• 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

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
Wow, thanks for the information dense response. Essentially, the wider dynamic range intended for theaters may reduce dialogue intelligibility at home because during silent scenes when it's only dialogue, the HE user turns up the volume only to be blasted during action sequences. So you have to add compression while emphasizing the dialogue frequency range? This must be why AVRs have their own version of dynamic range compression for "late night watching" mode.

This is correct. Commercial theaters are very quiet spaces compared to most home listening rooms. We also sit much farther from the speakers in a commercial theater than we do in our homes. Both of these things have to be compensated for, hence the slight compression to keep the dialog above the ambient level of the room, while not having the effects knock you out of your seat when it kicks in.

The Late night mode gives the end-user another level of compensation so they can "tailor" the dynamic range specifically for their home listening environment.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
3) Although I wasn't going to admit it, I'm one of those who runs without BM for my living room :oops: I know, I know..... I do have a sub though. It just sounds nicer in that room, and I have less consistency issues with BM bypassed, even if it isn't as good scientifically. (This is on the verge of a "stereo bass" debate, but let's not go there :D)

Coming clean as well, I don't use bass management on my LCR mains in my large room system. You are right, it does sound better (IMO), but it does present some acoustical challenges as well. I know full well when you have bass coming from the mains and the LFE simultaneously, there is going to be some boosting and cancellation occurring in the room. The interesting thing is when running Audyssey, it sets those speakers to large as well.
 

FishInLA

Member
Joined
Aug 12, 2021
Messages
28
Likes
22
Yes. Streaming presents all kinds of challenges when it comes to audio, so you could not use a mix designed for the disc on that platform.

Can you speak a bit more to this? My understanding of most workflows is that there is a theatrical mix, and then a home ent mix (Atmos nowadays for A-titles), and maybe deliverables for 5.1 and 2.0, but if you'd asked me if there is an entirely separate mix for disc vs streaming, or disc vs broadcast, I would've said no. Separate *encodes*, yes (Dolby TrueHD or DTS-HD Master Audio for disc vs Dolby Digital + for streaming), but an entirely separate mix?
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
My understanding of most workflows is that there is a theatrical mix, and then a home ent mix (Atmos nowadays for A-titles), and maybe deliverables for 5.1 and 2.0, but if you'd asked me if there is an entirely separate mix for disc vs streaming, or disc vs broadcast, I would've said no.


We start off with one master file. From that master file, you encode a master for disc and another one for streaming. These are not entirely new mixes, but encodes tweaked for each platform. A master file encoded for the disc could not be used for streaming, that is my point. A streaming master has much more compression and requires a completely different encoding setup than the disc. Perhaps "mix" was a poor choice of wording, and you are correct - encode is the proper word sort of. When I change anything from the original mix (and I do), it becomes a new mix. That comes before the encoding.
 
Last edited:
OP
A

audio2920

Active Member
Joined
May 21, 2021
Messages
233
Likes
288
I forgot to ask, when you're mixing to a "reference" HE playback system, what are your assumptions? for example, 5.1.4 or 7.1.4, LFE played full range because you're already adding LPF, etc. Understanding the mixer's assumptions will guide the HE enthusiast to be in alignment with the source material.

@Soundmixer has answered this, but just to second it; if there was such a thing it would be 7.1.4 with BM @ 80Hz as that's most common. Quite a few of my clients request this, and for those that don't, I might as well use it as a starting point anyway.
 
OP
A

audio2920

Active Member
Joined
May 21, 2021
Messages
233
Likes
288
Wow, this is fascinating - glad I stumbled on this thread. Thanks to our two pros for sharing their inside knowledge, and all for a lively discussion.

Glad to be of service; it also helps me clarify thoughts that otherwise rattle around in my head unchallenged!

This would have nothing to do with LFE though, right? This is just summation of LP'd content from the channels. However wouldn't Atmos include some inherent mitigation (pan law?) to reduce the disparity in the level of the full range channels if an object/channel was to be routed to multiuple channels based on how many speakers are defined in the decoder? Wouldn't that counteract this build-up?

Yes, it's primarily the LP'd content. But, if the LFE content were in phase with the LP'd content, you can see how the two would then sum more aggressively in BM than they do in room. In the initial cinema mix, we try not to duplicate too much content from the main channels in the LFE because we know this can cause issue down the road, but obviously there's always some sort of relationship phase-wise, and generally more positive than negative.

But no, there's no mitigation beyond the -3dB pan law. All the Atmos decoder does to render the "speaker" outputs from the objects is apply the age-old "Make the signal 3dB lower when sending to two speakers simultaneously" but it does so in 3D. The BM is after this panning, or "rendering", except in the case of Object Bass Management (earlier in this thread somewhere).

Is there always limiting happening in the HT side in the AVR? On the HT side I assume the BM happens after the signals exit the decoder, so it's really up to the implementation in the AVR (and the inherent limit of the speaker system I suppose) - wouldn't this be another stop betwen the 'chip' and the (even) wilder realm of room acoustics and calibration? I presume that you don't ride output limiters while mixing, so if a HT system had enough sub capacity, they should not experience this effect, right?

In Atmos, Dolby have built the limiters & BM built in on their side so it's not manufacturer specific. Like I say, knowledge of exact implementation in AVRs isn't my thing, but it occurs to me that even though the BM is part of the Atmos ecosystem, an AVR manufacturer could set it to "OFF" and implement BM post Dolby. I'm not sure if Dolby's "built-in" limiters can be bypassed - I suspect not.

We normally mix with the Dolby spec limiters engaged in our monitoring chain, so we should be hearing the same result as in the consumer unit (for a given speaker config). This threw me to begin with because like, what!? "We're putting compression between our mixing desk output and our monitors!? You're kidding?" But since the implementation is the same at the consumer end, it makes some sense.
 

Spocko

Major Contributor
Forum Donor
Joined
Sep 27, 2019
Messages
1,621
Likes
2,999
Location
Southern California
@Soundmixer has answered this, but just to second it; if there was such a thing it would be 7.1.4 with BM @ 80Hz as that's most common. Quite a few of my clients request this, and for those that don't, I might as well use it as a starting point anyway.
So this makes sense why some HE systems would experience muddled mid-bass between 80Hz and 160Hz where you have no BM and summing of 7 bed level speakers all set to 80Hz cut-off; and depending on room size/dimensions, these peaks could be quite significant at the MLP
 

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
So this makes sense why some HE systems would experience muddled mid-bass between 80Hz and 160Hz where you have no BM and summing of 7 bed level speakers all set to 80Hz cut-off; and depending on room size/dimensions, these peaks could be quite significant at the MLP
Sounds like we need an active solution, like the AVR mic is always on near your seat and listening for that. Assuming adjustments could be made without too much delay. I should pitch this to the AVR makers.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
So this makes sense why some HE systems would experience muddled mid-bass between 80Hz and 160Hz where you have no BM and summing of 7 bed level speakers all set to 80Hz cut-off; and depending on room size/dimensions, these peaks could be quite significant at the MLP

There is a missing component here that we have not mentioned. It is not just a crossover point of 80hz, it is also the steepness of the filters themselves. When 80hz was recommended as a BM crossover point, it was for a very specific speaker system with very specific crossover slopes in their BM software. That was the THX speaker system. THX required complimentary 24db per octave high and low pass filters in their BM. They also had required their main speakers to be -3 at 80hz as well, so that 80hz has some requirements. In PM3 studios (or any studio that does HT mixes), this is also the requirement for the speaker system.

The issue with excessive bass between 80-160hz should not exist if the manufacturers followed that advice in their BM software, even without the speaker component being there. By 160hz, the output would be 24db down from 80hz, and if this is not happening, then perhaps said manufactures AVR or processor is using a more shallow roll-off in their BM system. Some may be using 12 or 18db per octave rather than the 24db per octave.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
Although I wasn't going to admit it, I'm one of those who runs without BM for my living room :oops: I know, I know..... I do have a sub though. It just sounds nicer in that room, and I have less consistency issues with BM bypassed, even if it isn't as good scientifically.

I wouldn't quite say it's not as good scientifically. To smooth bass out you want multiple sources, and the typical BM -> one sub setup actually does the opposite. I would expect in a typical mix the LCR low bass is somewhat spread out among all 3 channels. But this is exclusively a 1 sub problem. If you have 2 or more it's going to work better due to optimal positioning and longer distance separation between subs. Assuming proper setup. Which raises another issue.

My biggest complaint about bass management is that the software solutions in most popular AVRs are VERY poor compared to the state of the art. Audyssey is one of the better ones and you still often have to screw around with its delays just to get proper phase matching at the crossover. That is ridiculous. If you want 2 or more subs, you need something better, hence the frequent use of miniDSPs. But miniDSPs require a lot of manual work.

It's been accepted for a couple of decades that the BM problem is mathematical optimization that should be done with software. Even Geddes wrote a software app to do it, and he's a PhD in acoustics with decades of experience tuning HT setups! Yet we still don't have even 1 reasonably priced, easy to use software+hardware package for it. It's pretty sad. The closest thing is Dirac's bass module, which is excellent, but only available in hardware starting at $4000. Far beyond the reach of most HT setups.

So I'd say state of the art bass management is scientifically very good, but the reality most people end up stuck with in their living rooms is not so good.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
I wouldn't quite say it's not as good scientifically. To smooth bass out you want multiple sources,

Not just multiple sources, but multiple sources playing the same signals.

My biggest complaint about bass management is that the software solutions in most popular AVRs are VERY poor compared to the state of the art. Audyssey is one of the better ones and you still often have to screw around with its delays just to get proper phase matching at the crossover.

Audyssey uses the BM within the AVR. It is not BM software in and of itself. I don't think the BM software in AVR's is necessarily poor, but infinished and certainly not standardized. You don't know what crossover slopes it uses unless you have a Denon or Marantz which use complimentary 24db per octave slopes in their BM program. The Storm processor has excellent bass management and so do the Altitude processors. Their BM software is far more comprehensive than the typical AVR or processor.


I would expect in a typical mix the LCR low bass is somewhat spread out among all 3 channels.

You will find it less in the center channel than in the Left/right channels. Back in the day, the use of the LFE was pretty sparse, and the bass in the mix was located in the main channels for the most part. Over the years, I am seeing less and less bass in the mains, and more of the bass of the mix in the LFE.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
Audyssey uses the BM within the AVR. It is not BM software in and of itself. I don't think the BM software in AVR's is necessarily poor, but infinished and certainly not standardized. You don't know what crossover slopes it uses unless you have a Denon or Marantz which use complimentary 24db per octave slopes in their BM program. The Storm processor has excellent bass management and so do the Altitude processors. Their BM software is far more comprehensive than the typical AVR or processor.

Sure, I should have said the Denon/Marantz implementation with Audyssey is one of the better ones, not Audyssey by itself. But this is really just an illustration of why it's still a bad solution. The room correction software SHOULD handle the bass management implementation. You can't separate the BM solution from room correction, they're both solving the same problem. That is why Dirac's bass module takes over the crossover from the AVR and handles it itself. The room correction software needs to have full control over the crossover frequency, slope, phase matching, and EQ of every subwoofer, and it should be able to set them all independently according to what produces the best result when it simulates thousands of options. I agree Storm/Trinnov do a great job but their pricing is so ridiculous that they have no real impact on 99.99% of the market.

Not just multiple sources, but multiple sources playing the same signals.

Yes, so the best solution would be to use not just the subwoofers but every speaker capable of playing low bass to play the same summed signal. Nothing I know does that yet, sadly, it was a theoretical goal for Dirac but not sure if they're still working on it.
 

Spocko

Major Contributor
Forum Donor
Joined
Sep 27, 2019
Messages
1,621
Likes
2,999
Location
Southern California
There is a missing component here that we have not mentioned. It is not just a crossover point of 80hz, it is also the steepness of the filters themselves. When 80hz was recommended as a BM crossover point, it was for a very specific speaker system with very specific crossover slopes in their BM software. That was the THX speaker system. THX required complimentary 24db per octave high and low pass filters in their BM. They also had required their main speakers to be -3 at 80hz as well, so that 80hz has some requirements. In PM3 studios (or any studio that does HT mixes), this is also the requirement for the speaker system.

The issue with excessive bass between 80-160hz should not exist if the manufacturers followed that advice in their BM software, even without the speaker component being there. By 160hz, the output would be 24db down from 80hz, and if this is not happening, then perhaps said manufactures AVR or processor is using a more shallow roll-off in their BM system. Some may be using 12 or 18db per octave rather than the 24db per octave.
This sort of detail from the source/content level is exactly what's missing! You have discussions on forums diving deep into the woods about how steep the crossover slopes should be and the obvious answer is "well, just match it to the content's slope" - and as you say, if the film industry practice is 24db per octave assuming -3 at 80Hz, this pretty much is how AVRs should set their default slopes to be (assuming they already are LOL) and then consumers never have to touch it unless told otherwise. Other than Bose satellites, I imagine that most modern LCR speakers are at least -3 @ 80Hz so this shouldn't be an issue or maybe I'm just being naïve LOL
 

Spocko

Major Contributor
Forum Donor
Joined
Sep 27, 2019
Messages
1,621
Likes
2,999
Location
Southern California
Yes, so the best solution would be to use not just the subwoofers but every speaker capable of playing low bass to play the same summed signal. Nothing I know does that yet, sadly, it was a theoretical goal for Dirac but not sure if they're still working on it.
And enthusiasts wonder why consumers have turned to soundbars en masse LOL
 

krabapple

Major Contributor
Forum Donor
Joined
Apr 15, 2016
Messages
3,168
Likes
3,712
I wouldn't quite say it's not as good scientifically. To smooth bass out you want multiple sources, and the typical BM -> one sub setup actually does the opposite.

That depends on if you mean 'smooth bass' across multiple listening positions, or just one main listening position. From what I understand it can be achieved for a single position -- which is what a lot of us here care most about -- with careful placement of 1 sub. But more easily achieved with 1+ subs.
 

Soundmixer

Senior Member
Joined
Mar 8, 2021
Messages
433
Likes
296
The room correction software SHOULD handle the bass management implementation.

Then you have another issue if the end-user doesn't use the room correction. I have read some comments on other forums where people did not like what Audyssey was doing and just turned it off.


You can't separate the BM solution from room correction,

Actually, you can. There was a time when automatic room correction didn't exist, and we were able to manually equalize the speaker system. I learned to do this from Bob Hodas who calibrated almost all of the dubbing stages and editing rooms in Hollywood.


The room correction software needs to have full control over the crossover frequency, slope, phase matching, and EQ of every subwoofer, and it should be able to set them all independently according to what produces the best result when it simulates thousands of options. I agree Storm/Trinnov do a great job but their pricing is so ridiculous that they have no real impact on 99.99% of the market.

Here is the rub. It cost big bucks to pull what you state here together, and that is why Storm/Trinnov costs so much. You need some powerful processing to do all of this, and that is why Storm/Trinnov processors are built off of computers rather than just SOC.
 

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
Yes, so the best solution would be to use not just the subwoofers but every speaker capable of playing low bass to play the same summed signal. Nothing I know does that yet, sadly, it was a theoretical goal for Dirac but not sure if they're still working on it.
Why would I want my surrounds trying to help with the LFE?
 

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,765
Likes
3,703
Then you have another issue if the end-user doesn't use the room correction. I have read some comments on other forums where people did not like what Audyssey was doing and just turned it off.
Yeah, that happens when people are lazy and don't take time to set things up properly. I find the bass corrections are always amazing; it is the target curve for the treble region that doesn't always match someone's speakers and room. The best solution in that case is not to turn it off but to limit the correction frequency, which is an adjustable setting.
 

Trell

Major Contributor
Joined
May 13, 2021
Messages
2,752
Likes
3,285
Yeah, that happens when people are lazy and don't take time to set things up properly. I find the bass corrections are always amazing; it is the target curve for the treble region that doesn't always match someone's speakers and room. The best solution in that case is not to turn it off but to limit the correction frequency, which is an adjustable setting.

Lazy? That is arrogant of you.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,923
Likes
7,616
Location
Canada
Why would I want my surrounds trying to help with the LFE?

The relevant variable is whether a speaker has enough LF output, not which channel it is. If you have small surrounds then of course you wouldn't, if you have big ones there's no reason not to. The output of those speakers is currently unused in most cases. Of course it's a complicated problem to use subs and speakers of differing capabilities, I assume that's why Dirac has not released it.

Here is the rub. It cost big bucks to pull what you state here together, and that is why Storm/Trinnov costs so much. You need some powerful processing to do all of this, and that is why Storm/Trinnov processors are built off of computers rather than just SOC.

You don't actually need powerful onboard processing at all anymore, no. Dirac and Genelec both use the cloud for computing their filters. Once the filters for every channel are calculated, the hardware for implementing those filters is quite cheap.

The high cost is mostly due to software development. But this is a sales/marketing choice. If you don't sell many units, you need extremely high prices. Dirac seems to be waking up to the fact that they can eat everybody else's lunch by trickling their solutions down to low/mid range consumer AVRs. Because everyone else in the business including Denon/Audyssey has been stagnant for over a decade.
 
Last edited:
Top Bottom