(Opus)
It's basically the pinnacle of audio encoding at this point, having
merged the best ideas from CELT, Silk and a few entirely new ones.
It would be hard to see how any proprietary codec vendor could
compete except where addressing a very narrow niche.
- -
Low delay AAC, in various versions?
What about EVS?
https://www.researchgate.net/profile/Anssi_Raemoe/publication/282605143/figure/fig6/AS:281480141651970@1444121503098/Combined-results-with-all-72-listeners-and-all-signal-types-with-increasing-bitrate-in.png
Opus is really good. But the “pinnacle”?
http://www.aes.org/technical/documentDownloads.cfm?docID=548
“A narrow niche? “ 😉🍷
I would see EVS (more or less) as the low-delay version of USAC.
Best,
Stefan Schreiber
- - -
Citando mgraves mstvp.com <[email protected]>:
Chris,
Actually, I too come from a broadcast background, having installed
graphics systems into production and master controls for over 25
years. I completely appreciate the demand for hard real-time and
zero latency.
I've tracked Opus since its earliest days in the IETF CODEC working
group. The standard has many operative modes. It's absolutely
capable of full-bandwidth, in both lossy and lossless modes.
You will find it both in the production/contribution side of the
house (remote codecs, STL, etc.) and distribution. It also dominates
video conference space.
It's basically the pinnacle of audio encoding at this point, having
merged the best ideas from CELT, Silk and a few entirely new ones.
It would be hard to see how any proprietary codec vendor could
compete except where addressing a very narrow niche.
Michael Graves
[email protected]
http://www.mgraves.org
o(713) 861-4005
c(713) 201-1262
sip:[email protected]
skype mjgraves
-----Original Message-----
From: Sursound <[email protected]> On Behalf Of Chris Woolf
Sent: Friday, May 31, 2019 5:45 AM
To: [email protected]
Subject: Re: [Sursound] wifi audio (was Re: Deconstructing soundbar
marketing B.S.)
On 30/05/2019 17:51, mgraves mstvp.com wrote:
The RF issue of range, carrier frequency, channel width is quite
separate from the deliverable audio path.
The Opus audio codec has revolutionized audio coding. It's able to
deliver full-bandwidth audio at bitrates not much more than what
was once typical of a telephone call. This means that the RF band
need not be large to deliver high quality audio over a digital link.
This answer is quite revealing of the different approaches and
requirements within our audio field. My background is broadcast
audio, so for origination purposes any digital coding has to be
lossless, and latency has to be ~very~ low. Lossy coding is fine as
a delivery format (and so would be OK for speaker feeds) but if the
sound has to be processed en route the psychoacoustic stuff doesn't
stand up. Likewise latency of 5-10ms can begin to alter performance,
depending upon how the foldback is returned to an artist.
I don't know Opus but having read up its spec (on Wikipedia) it is
lossy and so can only be used as a delivery format. I had to smile
at 30ms latency being reported as adequate for musicians to feel
"in-time" - not for the ones I've ever worked with. Likewise the
suggestion that 45-100ms is acceptable for lipsync is laughable -
that's up to 5 TV frames adrift. Maybe audiences have become inured
to low quality standards. Latency for "live interaction" at each end
of a phone line, and face-to-face a few feet apart in a room require
very different standards - Opus's suggestion of 150ms for VOIP might
just be acceptable for the first, but it would destroy the second
application.
I don't doubt that it is a clever and well-designed codec, and that
it is extremely useful, but one must keep in mind what it ~actually~
is rather than what it sounds like. Opus doesn't deliver full
bandwidth audio, any more than other digitally compressed systems
do. It delivers something that convinces most ears that it is a full
bandwidth, full dynamic range signal, but it must always be
remembered what is missing.
If you used such a system to deliver sound to speakers (assuming
there is a technique for maintaining multichannel phase coherence)
it should work perfectly well. If you used it for passing the output
channels of a microphone I doubt you would not remain happy for long.
Which also means that the statement "the RF issue of range, carrier
frequency, channel width is quite separate from the deliverable
audio path" must be very carefully qualified - it is only correct in
very specific circumstances.
Chris Woolf
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<https://mail.music.vt.edu/mailman/private/sursound/attachments/20190531/f2e08f40/attachment.html>
_______________________________________________
Sursound mailing list
[email protected]
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe here, edit
account or options, view archives and so on.
