HydrogenAudio

Lossy Audio Compression => Opus => Topic started by: wswartzendruber on 2012-10-07 22:01:29

Title: Opus, Coupling, and Neo:6
Post by: wswartzendruber on 2012-10-07 22:01:29
I routinely use DTS Neo:6 to listen to music.  I was listening to Amy Grant's "Oh How the Years Go By" track and noticed severe artifacts in the surround channels.  Pro Logic II exhibited similar results, although not nearly as noticeable.  When I listened to the same track with coupling disabled, the artifacts were far less severe for Neo:6 and similar for Pro Logic II.

The bitrate used was 128 Kbps and the encoder was opus-tools-0.1.5 with libopus-1.0.1.
Title: Opus, Coupling, and Neo:6
Post by: Speckmade on 2012-12-13 20:09:26
If I understand correctly, to me it sounds like nothing to wonder about. Like an all fabulous and fascinating human being leaving a piece of information as useful as his phone number painted on your car with lipstick and you wondering why the phone number is gone after your car comes out of the car wash. If you feed a stereo signal to some lossy codec I'd expect it to try to give back something perceptually close to the auditive impression of just that stereo signal.
Title: Opus, Coupling, and Neo:6
Post by: wswartzendruber on 2013-03-06 01:17:44
Might I suggest that there's an issue with the channel coupling?  Granted, this might just be part of the codec, and I understand that this could be the case.  Still, it shows that coupling introduces artifacts.

By the way, Neo:6 works by splitting the signal into ~19 subbands (on my receiver) and dematrixing those individually.