Fixing Rumours 5.1: Out of phase center channel of "Go Your Own Way" and out of phase LFE channel

[Dynamic Editor]

Hello! I have recently found that the center channel of Fleetwood Mac's "Go Your Own Way" off the DVD-A release of Rumours is out of phase with the rest of the music, and the LFE channel is out of phase with the rest of the mix.

The center channel appears to be around 2,880 samples too early (at 96kHz), and gives much of what it is shared between the center and front two channels a tube-like, double-tracked quality to them, especially the vocals. This I feel especially ruins the vitriol of Buckingham's vocals. The LFE as well seems to be out of phase across the album, arriving 2,049 samples early for "Go Your Own Way" specifically, and 392 or so samples late for the rest of the album (at 96kHz).

To show the difference, linked here are four 26-second snippets of "Go Your Own Way", as follows:

• Go Your Own Way (Original Snippet, Stereo fold of first three channels)
• Go Your Own Way (Fixed Snippet, Stereo fold of first three channels)
• Go Your Own Way (Original Snippet, 5.1)
• Go Your Own Way (Fixed Snippet, 5.1)

Keep in mind that the LFE is not included in the first two examples. (Note that I am sharing these links for the purposes of critique rather than piracy, which is why I cut them to 26 seconds.) I also noticed issues with other surround sound mixes as well, which I hope to post about soon!

 

[ar surround]

Ah ha! I can never make it through the 5.1 of this song, and I often skip it altogether. Perhaps your discovery of phase issues is the reason why. You provide a nice description of what it sounds like, but my summary description is that it sounds like shit.

Good find!

 

[ssully]

Let's see, for CD-rate audio, 1 sample = 1/44,000 th of a second. That's 22 micro seconds, far below audibilty for latency. ~400 times that is still just shy of 9 milliseconds -- latency that is on the bare edge of audibility. For the DVD audio where the SR is 96,000 /sec obviously the number will be more than twice as low. So is that LFE latency really a problem?

There's a better case for correcting the 2K sample @ 96kHz (~21 milliseconds) latency of the C channel

 

[Dynamic Editor]

Let's see, for CD-rate audio, 1 sample = 1/44,000 th of a second. That's 22 micro seconds, far below audibilty for latency. ~400 times that is still just shy of 9 milliseconds -- latency that is on the bare edge of audibility. For the DVD audio where the SR is 96,000 /sec obviously the number will be more than twice as low. So is that LFE latency really a problem?

There's a better case for correcting the 2K sample @ 96kHz (~21 milliseconds) latency of the C channel

I appreciate the criticism! Wouldn't the latency audibly affect the bass production in the subwoofer, though (assuming, of course, your speakers aren't built to playback those frequencies)?

 

[Cheezmo]

Surely you have seen this thread? It is a widespread problem...

DIGITAL - At Least a Dozen 2002/2003 Elliot Scheiner 5.1 Mixes Have a LFE Issue (info/list/fixes inside)

So I can't take full credit for making what I think is a pretty big discovery, but recently a user on Reddit put up a post that made a pretty watertight case that the LFE channel on Beck's Sea Change is 8ms delayed from the 5 main channels. That got me thinking because as we know, Porcupine...

www.quadraphonicquad.com

 

[ssully]

I appreciate the criticism! Wouldn't the latency audibly affect the bass production in the subwoofer, though (assuming, of course, your speakers aren't built to playback those frequencies)?

Like I said, at such a small latency between LFE and the rest, it is questionable whether it is audible. I see figures quoted ranging from 6 ms to 10 ms as the minimum audible. The LFE latency at 96khz SR that you are noting amounts to ~4 ms.

The Center vs rest latency you report, however, is certainly above those minima.

 

[Dynamic Editor]

Like I said, at such a small latency between LFE and the rest, it is questionable whether it is audible. I see figures quoted ranging from 6 ms to 10 ms as the minimum audible. The LFE latency at 96khz SR that you are noting amounts to ~4 ms.

The Center vs rest latency you report, however, is certainly above those minima.

Hmm... Alright! Good to know! Thanks again for your thoughts!

 

[Dynamic Editor]

Does anyone know if the out-of-phase center on "Go Your Own Way" is an issue for the SACD as well?

 

[ar surround]

Does anyone know if the out-of-phase center on "Go Your Own Way" is an issue for the SACD as well?

I thought that the presentations on the DVD-A and the SACD were supposed to be identical?

Also, I'm thinking that an alternate way to fix this to take a rather accurate demixing program (such as DeMix Pro or the pay-as-you-go LALAL.AI), extract the bass stem from the fronts and remake the LFE track choosing a LP filter and slope of one's preference.

Here is a spectral image of the GYOW LFE:

 

[Dynamic Editor]

I thought that the presentations on the DVD-A and the SACD were supposed to be identical?

Also, I'm thinking that an alternate way to fix this to take a rather accurate demixing program (such as DeMix Pro or the pay-as-you-go LALAL.AI), extract the bass stem from the fronts and remake the LFE track choosing a LP filter and slope of one's preference.

Here is a spectral image of the GYOW LFE:

View attachment 95701

But this doesn't fix the issue of the center channel audio of "Go Your Own Way" being out of phase, does it?

 

[ar surround]

But this doesn't fix the issue of the center channel audio of "Go Your Own Way" being out of phase, does it?

You are really going to make me work, aren't you? LOL

If I take the front channels and isolate the phantom center only, I hear that the only vocal is Buckingham's solo. So his vocal can be extracted from the front phantom center using one of the AI's (DeMix Pro or LALAL) and used to create a new center channel.

I'd probably do some EQ also to give the tune a bit more body.

 

[Dynamic Editor]

You are really going to make me work, aren't you? LOL

If I take the front channels and isolate the phantom center only, I hear that the only vocal is Buckingham's solo. So his vocal can be extracted from the front phantom center using one of the AI's (DeMix Pro or LALAL) and used to create a new center channel.

I'd probably do some EQ also to give the tune a bit more body.

Why not just adjust the start of the center channel? That's what I did. I feel like you're doing gymnastics when there's a much more straightforward and simple solution.

 

[ar surround]

Why not just adjust the start of the center channel? That's what I did. I feel like you're doing gymnastics when there's a much more straightforward and simple solution.

Because the track grates my ears in all of its incarnations. I never make it through without skipping to the next song. So I see no other way forward other than to tear it into pieces and rework the whole thing to my liking.

 

[Dynamic Editor]

Because the track grates my ears in all of its incarnations. I never make it through without skipping to the next song. So I see no other way forward other than to tear it into pieces and rework the whole thing to my liking.

Did you hear it with my proposed fixes?