Fixing Aqualung's Quad Mix: The Back Left Channel's Incorrect Polarity; Sometimes Laughably Out of Phase LFE; Swapped Back Channels of "Hymn 43"

[TVB]

You said the magic words, "most music simply doesn't contain such content". Why? The stuff that does is mind blowing! Coldplay's "fix you" on cd has some really low bass in the middle of the song where the bass is so low there is a fluttering sensation that I can feel as well as hear. The only way I even knew it was there was by pulling up the 30 hz slider on my soundcraftsman eq. It is really clean and very desirable. There are other songs with ultra low bass where a properly designed LFE channel could absolutely enhance the listening experience. When I listen to a surround music disc MY DREAM is to control/enhance the bass to my sub by adjusting the volume of the LFE channel, but you can't do that if nothings there! What a TRAVESTY that the LFE was only designed for movies and not music. C'mon all you music production guys, TAKE US TO THE NEXT LEVEL and start putting low frequencies in the LFE channel of music discs. Just dreaming, but how about 2.1 digital discs?

 

[ssully]

if you had to 'pull up the 30Hz slider' on your EQ then it wasn't really meant to be 'heard' that way.

And even then, that doesn't mean it requires a separate LFE channel.

 

[ssully]

It's at the bottom of the first post for all the tracks, but to actually answer your question, for the 96kHz blu ray, I moved the LFE 267,088 samples to the right and inverted its polarity (2 sec and 75,088 samples).

Oops. It took me this long to re-read your original post, and this one, to see that you not only shifted the LFE, you inverted its polarity too. That should eliminate the 'macro OOP" I saw.

 

[Fourplay]

MY DREAM is to control/enhance the bass to my sub by adjusting the volume of the LFE channel, but you can't do that if nothings there!

Actually, I have done just that in my system. My method is to use a routing switcher. One position of the router is a straight pass-through of the LFE out signal of the Oppo to the amp. If I find that to be insufficient, with the push of a button I switch to a summed signal of the four main channels (excluding the center channel), and this signal is run through a DBX brand sub-harmonic synthesizer which has several adjustments including gain. It does not work perfectly in every situation, but in most cases it works very well, and I can easily dial in my LFE to taste from my listening position (using a monopod to twiddle the dial).

One example of where I use this is at the climax of One Of These Days from the hidden Meddle mix. The later part of the song is comparatively anemic, but I can preset the DBX and with a touch of a button I can affect the change pretty much right on cue.

"One of these days" I will post my setup in the "show us your gear" thread. I use the same router for channel isolations and reassignments on the fly. Works great!

 

[jimfisheye]

I've seen enough mayhem at this point to know these posts might be serious. I've got slightly different numbers but this is fully out to lunch as described! This is just crazy how the Lfe can be not just out, and not just way out, but a moving target song to song too!

How does this happen?
Someone patching external stuff and screwing up their latency?
Just how do you do this?!

I must have been interested in the 5.1 remix because I sure should have thought the new 4.1 copy of the quad sounded bad. But it's also weird because it's still hifi and some songs even sound overall louder in the vocal presence. Pulling out the old copy of that shared Q4 to DVDA copy that a bet a few people have heard instantly sounded amazing vs the new one.

I remember comparing the bluray and the dts2496 copies of the 4.1 and finding them the same. Apparently I never listened to the 4.1 copy while doing that!


Reverse the polarity of the Ls channel. (ch 3 in quad, ch 5 in 5.1 format)
This is the main one and it's easy. The difference is really shocking and obviously correct if you A/B. The original returns except a little brighter.

Reverse the polarity of the Lfe channel. (All songs) (ch 4 in 5.1 format)
Each song needs a different correction!
Song number and nudge amount in samples (96k) [- = left, + = right]
1. -2128
2. +242106
3. -27069
4. +99161
5. -27131
6. -27131
7. -27069
8. -27069
9. -27278
10. -27097
11. -27097

I also created a low passed track from the front L/R pair to work from visually. Waves linear phase multiband eq. It's hard to call it precisely to the sample. There's still some non linearity. I'll say I'm accurate to 2 samples.

Then there's the business of the rear channels in Hymn 43.
The argument is convincing. Hard to imagine intentional. Not impossible. A mix can be anything it wants to be. It's the same way on the original Q4 for what it's worth. So the mistake went back to the original release. The Ls (ch 3) was correct polarity on the original though. That was a new mistake.

Speaking of that original copy. The one I have anyway.
There's some azimuth error offset between channels. An offset of -18 samples across the channels corrects it. Nudge channel 2 18 samples left (-18). then ch 3: 2 x 18 and ch 4: 3 x 18.

You have to break the thing out into mono channels in 4 tracks in your favorite DAW to do all that. Then render it.

Notice the vocal fall into the center. It sounded panned left before. (because the R ch was delayed)

That old copy sounded a lot better next to the new one before corrected. The new corrected probably sounds better now. Maybe the old one still has a right to stick around though?

But how does this level of crazy happen with a release?

 

[jimfisheye]

I can't be sure without the actual master in front of me to compare to but it sounds like the Lfe contains elements removed from the mains channels as opposed to an errant duplication and thus needs to stay. I think this 4.1 mastering has also been high shelf boosted.

But this is like getting a new Quadio release of this all the same!
Thank you @Dynamic Editor for catching this!!!
Errantly flipping the polarity of the Ls channel really messed with the entire sound! Just that fix alone is a big deal here!

I really want to know what's behind some of these kind of mistakes! I've seen all kinds of software bugs. OldtoolsHD crashing and burning so bad is what led me to Reaper DAW 15 years ago. I suppose I already know... Nobody ever listened to the final master for a safety check. I'm still really curious about what software bug or mistake led to this in the first place though!

That old copy of the ripped Q4 to DVDA is about 4db louder than the 4.1 remaster for anyone keeping score and wanting to A/B between them.