• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Why bass management makes my life tedious

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
This is not a playback discussion based on personal opinions or preferences. This is a discussion on how the content is produced.
The bass EQ movement is not a real movement. Folks have been EQ'ing their subs for decades.
Did you read the post I linked? It's not about equalizing subs, it's about boosting the LFE content that for some reason is filtered out of some Blu-ray releases. That is where you guys do your work. Measurements are also provided.

I couldn't help but draw a link in my mind when I saw you guys talking about reducing bass as speaker count increases. Maybe there's a link? If not, what else would contribute to what they are experiencing?
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
Did you read the post I linked? It's not about equalizing subs, it's about boosting the LFE content that for some reason is filtered out of some Blu-ray releases. That is where you guys do your work. Measurements are also provided.

So let's go with the mission statement here.


"Bass EQ or BEQ has been talked about on Data Bass forums... since ~2014
for Restoring some Bluray/4K UHD movies that have had their lower frequencies severely filtered.

The ULF content is still there on the disc for the taking,
and with the right software the ULF/LFE levels can be restored.

This is not made up content, noise or filler, it's real low frequency material that "some" sound engineers decided no one needed to hear/feel. So the lowest frequencies output levels were just reduced to soundbar levels, But are still ripe for the picking with the right software. "

So my question would be, does the original poster have the measurements of the actual film printmaster to compare what LFE amount has been rolled off on the disc? The answer to that based on your link is no. There are no comparisons shown here, so the OP is assuming that output has been rolled off for some nefarious reason. That is an assumption that cannot be quantified and certainly hasn't with this link.

What the OP is proposing here is a boost in the lowest of frequencies to compensate for the natural roll-off of LFE content. That roll-off is not there by accident, and it most likely exists in the original printmaster as well. I know I don't make a practice of rolling off the LFE of any content I have mastered for disc (streaming is a different animal).

I couldn't help but draw a link in my mind when I saw you guys talking about reducing bass as speaker count increases. Maybe there's a link? If not, what else would contribute to what they are experiencing?

Our discussion here is about the production side, not the reproduction side.
 
Last edited:

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
Actually, it is not about boosting the LFE content that has been filtered out, it is about creating a house curve that conforms to our hearing insensitivity at low frequencies. If low-frequency content has been filtered out, there is nothing you can do to your sub to bring it back.
But that is what they are doing; boosting lower-level content that is still there. Not all Blu-rays exhibit this. My question is, what are the reasons why it would be reduced so much in level for the Blu-ray compared to the theatrical version?
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
But that is what they are doing; boosting lower-level content that is still there. Not all Blu-rays exhibit this. My question is, what are the reasons why it would be reduced so much in level for the Blu-ray compared to the theatrical version?

What they are attempting to do is compensate for the natural roll-off of the LFE content by applying a huge EQ boost during playback. That is not the same thing we are discussing here.

I saw no comparison between the actual theatrical soundtrack and the disc. The OP does not have access to the original printmaster to make that comparison. Based on that, there is only an assumption that the bass has been filtered or purposefully rolled off.
 
Last edited:

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
You may be right that they don't have measurements of the original theater masters. I shouldn't assume that. But, there is something going on with certain disc releases. It seems to be more prevalent from some studios, and more so from Atmos titles. It seemed to tie in to your discussion on countering bass build-up as speaker counts increase (Atmos!).

I hope this post will clarify what I am referring to.


Here is an example from the Atmos version of War of the Worlds. It is the average bass levels from the movie, from the linked video:

BEQ1.png



And here is a comparison between the DTS-HD (top left) and Atmos (top right) tracks. Look at the chunk of energy taken out below 30 Hz just for the Atmos track. What gives?

BEQ2.png



So my question is:
  • Could this be an attempt to counter the bass build-up you are discussing on higher speaker-count systems?
  • If not, why would it be done?
Thank you.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
You may be right that they don't have measurements of the original theater masters. I shouldn't assume that. But, there is something going on with certain disc releases. It seems to be more prevalent from some studios, and more so from Atmos titles. It seemed to tie in to your discussion on countering bass build-up as speaker counts increase (Atmos!).

I hope this post will clarify what I am referring to.


Here is an example from the Atmos version of War of the Worlds. It is the average bass levels from the movie, from the linked video:

View attachment 134206


And here is a comparison between the DTS-HD (top left) and Atmos (top right) tracks. Look at the chunk of energy taken out below 30 Hz just for the Atmos track. What gives?

View attachment 134207


So my question is:
  • Could this be an attempt to counter the bass build-up you are discussing on higher speaker-count systems?
  • If not, why would it be done?

Thank you.

The simple answer to your question is no, it is not what we are talking about. This comparison IMO is very poor. My first question. Did both encodes come from the same source? The OP cannot confirm this. Without this information, this comparison is not valid. It is likely the DTS mix was from an earlier release(the original Bluray release), and the UHD disc's mix is a newly created one. You cannot compare the two on a one-to-one basis.

When you listen to the first two minutes of the video, the gentleman clearly states that BEQ compensates for the natural content roll-off in the LFE channel. That is a reproduction issue, not a production issue. What you see here is not something based on the encoding process, but is in the mix itself.
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,168
Likes
3,713
I have another issue to raise about LFE content production (not having to do with bass management per se). It's something I've seen in e.g. 5.1 music mixes that exist in Dolby/DTS lossy format and also in some PCM format (like offered on BluRay, or DVD-A). Possibly this issue deserves a separate thread so as not to dilute this one

To wit:

DTS and DD LFE contents are, by spec, steeply rolled off above some low frequency (iiirc 120 Hz?). This is quite apparent in frequency/amplitude plots of LFE channels. In addition, typical AVRs these days have an 'LFE LPF' setting, so users can set the low pass filtering of the LFE output there. On top of that there is typically a filter/crossover setting on the subwoofer itself (sometimes user-defeatable, but always user-adjustable). But the rolloff enforced on the DD/DTS version itself means that at most , the user could allow LFE content bandwidth only up to ~120 Hz to emit from the sub.

HOWEVER: PCM LFE of the same mix, released on BluRay (or perhaps DVDA...or even perhaps DSD, though I forget whether I've seen it there) sometimes shows full range signal in the LFE channel (as seen in freq/amp plots of the LFE channel contents). E/g, a full range bass guitar and/or drum part has been mixed to LFE, but has not been rolled off.

The effect of this (aside from making it super easy for musicians to learn bass and drum parts!), is that the producer of the BluRay version is depending on the listener's system to apply some low pass filtering. Without proper filtering, the subwoofer could easily become localizable. (Also , bass/drum content may even be *duplicated* on other channels, resulting in other issues...), and certainly they are not hearing what was 'intended'.

I could provide screen capture examples of this, if need be.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
I have another issue to raise about LFE content production (not having to do with bass management per se). It's something I've seen in e.g. 5.1 music mixes that exist in Dolby/DTS lossy format and also in some PCM format (like offered on BluRay, or DVD-A). Possibly this issue deserves a separate thread so as not to dilute this one

Okay, so with the home theater-based Dolby Digital encoder back in the day, you had to engage the LFE filter (120hz) before the encoding, it was not a default setting. That was not the case with DTS. 80hz was the default, and you could choose 120hz as well. I know of at least one music disc that did not have the LFE filter engaged before decoding, and it was confirmed by Dolby.


HOWEVER: PCM LFE of the same mix, released on BluRay (or perhaps DVDA...or even perhaps DSD, though I forget whether I've seen it there) sometimes shows full range signal in the LFE channel (as seen in freq/amp plots of the LFE channel contents). E/g, a full range bass guitar and/or drum part has been mixed to LFE, but has not been rolled off.

I would say this is an aberration rather than a standard practice. With PCM LFE, there are plugins you can use to low pass frequencies in the LFE channel, so there is no reason whatsoever that there would be full-range content in the LFE unless this was the sound mixers' intent (which is kinda nuts), or they forgot to use a filter. Also, Chesky Records used the LFE channel as a full-range height channel on some of their SACD offerings.

Without proper filtering, the subwoofer could easily become localizable. (Also , bass/drum content may even be *duplicated* on other channels, resulting in other issues...), and certainly they are not hearing what was 'intended'.


I absolutely agree with this statement.
 
Last edited:

markus

Addicted to Fun and Learning
Joined
Sep 8, 2019
Messages
646
Likes
653
I have another issue to raise about LFE content production (not having to do with bass management per se). It's something I've seen in e.g. 5.1 music mixes that exist in Dolby/DTS lossy format and also in some PCM format (like offered on BluRay, or DVD-A). Possibly this issue deserves a separate thread so as not to dilute this one

To wit:

DTS and DD LFE contents are, by spec, steeply rolled off above some low frequency (iiirc 120 Hz?). This is quite apparent in frequency/amplitude plots of LFE channels. In addition, typical AVRs these days have an 'LFE LPF' setting, so users can set the low pass filtering of the LFE output there. On top of that there is typically a filter/crossover setting on the subwoofer itself (sometimes user-defeatable, but always user-adjustable). But the rolloff enforced on the DD/DTS version itself means that at most , the user could allow LFE content bandwidth only up to ~120 Hz to emit from the sub.

HOWEVER: PCM LFE of the same mix, released on BluRay (or perhaps DVDA...or even perhaps DSD, though I forget whether I've seen it there) sometimes shows full range signal in the LFE channel (as seen in freq/amp plots of the LFE channel contents). E/g, a full range bass guitar and/or drum part has been mixed to LFE, but has not been rolled off.

The effect of this (aside from making it super easy for musicians to learn bass and drum parts!), is that the producer of the BluRay version is depending on the listener's system to apply some low pass filtering. Without proper filtering, the subwoofer could easily become localizable. (Also , bass/drum content may even be *duplicated* on other channels, resulting in other issues...), and certainly they are not hearing what was 'intended'.

I could provide screen capture examples of this, if need be.

On newer delivery formats the LFE is defined as a full bandwidth channel. The 120Hz LFE filter during encoding isn't mandatory. So there are plenty of recordings with higher frequency content never intended to end up in the encoded bitstream.
The solution is to apply a low-pass filter in the reproduction device. That isn't ideal as this presents an arbitrary phase shift (LP might or might not been switched on in encoding) which can lead to cancellation effects. Still better than doing nothing.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
On newer delivery formats the LFE is defined as a full bandwidth channel. The 120Hz LFE filter during encoding isn't mandatory. So there are plenty of recordings with higher frequency content never intended to end up in the encoded bitstream.
The solution is to apply a low-pass filter in the reproduction device. That isn't ideal as this presents an arbitrary phase shift (LP might or might not been switched on in encoding) which can lead to cancellation effects. Still better than doing nothing.

I have said this for decades about applying the LPF in the AVR or processor at 120hz with movie soundtracks. We already do that on the production side, so there is no reason in the world to apply that to the reproduction side for the very reason you mentioned. I would rather do nothing until I have to do something when it comes to the LPF. Leave it off until there is a need to turn it on.
 
Last edited:

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
I have said this for decades about applying the LPF in the AVR or processor at 120hz with movie soundtracks. We already do that on the production side, so there is no reason in the world to apply that to the reproduction side.
How confident are you that everyone does it? Confident enough that I can raise my LPF for LFE to 250 Hz and not hear a difference?
 

markus

Addicted to Fun and Learning
Joined
Sep 8, 2019
Messages
646
Likes
653
I hope no one does that, I do not know of any processor that would not clip the signal internally for such a signal.

For 7.1, the worst case is approx. 125dB, and some processors can handle that, you need to keep trim levels at minimum though. I have not found any movies with such low-frequency level, but some can have both LFE and LCR maxed out, gives you around 120dB or so.

It would be desirable SSP's allowed the user to properly monitor the redirected subwoofer channel. That way one could adjust required headroom in order to optimize gain for the whole system. It's unfortunate no manufacturer provides such a feature.
 
Last edited:

markus

Addicted to Fun and Learning
Joined
Sep 8, 2019
Messages
646
Likes
653
I have said this for decades about applying the LPF in the AVR or processor at 120hz with movie soundtracks. We already do that on the production side, so there is no reason in the world to apply that to the reproduction side for the very reason you mentioned.

Unfortunately there are recordings that do not have this filter applied on the production side. So what to do? Have a process in the SSP that monitors the LFE for higher frequency content and dynamically engage a LP?
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
Unfortunately there are recordings that do not have this filter applied on the production side. So what to do? Have a process in the SSP that monitors the LFE for higher frequency content and dynamically engage a LP?

The lion's share of recordings DO have the filter applied. I would leave the filter off until I actually needed it. I wouldn't want to compromise the fidelity of the majority just to compensate for a small minority.
 

markus

Addicted to Fun and Learning
Joined
Sep 8, 2019
Messages
646
Likes
653
The lion's share of recordings DO have the filter applied. I would leave the filter off until I actually needed it. I wouldn't want to compromise the fidelity of the majority just to compensate for a small minority.

Checking for unwanted LFE content before actually watching a movie is not a practical solution ;)
Last time I checked a couple of Blu-ray discs it didn't appear to be an issue only affecting a small minority of movies. But admittedly that was some years back.
 
Last edited:

markus

Addicted to Fun and Learning
Joined
Sep 8, 2019
Messages
646
Likes
653
It is limited to 12 dynamic objects for streaming, and 16 for the disc. I don't think there are many(if not any) situations where all of the channels are pushed to reference levels simultaneously. I would also not look at peak levels established for cinemas to be applied to HT mixes. They are monitored at lower levels than cinema-based mixes are.

Worst case 12 elements would translate to 115dB SPL (LFE) + 12 x 105dB SPL = 128dB SPL?

I'm still not confident this is how the worst case scenario for the subwoofer channel in a bass managed system could be calculated. Once everything gets decoded and rendered it becomes impossible to have 0dBFS in all speaker channels?

@audio2920 ?
 
OP
A

audio2920

Active Member
Joined
May 21, 2021
Messages
233
Likes
288
I prefer to mix/master so it is not an issue. Bass build-up is a known problem ever since the introduction of discrete 5.1 and bass management. Back then there were no tools to address this issue, and that is why some movies transferred directly from the cinema had too much bass when played back. This is why we started doing very specifically made-for-home theater mixes, so we could address issues like this, and have a better translation of the large room mix to a smaller room.

Agreed. We're both saying you have to do an HE mix, which we all do. And I'd go on to suggest that somewhere between your "let the chips fall" argument and my "try and make it play nicely on all systems" is where things land in practice, while using 7.1.4 with BM as the primary format for mixing HE.

And another "Yes" from me; BM handling is not an Atmos/Immersive specific thing.

I could be wrong, but the limiter is dynamic. I don't think Dolby would create a limiter that specifically addresses bass build-up as a static limiter. Dolby is not clear about this, but it is clear they knew about the issue and designed a fix for it.
So, my understanding of the limiter is that's it's just a peak limiter applied to the "speaker" output of the renderer, before passing that audio to the next part of the AVR/system (either the next block of an SoC solution, or another chip) BUT this is before the AVR's attenuation of course, such that it's effect is the same at all listening levels. AFAIK the renderer works in floating point so it can just do the render to whatever your room config, then limited the sums back such that no output leg exceeds 0dBFS (actually 0dBTP? Dunno, and irrelevant).

I also believe the limiter is "channel linked" to try and maintain positional and spectral balance. (Not sure to what extent, maybe 100% link or maybe less but as you say, details on the limiter are quite thin on the ground so, I too could be wrong!). Since it's speaker based, not object based, in a mix where there's sound going all over the place, a room config with less speakers will likely cause the limiter to work harder than a higher speaker count would.
 
OP
A

audio2920

Active Member
Joined
May 21, 2021
Messages
233
Likes
288
Worst case 12 elements would translate to 115dB SPL (LFE) + 12 x 105dB SPL = 128dB SPL?

I'm still not confident this is how the worst case scenario for the subwoofer channel in a bass managed system could be calculated. Once everything gets decoded and rendered it becomes impossible to have 0dBFS in all speaker channels?

@audio2920 ?

So the short answer is, 115dB, because it's my understanding that the limiter (in Atmos) will kick in above that.

That's certainly true of the implementation in the mastering unit (RMU) we use in the studio. I guess it's possible that in a particular AVR the implementation of bass management could be done post Dolby's limiter, but I'm not sure a manufacturer would do that in practice, as they'd need some other way of managing the headroom.

But as @Soundmixer says, it's not even 115dB because we don't mix HE at 85dB ref. It comes out lower than that for all the current loudness normalisation specs.

However, purely academically, I think the LFE uses up one of those 12 spatial objects, so you're left with 11, and the maths would be something like:

LFE = 115dB
11 Objects x 105dB = 20log11 + 105 = 126dB
126dB+115dB = 20log(10^(115/20)+10^(126/20)) = 128dB

I think..... :) So the limiter would pull 13dB off the LF output under these conditions.

Last thought - and it's outside my knowledge area and also irrelevant - I'm not sure the spatially encoded objects actually have a 0dBFS limit while "inside" the renderer and transport stream. If we only sent 12 objects to the RMU, then 0dBFS is the max for each one, as our MADI/Dante stream we feed it with is 24-bit fixed point. But if we were to send like 100 objects all really loud, would the spatial encoder maintain their level when summing these down to 12? I think it might, using 32-bit floating point which gives like 1000dB headroom above 0dBFS. The limiter would then just work really really hard on decode/render. I've no idea if this is what happens, nor do I intend to find out :D
 
Top Bottom