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

[MrSmithers]

Some draft waveforms - can go into more detail later to establish the exact timeframes... Right back to some music and booze...

Document - correct...

Boz Scaggs

Automatic For The People

Green

Up

Monster

New Adventures in Hi-fi (This looks like a much smaller gap?)

 

[steelydave]

Hmm... I think it could be interesting to see if the LFE channels on the affected mixes are all low-pass filtered, or full-range.
Or who mastered them.

If they're Low Pass Filtered that would add in a Group Delay, so that could be why one is seen,

Very approximately, the Delay, Td, in milliseconds (ms), where n is the filter order, and fc is the 3dB cut-off frequency in Hz of the filter.

Td = 125n/fc

Assuming a Butterworth filter, so maximally flat, a 100Hz cut-off frequency, and wanting to reduce the signal by at least 30dB at 500Hz, you would require a 3rd to 6th order Low Pass Filter.

For filter orders;

3rd: Td = 3.75ms
4th: Td = 5ms
5th: Td = 6.25ms
6th: Td = 7.5ms

I listened to the LFE channel on all 12 of the affected mixes and all of them are low-pass filtered - if this process induces a delay like Dunc has suggested, that may be an explanation for the issue.

Nice find,but isn't it a little funny that many of these discs are on top of the polls and nobody noticed any problems before?I have most of them and I haven't noticed any problems with bass sound on these.

Sure, I'm not saying they don't already sound great, and if you're happy with yours than by all means ignore this thread and save yourself the time and energy. I can't account for how other people hear or rate music, I've seen enough audiophiles endorse nonsense snake-oil technologies to supposedly enhance sound, and even more pictures of people's home setups with terrible speaker alignments or lack of room treatments that I only place a certain degree in these qualitative assessments. Like I said in my original post, this isn't a night-and-day quality improvement like a fully phase-inverted channel, but if something can make your listening experience even 10% better, why dismiss it entirely out of hand?

Beck: Sea Change (BDA) waveform:

Issue1: L, R & LFE phase inverted with C, Ls & Rs
Issue 2: LFE not aligned

EDIT: After fixing this yesterday there was a huge boost in bass!

It's only the FL & FR channels that are inverted, the LFE is correct - if you're inverting your LFE, that may explain why you were saying to me in PM that it was only a 2ms and not 8ms delay that was needed. It's definitely 8ms if you leave the phase of the LFE as-is.

Isn't that the kicker? Some of the titles in question are rated extremely high.

Yes, but like I said in my initial post, if you're running a full-range setup with 'large' speakers and no crossover, you most likely won't notice this issue at all. The Jeff Beck Blow by Blow SACD was similarly highly rated in the polls for 10 years until I discovered the phase inversion issue, and now everyone seemse to accept that I was right. I'm pursuing this because it makes my listening experience better - to anyone who thinks everything sounds fine, I would absolutely say steer clear of this thread and save yourself the grief of doing these fixes.

 

[edisonbaggins]

@steelydave BBB is a great example of an album that sounds anemic on a full-range and bass-managed system. It's the phase inversion of main channels that kills the bass.

I've heard some of these LFE shifted albums on both types of systems and they don't disappoint the way the Sony BBB does.

 

[DuncanS]

I have 4 full range floorstanders + centre and no LFE, so I have never noticed anything Bass wise on the albums that have the error. I doubt if many people who aren't Electronic Engineers (as I am) would even consider that a filter would add a delay, so they wouldn't think about having to correct for it. Hence it would be missed.

 

[steelydave]

@steelydave BBB is a great example of an album that sounds anemic on a full-range and bass-managed system. It's the phase inversion of main channels that kills the bass.

I've heard some of these LFE shifted albums on both types of systems and they don't disappoint the way the Sony BBB does.

Yes I agree with you, phase-inversion problems cause a 100% loss of the affected frequencies. The phase-inversion issues caused by the offset LFE aren't as total, but they are absolutely noticeable - the difference on my phase corellation meter between the original files and the corrected ones is about 50-75% out of phase to nearly 100% in phase, and my headphone listening tests back this up, the corrected files have an audible impact on the tightness and quality of the low-end extension.

 

[HomerJAU]

It's only the FL & FR channels that are inverted, the LFE is correct - if you're inverting your LFE, that may explain why you were saying to me in PM that it was only a 2ms and not 8ms delay that was needed. It's definitely 8ms if you leave the phase of the LFE as-is.

Got ya Dave. Thanks (I'll edit my original post to remove confusion)

 

[edisonbaggins]

Yes I agree with you, phase-inversion problems cause a 100% loss of the affected frequencies. The phase-inversion issues caused by the offset LFE aren't as total, but they are absolutely noticeable - the difference on my phase corellation meter between the original files and the corrected ones is about 50-75% out of phase to nearly 100% in phase, and my headphone listening tests back this up, the corrected files have an audible impact on the tightness and quality of the low-end extension.

Yeah, my main man. Don't nobody get me wrong. Stands to reason that LFE shift causes noticable differences for some listeners on some systems.
Would I gladly download corrected versions of all affected albums? Sure!
Though, because the sound is already satisfying for for me, and the visual experience is pretty intense, for albums like the corrected IA BD, I'll stick with that, mainly.

 

[haikubass]

This is the opening track off Mutt Slang II, Trolls - looks alright to me:

 

[steelydave]

@mr

Yeah, my main man. Don't nobody get me wrong. Stands to reason that LFE shift causes noticable differences for some listeners on some systems.
Would I gladly download corrected versions of all affected albums? Sure!
Though, because the sound is already satisfying for for me, and the visual experience is pretty intense, for albums like the corrected IA BD, I'll stick with that, mainly.

Yeah man, we understand each other. I think especially for people who prefer disc-based playback, this whole discussion is a moot point. But I thought it was worth raising for everyone else, both file-based end users who want to fix their own discs, and any engineers who might be browsing the forum so that they can maybe become aware of a possible issue so it doesn't crop up in the future.

Imagine how many problems would've been prevented over the last decade (or two!) if every mixing and mastering engineer had been aware of the critical necessity of simply checking 5.1 mixes for phase coherency? I think we're incredibly lucky that file-based playback has evolved as it has so that we even have the opportunity to remedy these issues now. For the first 10 years or so as a surround listener, the copy-protection for both SACD and DVD-A was unbroken, so whatever was on the disc stayed on the disc. Power to the people!

(@MrSmithers and @haikubass thanks for the screenshots, I'll update the table in the first post later this evening and then we can work toward figuring out the delay times at some point in the future.)

 

[haikubass]

Here's Steve Lukather, 1st track off Candyman - I'd say that LFE is delayed by 2 ms (see the selection):

 

[Cheezmo]

I looked at the two Beck titles first and completely confirm what was discovered here.

Except, the phase appears correct to me on the last track (Emergency Exit) of Guero.

 

[steelydave]

I looked at the two Beck titles first and completely confirm what was discovered here.

Except, the phase appears correct to me on the last track (Emergency Exit) of Guero.

Wow, I think you're right actually - visually Emergency Exit does look correct compared to all the other tracks. Will double check with the phase meter later to confirm.

 

[HomerJAU]

I should have a new Channel Alignment tool added to Music Media Helper later today to enable easy fixing of affected albums. This will only support FLAC, WAV and DSF files.

I’ve just added support for DSF files to MMH’s Channel Remix tool ‘Invert Phase’ option. Previously only FLAC and WAV were supported. I just need to update docs and I’ll release the new version (4.1.0).

 

[edisonbaggins]

Wow, I think you're right actually - visually Emergency Exit does look correct compared to all the other tracks. Will double check with the phase meter later to confirm.

Have mercy!!!

 

[Cheezmo]

Have mercy!!!

No kidding. Once a can of worms like this is opened it only gets worse and worse as exceptions are found, etc.

 

[Fourplay]

As I recall the original SACD of Norah Jones - Come Away With Me also suffered from an LFE phase inversion. It is great to call these out so they can be repaired. Perhaps once the offenders are all nailed down, a separate concise roundup could be posted. Thanks for your very useful thread!

 

[Eclectic]

What a great piece of empirical research. Well done, steelydave!

 

[winopener]

Maybe a specific area of the forum can be created, "Fix The Mix", including there all the finds done in the past about authoring problems.
One section for analog (which will be 99% Q8) with list of swapped channels etc;
One section for digital, which can include:
- the above issue of LFE phase invert and /or delay
- the messed channels configuration (example: TYA A space in time DVD, swap f/r)
- the "we tried to 5.1fied a quad mix but we did a dud" (example: Deep Purple Stormbringer dvd)
- other stuff (example: the l/r delay on the Sly&FS GH stereo track)

I mean, there's a wealth of information in this forum that is really precious; having a single point for collecting all this stuff will be important for the years to come, especially for the newcomers.

 

[MrSmithers]

Maybe a specific area of the forum can be created, "Fix The Mix", including there all the finds done in the past about authoring problems.
One section for analog (which will be 99% Q8) with list of swapped channels etc;
One section for digital, which can include:
- the above issue of LFE phase invert and /or delay
- the messed channels configuration (example: TYA A space in time DVD, swap f/r)
- the "we tried to 5.1fied a quad mix but we did a dud" (example: Deep Purple Stormbringer dvd)
- other stuff (example: the l/r delay on the Sly&FS GH stereo track)

I mean, there's a wealth of information in this forum that is really precious; having a single point for collecting all this stuff will be important for the years to come, especially for the newcomers.

That'd be really good. Like a separate tab or something? To have an issue/fix HQ would be cool. I have some of the titles listed as -

Neil Young - Harvest - Channel assignment errors on certain tracks
Porcupine Tree - In Absentia - Inverted LFE and centre timing issue
Coldplay - A Head Full Of Dreams - Added LFE info in fronts
Jeff Beck Group - Incorrect speed on first 2 tracks
Jeff Beck - Blow By Blow - Incorrect phasing (corrected on Analogue Productions SACD)
The Beach Boys - Pet Sounds - Wrong Channel assignment track 1
Ten Years After - A Space In Time - Swapped fronts/Rears
Keane - Hopes And Fears - Swapped fronts/Rears
George Benson - Breezin’ - LFE issue
Guns N' Roses - Appetite For Distuction - My Michelle channel delay issue

 

[MrSmithers]

I think New York Voices should be ok? At first it looks like it's spot on, but I zoomed in macro style and it seems to be like half a millisecond out, if that? ☺



Time for your close up...