you mean by "best audio quality" since there's some psycho acoustics involved. Maybe this paper helps you in this: ceur-ws.org/Vol-2654/paper25.pdf $\endgroup$
    – Juha P
    Dec 29, 2021 at 12:19
  • $\begingroup$ If you put $N$ audiophiles in a room, will the number of opinions about different audio recording methods increase as $\mathcal O (N^2)$, $\mathcal O (e^N)$, or some faster-growing amount? $\endgroup$
    – TimWescott
    Dec 29, 2021 at 19:05

1 Answer 1

3
$\begingroup$

First of: "AAC" is not a single codec; there's AAC-LC, HE-AAC, and probably a lot more.

Since "audio quality" is a perceptive quality, there's no objectively "best" answer. It's, however, very likely that in all direct human comparisons, MP3 at 128 kb/s would fare worst.

Also "audio quality" means something very different in different use cases: is "audio" speech? Is "audio" classical music? Is "audio" distortion-heavy rock music? Is it the sound of an action movie? Bird song recordings for ornithologists?

Also note that you're not actually comparing the coding standards, you're comparing the quality coming out of any given encoder. It's not guaranteed that two different encoders would make an identical bitstream out of the same uncompressed audio – filtering and things like pre-emphasis leave room for choice, before the digital signal undergoes the specified parts of a codec.

Your best bet is hence to get a corpus of the type of audio relevant to you, and let a group of people relevant to you listen and compare in a controlled environment. A taste of what the results of that might look like is the Hydrogen audio multiformat listening comparison.

Honestly, though: I have yet to personally meet anything were I can tell 192 kbps HE-AAC from uncompressed audio. I wouldn't throw away my CDs after saving 64 kb/s Opus, though. So that's a subjective hand-waving answer. But, on the other hand, grabbing your "favorite" sample of audio that you think would make for a good comparison, and throwing encoders at the specified settings at that, would give you a solid start (and would probably take less time than this answer took to type ;)), so I'd recommend you go ahead (this is all assuming you don't have a hearing impairment).

$\endgroup$
1
  • $\begingroup$ "(this is all assuming you don't have a hearing impairment)" -- no, that makes it easy. And you save money on stereo equipment, too. $\endgroup$
    – TimWescott
    Dec 29, 2021 at 19:39

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.