Skip to main content

Topic: IETF Opus codec now ready for testing (Read 234518 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • Gainless
  • [*][*][*]
IETF Opus codec now ready for testing
Reply #550
Has there been any look on fixing up killer samples btw? The unconstrained VBR has definately flaws in some cases, which makes it still safer to use older versions/constrained VBR at high bitrates.
  • Last Edit: 10 February, 2013, 10:15:12 AM by Gainless

  • Garf
  • [*][*][*][*][*]
  • Developer (Donating)
IETF Opus codec now ready for testing
Reply #551
If somebody hasn't seen there is an official Opus 1.1 alpha.

https://ftp.mozilla.org/pub/mozilla.org/opu...alpha-win32.zip

Also some kinda settings to play with
https://wiki.xiph.org/Opus_tuning


Imo official opus binaries should be stickied in a thread of it's own. As posting on a topic like this could likely cause it to be buried.


Maybe a sticky with the latest binaries etc...

  • IgorC
  • [*][*][*][*][*]
IETF Opus codec now ready for testing
Reply #552
It would be really handy. Like "Nero AAC Recommended Settings" http://www.hydrogenaudio.org/forums/index....showtopic=44310
  • Last Edit: 10 February, 2013, 12:38:02 PM by IgorC

  • DonP
  • [*][*][*][*][*]
  • Members (Donating)
IETF Opus codec now ready for testing
Reply #553
I encoded a bunch of various music at bitrate=100 using the "babyeater" release.  The majority showed in foobar properties as between 100 and 110 kb/s, some much higher, and none below target rate.  Then I did the math using the whole file size (including metadata and whatever else besides the audio data) and play time on 2 of the files and found in both foobar was reporting the bitrate higher than it should by a few kb/s.

With the case below, that would be (4 969 864 bytes) * (8 bits/byte) / (1024 bit/kbit) / (374.7 seconds) = 103.6 kb/s


  • nu774
  • [*][*][*][*][*]
  • Developer
IETF Opus codec now ready for testing
Reply #554

  • Jplus
  • [*]
IETF Opus codec now ready for testing
Reply #555
Thanks for pointing that out, nu774. So then it becomes 4969864B * 8b/B / 1000b/kb / 374.7s = 106.1kb/s and Foobar2000 turns out to do it right. Now I'm starting to wonder whether my bitrate numbers from opusinfo were calculated using base 10 or base 2.

After checking: opusinfo does it in base 10, too. Note that I used the "without overhead" bitrates from opusinfo in the post that I linked to above. With overhead the numbers are 0.5-2kbps greater.

  • db1989
  • [*][*][*][*][*]
  • Global Moderator
IETF Opus codec now ready for testing
Reply #556
Done:

I propose that new posts in this thread which do not continue existing discussion here be split off, as Opus has it own subforum.
Opus 1.1 alpha version (not BABYEATER), including softrunner’s post about this version and spectrograms.

Mods, could we get a threadsplit for these quality level posts? I really think this deserves its own thread.
Can audio encoders target quality w/o caring about bit rate/file size?

Some replies somewhat spanned over both topics, as evidenced by my having to begin the latter with another post that (handily!) quoted softrunner’s initial message – if only we could split single posts, too  – but I think this works pretty well.

I agree with the sentiment that things should be posted to separate threads if at all possible since there’s a dedicated subforum for Opus now. I’ll see if anything from the past couple of pages here stands out as being split-worthy, but you’ll excuse me if I don’t look too far back.