1. Join Now

    AVForums.com uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Question about bass management

Discussion in 'AV Receivers & Amplifiers' started by NicolasB, Dec 19, 2002.

  1. NicolasB

    NicolasB
    Well-known Member

    Joined:
    Oct 3, 2002
    Messages:
    6,070
    Products Owned:
    1
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    Emily's Shop
    Ratings:
    +669
    A while ago I was reading a quite authoratitive-sounding article (I forget where) on bass management. It made a rather extraordinary claim.

    To clarify what I am talking about:

    A typical home cinema processor will be hooked (via an aplification stage) into 5 conventional speakers and a subwoofer. The actual sound track of the film consists of 5 full-range channels, plus a low frequency channel. In many systems some or all of the main 5 speakers can't go low enough to play back the whole of the surround channels, so the processor redirects low-frequency components from the main 5 channels. Anything below a certain frequency (typically 80Hz) will be filtered out of the signals fed to the main speakers, mixed in with the low frequency channel, and sent to the subwoofer.

    At least, that's what I would have expected. But according to this article the process is actually rather different. What it says is that when the processor is determining what to send to the subwoofer it mixes the LFE channel in with the full range channels first and then applies the 80Hz low-pass filter, before sending the result to the subwoofer channel.

    Why would this make a difference? Because the subwoofer channel doesn't necessarily top out at 80Hz. In theory there could be sounds in it as high as 120Hz. If bass management is done in this way then anything between 80Hz and 120Hz in the LFE channel will be thrown away. This is not too calamitous in practice (says the article), because although in theory there can be stuff above 80Hz in the LFE channel, in practice there tends not to be. But where this would really bite is if you set the threshold frequency lower than 80Hz. Let's say your main 5 speakers are all floorstanders which can get down to 40Hz in reasonable comfort, so you set the bass management to only re-direct frequencies below 40Hz. That would mean anything above 40Hz in the subwoofer channel would be thrown away!

    So, this raises two questions:

    1) Is all this actually true, or is it all complete and utter bullsh*t?

    2) If it is true, what can be done?

    One possibility might be to tell the processor that your two main front speakers are "large" (i.e. full range) that the others are small (i.e. need to have bass redirected) and that there is no subwoofer present (i.e. all bass including the LFE channel should be sent to the front left & right channels instead). You could then hook the front L&R channels into something like Paradigm's X30 controller before feeding the signal into the amp, and the X30 would redirect all the bass frequencies from the fronts to the sub. The problem with this, of course, is that effectively having two bass cross-overs in the system could cause problems. (Maybe if the processor cross-over was set to 60 Hz and the X30 cross-over to 80 Hz you could get away with it?)
     
  2. Lowrider

    Lowrider
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    I believe anything from receivers, but I bet the sort of processors you are looking at don´t work like that...

    The best way to connect speakers, at least most, and the 803 surely are included, is all small, and xover around 50hz, knowlegeable people with much larger speakers than yours do this...

    Very few speakers will have energy down to 20hz, and most will gain clarity and dynamics not having to play under 50hz or so...
     
  3. NicolasB

    NicolasB
    Well-known Member

    Joined:
    Oct 3, 2002
    Messages:
    6,070
    Products Owned:
    1
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    Emily's Shop
    Ratings:
    +669
    I know. That's precisely why I am concerned about this issue. (If it is an issue).

    Of course we could all get a system based on seven of the original B&W Nautilus speakers - they're flat (to +/- 0.5 dB) down to 25Hz and only hit -6dB at 10Hz. Power them with four Linn Klimax Solo monoblocs each and this would only cost about £200,000. :)
     
  4. Lowrider

    Lowrider
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Wake up... :p
     
  5. alexs2

    alexs2
    Well-known Member

    Joined:
    Nov 13, 2002
    Messages:
    13,896
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Ratings:
    +1,674


    The only solution?.....turn on the system and use it....fiddle a bit with it.....now I'm going back to sleep ZZZZZZZzzzzzzzzz.......
     
  6. paiger

    paiger
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    I have no sub but set my floorstanders to small with the Xover at 60. I find if I set it lower or have them on large, the whole thing loses clarity. This may be a fault in my cheapo amp not being able to drive them properly. I would rather lose a little low end than lose clarity higher up.

    For SACD, the problem is worse because the Xover is fixed at 120Hz so there is no way I can set them to small. I end up with boomy bass due to my house having a suspended wood floor. Guess I need a seperate BM box for this.

    S
     
  7. Stuart Wright

    Stuart Wright
    AVForums Founder Staff Member

    Joined:
    Jan 24, 2000
    Messages:
    13,925
    Products Owned:
    6
    Products Wanted:
    2
    Trophy Points:
    167
    Location:
    Birmingham, UK
    Ratings:
    +9,660
    alexs2 - the phrase 'if you don't have anything useful to say - shut up' springs to mind.

    NicolasB - I hope it is bull****. What's the point of having a separate LFE channel if you're just going to mix it with any others?

    I'm not an expert on the design of the electronic routing of sound within AV amps but my gut instinct (and I have a large gut) is that it's crap.
    The way my Denon allows you to set up test tones would imply that with bass management set up to have small speakers and a sub (as 99% of all speaker systems *should* be - to start off with at least) all 6.1 channels are separate and there's none of this mixing rubbish going on.

    Where is the article?
     
  8. alexs2

    alexs2
    Well-known Member

    Joined:
    Nov 13, 2002
    Messages:
    13,896
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    136
    Ratings:
    +1,674
    Point noted,thanks Spectre....guess it just followed on from Lowrider's suggestion to wake up....in any case was only suggesting that listening and adjusting may be the best solution,regardless of the internal programming.

    Will attempt to provide only useful comments(which I usually try to do)...peace and Happy Xmas.
     
  9. NicolasB

    NicolasB
    Well-known Member

    Joined:
    Oct 3, 2002
    Messages:
    6,070
    Products Owned:
    1
    Products Wanted:
    0
    Trophy Points:
    136
    Location:
    Emily's Shop
    Ratings:
    +669
    Well, you'd have to do some mixing some time unless you have all of your main speakers set to large (i.e. you're not redirecting anything from the main 5 channels to the sub and you expect your main speakers to go down to 20Hz). But the sensible way to handle redirecting bass from the main speakers to the sub would be to take the main signals, apply a low pass filter, and then mix the LFE channel in afterwards rather than mixing first and then applying the filter.

    I can't remember. :) If I track it down I will post a link.
     
  10. The Nightfly

    The Nightfly
    Active Member

    Joined:
    Aug 31, 2002
    Messages:
    801
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    31
    Location:
    The New Frontier
    Ratings:
    +23
  11. EvilMudge

    EvilMudge
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    NicholasB,

    At first I agreed with you, the audibly correct way to manage the bass would be the one you suggest. However I then thought about the DSP power needed, and I'm guessing it's only within the realm of the stupidly expensive DSP equipped units out there (Theta or Krell would be two that might qualify). I'm not even sure the MC-12 has enough power to do all the bass management and then do the post processing expected of it.
     
  12. Charlie Whitehouse

    Charlie Whitehouse
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    This is a very interesting topic. At first, I thought this has to be bull***t, but I suppose it could be true in some cases, but definitely not all.

    I can only speak reasonably authoritatively about the Theta processors since I have owned one for a long time and am intimately familiar with all its setup options.

    In the case of Theta, I don't believe what is suggested in the article can be true at all. These processors do not have any parameter to set any single low-pass crossover frequency for the sub channel. I have tested the output from this channel and there is definitely no 120Hz brickwall filtering of any kind going on. Using bass management with shallow crossover slopes, there is content easily identifiable via this output at 500Hz+.

    Both Casanova and Casablanca processors allow not only the crossover frequencies to be set, but also the filter slopes (6/12/18/24dB per octave) as well as the crossover type (Phase Perfect/Linkwitz-Riley/Butterworth). In the case of the Butterworth type, the high-pass and low-pass filter slopes can be set independently. Also with any of the crossover types, the phase of the low-pass signal sent to the sub can be inverted. All of the above options can be independently selected for the bass management option of each of the main channel groups. All this flexibility is only going to work if the low-pass elements crossed-over from each of the other channels are mixed with the LFE AFTER they have been generated. Applying some notional overall low-pass filter and slope to a sum of LFE and all crossed-over channels' full range signals, as postulated in the article, wouldn't work at all. Hence I don't believe this article could possibly apply to the Theta processors.

    As for relative processing power of different processors, it is a little difficult to measure relative powers since they are always quoted in different measures (MHz, MIPS, MFLOPS). In any case, from my own experience in the past as an assembler programmer, by clever and efficient programming, it is always possible to get a quart or two from a pint pot. Unfortunately, the reverse is also true.

    Since the Lex MC12 is a more recent development, I am prepared to concede that it probably has more raw processing power in its 4 SHARC Analog Devices 32 bit DSP's (with the possibility to quadruple this with further card upgrades). The TAG has a 66MHz ADSP-21065L 32 bit SHARC, with the option to add another, I think. The Theta Casablanca II has 2 x Motorola 56362's which are 24 bit 100/120MHz devices. These deal with all the bass management and some of the 'matrix' surround decoding modes. It also has assorted other 56009, 56004 and Actel processors to deal with DTS and FIR filtering for its DACs etc. In any case, if the Thetas can do it, I see no reason why the MC12, for one, can't.

    As for the Krell processors, I wouldn't think that they do any more than many middle of the road amplifiers/receivers. I don't know the specifics of the new Showcase processor, but previous efforts from the HTS range have done little more than use off-the-shelf chip solutions which leave much to be desired. I certainly wouldn't be over-awed by them, and they in no way bear technical comparison to Meridian/Tag/Lexicon/Theta offerings.

    Just to satisfy my curiosity, since I haven't played around with many other products, for how many other amplifiers/receivers/processors out there do you actually set a single low-pass crossover frequency parameter for the sub channel as indicated in this article?
     
  13. EvilMudge

    EvilMudge
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Thanks Charlie,

    Krell was just a guess, Theta was an educated guess :)
    I'd also like to apologise to MC-12 owners, as this is a quality piece of kit, and as I recall has the ability to assign different Xover values to different pairs of speakers (read this on SMR's site I think.)

    And to echo your question - how many of you high end owners out there can set individual crossovers for different speakers? Adzman, can your beloved Meridian do this?

    There is another way of doing the processing which avoids most of the overhead of the Theta method. First the full range channels are decoded. Then the small channels are summed, filtered, then added to the LFE output. This is only slightly more time consuming than adding them all together at the start then filtering them (1 extra ADD operation.) It's possible that this is what goes on in the more powerful mid-range units.
     
  14. Lowrider

    Lowrider
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    The current TAG offer does have a single xover freq for all speakers, the LFE content stays unchanged, I guess...

    I don´t think it is that bad, just limits your choice of speakers a bit... but any quality speaker will go down to 50hz easy...
     
  15. cybersoga

    cybersoga
    Active Member

    Joined:
    Apr 7, 2002
    Messages:
    1,628
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    48
    Location:
    UK
    Ratings:
    +25
    I've got my Sony setup so my fronts (set to small) cross over at 50hz, the center at 50hz, the rears at 70hz. There is also LFE cut option which I left at the max setting of 120hz (std). The subwoofer's LPF is disabled. Can I assume that I am hearing the whole 1-120hz of LFE mixed with the bass from the fronts <50hz & rears <70hz coming out the sub?

    The Marantz I had previously didn't have any bass management options.
     
  16. Charlie Whitehouse

    Charlie Whitehouse
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    That's the whole problem. From this article, it would appear not. You are getting 1-120Hz of the LFE + Front + Rears from the sub output, effectively doubling up from 50-120Hz for the fronts and 70-120Hz from the rears! :(
     
  17. cybersoga

    cybersoga
    Active Member

    Joined:
    Apr 7, 2002
    Messages:
    1,628
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    48
    Location:
    UK
    Ratings:
    +25
    You sure this effects me? My Sony lets me choose the cross over frequency (HPF) independantly for each speaker, as well as the LFE cutoff frequency as a seperate option. Wouldn't the LFE cutoff option only affect the LFE? that's what it seems to suggest, but I suppose it could be misleading. Or are you saying that i'm getting the LFE above 50hz doubled up through the front speakers?
     
  18. Charlie Whitehouse

    Charlie Whitehouse
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    NO. The LFE will only come out the SUB output socket (unless of course you configure your system without a sub when it will come out front L&R normally).

    But the implication of this article is that with your LFE cutoff set to 120Hz, because the full range signal from front & surround speakers is added to LFE before applying a single crossover frequency of 120Hz, then everything from 50-120Hz in the front channels will be replayed by your front L&R AND sub, and for the surrounds, everything from 70-120Hz will be replayed by surrounds AND sub, effectively 'doubling up' at these frequencies.

    All this could be total bullsh*t, but given how much store has been set by this site's authoritative technical analysis of the Chroma Bug issue, we probably ought to take at least some notice. :(
     
  19. cybersoga

    cybersoga
    Active Member

    Joined:
    Apr 7, 2002
    Messages:
    1,628
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    48
    Location:
    UK
    Ratings:
    +25
    I understand now... not a lot we can do about it if it's true though is there?
     
  20. EvilMudge

    EvilMudge
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Thought about this some more, and I do start to see where the whole argument was going. Designing a 5 channel system is fairly easy. The speakers are all going to be in roughly the same place, so the DSP only needs to correct for distance from the listening position and relative volume.
    Add in a seperate channel, that has to produce sound which must arrive in phase with five other speakers and the problem increases dramatically.
    Where is your subwoofer? Mines sitting next to my FL speaker. It's phase is set (correctly I might add) so that it matches my centre - but not the rears or the front stereo pair.
    I set it this way because it's more important to me to have the most active channel getting the maximum use of the sub.
    I'm willing to bet that some people have theirs off to one side, whilst others sit the sub behind them. The only way to have the whole work out easily is to have the sub sit dead centre in the middle of the room, or go with WATT/Puppys!
     
  21. Charlie Whitehouse

    Charlie Whitehouse
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Quite apart from the phase differences caused by the different speaker distances, there are inevitable phase anomalies caused by the group delay in the filter process itself.

    To give you an idea of how the Theta 'Phase Perfect' crossover works, it applies a normal high-pass filter to the original signal to derive the signal to go to the 'small' speaker. Rather than apply a corresponding low-pass filter to the original to derive the sub output, the generated 'small' speaker signal is digitally subtracted from the original to derive the sub output. The advantage is that if both are summed you get an exact replica of the original signal.

    You still have to deal with the speaker distance time delay problem but assuming that you can set delays to compensate, this method theoretically gives the best sounding results. :)
     
  22. Lowrider

    Lowrider
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    By the way, the TAG AV192R + the dual sharc upgraded AV32R will have different xover AND 5 band parametric equalizer per channel... :rolleyes:
     
  23. cybersoga

    cybersoga
    Active Member

    Joined:
    Apr 7, 2002
    Messages:
    1,628
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    48
    Location:
    UK
    Ratings:
    +25
    this site has some sine wave test tones. If I play the 30hz tone, the sound only comes out of the sub. If I play the 100hz tone, the sound only comes out of the front speakers, so it appears that my reciever is working as expected.
     
  24. Charlie Whitehouse

    Charlie Whitehouse
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    An interesting result. There is more to this than meets the eye.

    Here is an extract from the Dolby Digital 5.1-Channel Production Guidelines where they are talking about setting up a monitoring environment that allows mixes to sound correct when replayed in the consumer environment...

    "Section 3.3 Bass Management

    ..... For consumer applications such as DVD and Digital TV, the consumer decoders add a slight twist to the equation. Consumer decoders take the LFE signal and add any channels in need of bass management, as determined either by product design or user selection. The five main channels are then high-pass filtered at either a fixed frequency of 80 Hz or a selectable frequency of 80, 100, or 120 Hz. The summation of the LFE and any other channels is low-pass filtered at the same frequency. If the crossover frequency is fixed at 80 Hz, as is standard in lower priced decoders, information in the LFE channel between 80 Hz and 120 Hz will be reproduced at a lower level than it is recorded at. To replicate what the consumer will hear, a third order 80 Hz filter in the LFE audio signal path to the recorder is recommended....."

    Which supports what the article says. However, what is not clear is how the single low-pass frequency for the sub output is being determined. Dolby's statement above implies a single low-pass frequency also. Is it being automatically selected by the processor/receiver based on whatever the lowest frequency selected for any of the crossed-over channels is? Or is it stated explicitly in the configuration? Either of which would still explain your results Cybersoga - one thing you could try is to set the crossovers for the main L&R as high as possible, and the centre as low as possible then play a sine wave somewhere between the two - if the lowest selected crossover (centre) determines the low-pass frequency for the sub, the test tone should be reproduced at very low level only because the sub is being prevented from going that high. I don't really expect any different results though.

    Or maybe the article does not apply to any processors where different crossover frequencies can be selected for each channel? :confused:
     
  25. cybersoga

    cybersoga
    Active Member

    Joined:
    Apr 7, 2002
    Messages:
    1,628
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    48
    Location:
    UK
    Ratings:
    +25
    I think this might be the case. On my new Sony, I can set the cross over frequency for the front, center, rear, and there is a seperate setting for LFE low pass filter.
    The Marantz SR5200 I previously had didn't have any of these options, not even a single cross over setting, so I supposed it's fixed at 80hz like the article suggests. It seems to me that this would be applicable to recievers that either don't have any sub crossover options or only have one setting that affects both LFE low pass filter and speaker/sub crossover(s). I may get the Marantz (waiting to be sold) back out of the box and try the same test and see what happens.

    Maybe somone else could give it a try and mention what reciever you have?
     
  26. Lowrider

    Lowrider
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    I got this answer from TAG, all I have to do now is wait for that upgrade... :blush:

    Hi Antonio,

    AV32R currently applies the crossover frequency selected in the subwoofer menu system to the LFE channel, as the Dolby notes suggest. AV192R has a much more flexible bass management system that allows the LFE low pass filter to be set independently of the bass management crossover frequencies for the various groups of loudspeakers (which are also independent of one another and selectable in 10Hz steps between 40 and 150Hz). This will become available for AV32R as part of the Dual Processor upgrade.

    THX recommend an 80Hz low pass filter be applied to the LFE channel, despite the notional 120Hz allowed for this channel - this is mainly to prevent localisation of the subwoofer. Disc authors generally avoid putting content above 80Hz in the LFE channel for the same reason.
     
  27. cybersoga

    cybersoga
    Active Member

    Joined:
    Apr 7, 2002
    Messages:
    1,628
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    48
    Location:
    UK
    Ratings:
    +25
    I found a tone generator very useful for testing, my pc is hooked up via SPDIF so should be pretty acurate. you can also make wav files, burn them to cd and test that way.
     

Share This Page

Loading...