INTERACT FORUM

More => Old Versions => JRiver Media Center 23 for Windows => Topic started by: etc6849 on September 10, 2017, 03:01:11 pm

Title: Do most use JRSS upmixing and how does it work?
Post by: etc6849 on September 10, 2017, 03:01:11 pm

Are most with 5.1 setups using JRSS to up convert all stereo tracks?

What exactly is it doing?  Is there a technical write up talking about it somewhere?  All I found in the wiki was under mixing (and it doesn't say much other than how to set it up).  How can something like this work so well without knowing a transfer function approximating the room the recording was in?

Is the delay it seems to add to make the rear channel content (for expanding the stereo playback image) based on a set value or does its algorithm modify this depending on the content's soundstage (for the file being played back)? 

I assume it is doing something similar to what I've seen people do with reflectors/acoustics to create a Haas effect (where there is an added perception of space to the stereo image).

I ask because it is expanding my stereo image in a way that is totally believable and life-like.  Far beyond any AVR or HT processor I've heard.  Not that I use an AVR or home theater processor anymore; because the pro gear I have blows away any consumer gear regardless of price point.

Title: Re: Do most use JRSS upmixing and how does it work?
Post by: Mr. Tuesdays on October 11, 2018, 04:52:14 pm

I know this is over a year old, but I have not found any answer to this elsewhere, so thought I'd post some of my findings.

After extensive tests looking at how JRSS does its upmixing, this is what I found if starting with a 2 channel (stereo) recording upmixing to 5.1:

• If both channels are identical, Left (1) and Right (2) are unchanged.  Center (3) will contain the mono sum, in this case same as Left and Right.  Sub (4) contains low mono. Surround Left (5) and Surround Right (6) contain copies of L&R, but lower in level and delayed by 20ms. This is not normal decoding behavior, no center channel should appear in the surrounds.
• If Left and Right are different then they remain unchanged after upmix.  Center will contain pure Left and Right attenuated, along with the Center (mono) signal at a higher level. Surround Left will contain mostly Left with some center and Right.  Right surround will be similar, only with Right front prominent with Left front and the center at lower levels. Surround is delayed 20ms.
• When the stereo signal contains an out of phase signal (normal surround encoding) that signal will remain in the front channels and will appear in both surround channels at levels identical to the front channels.  The surround channels will also contain some left, center and right. Surround is always delayed. There will be no surround (out of phase) in the Center channel.

Basically, Left and Right remain unchanged, Center gets a mono sum, surround channels get the out of phase signal from the front, as well as some of both left and right, at a lower level.  Surround is steered by levels in the front channels, more front left will end up in surround left, more right front in surround right. There is always some front mono sum in the surround, at a level that is unusually high for most decoding. The surround channels have high crosstalk from the front channels.

Across the front channels any sounds panned hard left or right will end up also in the center channel (but at lower level) pulling them toward the center.  Not unpleasant, but not as wide as pure 2 channel. Might be good for bringing all dialog more toward center.
Decoding to 4 channels is the same, minus the center and sub channels.

Perhaps more than you wanted to know, but that's what I've found.  :P