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

[ssully]

Thought I'd check the DVD-A of Gaucho:
1: not sure if the center or the LFE (or none) that's off on this one
2: I listened just to the center at this point in the song, and it appears to be mostly bass.
3: I'm not sure how all Low Pass Filtering works, but I included Spectograms of the song Babylon Sisters to see what it looks like, I know if I utilize LPF in Audacity it just gives you specific set Roll-off db per octave and looks crude to me in application. Also applied to center channel as an example at 20K @48db in the third pic.
*Yes, we need a section of the forum to dive deeper into this.
If none of this seems correct or applicable, please delete!

View attachment 63977

This looks fine. I don't get what you're talking about.

 

[AYanguas]

While I have the capability to fix this issue on my rips, I just don't have the ambition to undertake it as yet. I too would like too hear an explanation from the pros before doing so.

+1

Difficult to get any good pros stating that you will have a really better sound.

The amount of improvement that one can get can be evaluated by listening comparatively the samples of "Sea Change" that were posted above. If you listen more or less "the same" all samples, then those few milliseconds don't change your listening experience at all.

 

[ssully]

While it is good to know the discs and tracks that this problem is on. Once I find out that the LFE on a 5.1 mix contains bass that is low-pass filtered (upmixed) from the other channels, and not mixed that way from the multis, I am inclined to keep that channel turned down or off. Sub is off 90% of the time anyway.

Low pass filtering is the same thing that any subwoofer with a 'crossover' dial or switch does: allowing input frequencies *below* a certain threshold, to pass through to the speaker, rejecting everything above.

AVR 'LPF for LFE' controls do the same but only for LFE content.

As for how LFE is created, conceptually it could be either of these
-- lowpassing a full range source and also *highpassing* that source so it does not duplicate the LFE bass
-- lowpassing a full range source without changing it, resulting in 'duplicated bass'

this applies both to a source that is the original multitracks, or some intermediate mix. (even if you mix bass and bass drum directly to LFE at full range, you will still have to duplicate their higher frequency content in some other channels)

 

[DuncanS]

Thanks for the link!

Yup, the delay is likely some combination of group delay and phase shift*, although admittedly that's still part of what I don't understand about it. Is group delay just a type of phase shift, or should they be thought of as separate phenomena? I need to sit in a quiet room, with a liter of coffee and a good book or 3, and figure this stuff out one of these days. Maybe some biscotti too.

* (plus maybe a bit of random digital latency weirdness just for good measure)

If you put any frequency into any filter you will get a phase shift with respect to the input. Group delay is the effect of Phase Shift over the frequency range in the pass band, it shows up non-linearity in the filters phase response. In a low pass filter you will see bigger phase shift at the cut-off and beyond than you will in the pass band.

 

[DuncanS]

The low pass filter in the subwoofer unit will add a group delay.

 

[AYanguas]

If you put any frequency into any filter you will get a phase shift with respect to the input. Group delay is the effect of Phase Shift over the frequency range in the pass band, it shows up non-linearity in the filters phase response. In a low pass filter you will see bigger phase shift at the cut-off and beyond than you will in the pass band.

So, this means that if the LFE phase shift is caused by group delay, the proposed correction here, aligning the waveforms, would only be correct for that frequency at we see the peak of the waveform, But for other frequencies will not be that same delay, so we will be introducing some kind of 'frequency distortion'. Of course inaudible, I think. But just to understand the concept

 

[ssully]

Nope, its why I don't have a LFE sub as its an effect channel, for it to work properly you should be feeling its effect! I'm not that much of a fan of the Centre either, as its really there to centre the speech in the middle of the screen in cinemas as people are sitting across a wide area.

As I've posted before , a dedicated center channel (and proper mixes using it) have improvement effects vs stereo that go beyond simply stabilizing the image for people not sitting at the best listening position*. Floyd Toole and James Johnston have both discussed them.

An LFE channel, though , really was just meant for effects that were loud and have very low frequency components. There might be some electronic music that fits the bill, but most music doesn't.

* and for that matter, do you keep your head perfectly still at all times when listening?

 

[ssully]

@steelydave, you are the only one in this whole thread who's brought up that it's the playback system's job to send the appropriate low frequency material to the subwoofers.

It's called a "Low Frequency Effect" channel, not a "Subwoofer" channel. In their wisdom, the creators of 5.1 sound knew that it needs to be the responsibility of the playback receiver to crossover music material to the subwoofer. People who mix music and movies don't know what kind of subwoofer / satellite speakers you have, and what the crossover points are. Your receiver knows. To arbitrarily have some frequency range of music in the LFE channel is just complete folly, but somehow only a few 5.1 sound mixers seem to understand this.

With the possible exception of the canons in the "1812 Overture", no one should be mixing 5.1 music with any material in the LFE channel.

It kind of looks like most of these mixes are sending the musical low frequencies to the LFE channel, but not applying a similar hi-pass filter to the rest of the channels. This means it might be appropriate to mute the LFE material, let your receiver decide what to send to the subwoofer, and boost the subwoofer volume to compensate for the junky LFE data that we blocked.

Am I crazy?

No, you aren't crazy. Using the LFE for music is usually unwarranted, and only creates potential issues that are unnecessary.

 

[ssully]

Yes I completely agree with you, there's no point in lining up waveforms just for the sake of it. Believe me, I wouldn't have even started this thread if it was just an OCD thing.

Yes, cancellation is a real thing -- though I haven't got around yet to comparing the effect on Sea Change, a DVDA mix that never struck me as lacking bass. I'm planning to do that comparison, hopefully in some blind fashion.

There's also the fact that some mixes have this issue...and some don't. Suggesting that it was something that it was felt needed to be 'corrected'....

 

[DuncanS]

So, this means that if the LFE phase shift is caused by group delay, the proposed correction here, aligning the waveforms, would only be correct for that frequency at we see the peak of the waveform, But for other frequencies will not be that same delay, so we will be introducing some kind of 'frequency distortion'. Of course inaudible, I think. But just to understand the concept

Yes you are in effect only aligning at one frequency, you're are not adding a frequency distortion (so not adding anything to the signal) but in effect producing phase 'errors' across the frequency band of interest. It is possible to design "All Pass Filters" to do time/phase correction, they can get quite complex. Better to dump the LFE!

 

[AYanguas]

Yes you are in effect only aligning at one frequency, you're are not adding a frequency distortion (so not adding anything to the signal) but in effect producing phase 'errors' across the frequency band of interest. It is possible to design "All Pass Filters" to do time/phase correction, they can get quite complex. Better to dump the LFE!

OK. It is phase 'errors' the concept I was thinking about, not frequency distortion.

Thank you for the explanation.

 

[Jim the Oldbie]

So far, I think the one solid conclusion here upon which we can all agree is: Music + LFE = one big hot greasy MESS.

 

[mkt]

“The LFE channel should never be used to carry the bass content of the main speaker channels”
http://www2.grammy.com/pdfs/recording_academy/producers_and_engineers/5_1_rec.pdf

 

[quicksrt]

“The LFE channel should never be used to carry the bass content of the main speaker channels”
http://www2.grammy.com/pdfs/recording_academy/producers_and_engineers/5_1_rec.pdf

Yes, and I would expect that (may be wrong here) any settings that send low frequencies to a LFE, would also take some of the lower end away from those main speakers. In other words, if they want to send some low info (generated from the mains), they should not subtract anything to do so.

I don't know, I'm just not into upmixing as a stylistic choice.

 

[Eclectic]

I just stumbled on something amusing. The Producers & Engineers Wing of The National Academy of Recording Arts & Sciences, including Scheiner, were familiar with this issue in 2004. Take a look at Section 5.3 of their Recommendations For Surround Sound Production. Here's the relevant prose:

"Even on projects destined for release on SA-CD or DVD-Audio, the surround mastering engineer should filter the LFE channel, despite the fact that the specification does not require such filtering. Limiting the frequency range of the LFE channel to only the bottom few octaves allows for more consistent playback in different environments. (See section 4.6.1) However, this does raise the spectre of potential latency issues, since the main channels remain unfiltered. If the LFE channel is delayed even slightly, it can completely cancel redirected bass content from the main channels when monitored on a bass-managed system. Therefore, the surround mastering engineer must listen on a bass-managed system to check the results of combining the low frequency signals. In some cases, it may be necessary to time-shift the main channels in order to ensure that all channels remain in absolute perfect phase after the filtering process. "

 

[MrSmithers]

I just stumbled on something amusing. The Producers & Engineers Wing of The National Academy of Recording Arts & Sciences, including Scheiner, were familiar with this issue in 2004. Take a look at Section 5.3 of their Recommendations For Surround Sound Production. Here's the relevant prose:

"Even on projects destined for release on SA-CD or DVD-Audio, the surround mastering engineer should filter the LFE channel, despite the fact that the specification does not require such filtering. Limiting the frequency range of the LFE channel to only the bottom few octaves allows for more consistent playback in different environments. (See section 4.6.1) However, this does raise the spectre of potential latency issues, since the main channels remain unfiltered. If the LFE channel is delayed even slightly, it can completely cancel redirected bass content from the main channels when monitored on a bass-managed system. Therefore, the surround mastering engineer must listen on a bass-managed system to check the results of combining the low frequency signals. In some cases, it may be necessary to time-shift the main channels in order to ensure that all channels remain in absolute perfect phase after the filtering process. "

Wow, that just about sums everything up perfectly!

 

[MrSmithers]

I've got a 4ms LFE delay for Boz Scaggs - Dig.

Listened last night and the bass sounded neater. Not night and day like Beck but an improvement still.

 

[AYanguas]

Now, I'm playing with the SpecWeb 2.2 tool for upmix from 2.0 to 5.1. Really nice tool and very easy with the "defaults". results are impressive! Much more aggressive in the voices and instruments separation than our AVR upmixers.

The thing related to this thread is what the SpecWeb manual says about the LFE channel. I have remarked what could be a reason for generating content for the LFE in music.

"""""""""" From SpecWeb 2.2 Guide.pdf """"""""""

The LFE channel is the .1 in 5.1. LFE stands for Low Frequency Effects. It is not the subwoofer channel. It’s also called the .1 channel because it is not a full frequency range channel, typically only carrying sounds up to about 120Hz.

The LFE channel is NOT the “subwoofer channel. Whatever system you use to listen to 5.1 music does “bass management” and (assuming you don’t have full range speakers) extracts the subwoofer signal from the 5.0 surround channels, and sends it on to your subwoofer.

Technically speaking, the LFE channel should not really be used for surround music. It’s meant to have sounds that DON’T appear in the other 5 channels (special effects sounds). However, in stereo to surround conversions most people expect 5.1 vs. 5.0 and you can add some LFE to give the deep bass a little kick.

If you do decide to go with 5.0, SpecWeb will run a little faster, and will create an empty LFE channel, so your output files are still in ITU order (LF, RF, C, LFE, LS, RS).

If you stick with 5.1 (the default) we have ensured that any signal in the LFE channel is time and phase aligned with the other five channels. This is to avoid any low frequency cancelation in your subwoofer or listening environment, due to adding sounds in the LFE that are also present in the other 5 channels.

SpecWeb’s LFE is an infinite slope digital filter (runs in the frequency domain). Any sounds below 20Hz or above 90Hz are rejected. The frequency response between 20Hz and 90Hz is flat.

"""""""""""""""""""""""""""""""""""""""""""""""""""""""

 

[quoddiophile]

There is another potential issue with the Toto IV SACD. At the end of each track (except for some reason, track 8 "We made it"), there is a clearly visible, identical artifact on the Roon waveform. It results in a loud and annoying popping noise, almost like a very loud, sharp drum thwack. Any ideas as to how to get rid of this? These artifacts do not appear on the Roon waveform for the original uncorrected version. Audacity does not work with dsf files. Also, I have to wonder if this is an issue on other SACDs but I cannot check as this is the only SACD on the list that I own.

Jim

 

[Cheezmo]

Some of this stuff is just bonkers. Tarkus' LFE is inverted and delayed 20ms.