AVIA Note: Subwoofer Calibration (1 Viewer)

[Jeffrey Forner]

Barry;

The 20 Hz signal is as low as you can go on Avia (unless there really is a hidden signal that goes deeper).

I think the idea of the frequency sweeps is to find any frequency gaps between the speakers and the subwoofer. Since true full-range speakers are rated down to 20 Hz, I doubt that there's any real reason for them to make the sweep go any deeper.

 

[Barry Barnes]

I think the idea of the frequency sweeps is to find any frequency gaps between the speakers and the subwoofer. Since true full-range speakers are rated down to 20 Hz, I doubt that there's any real reason for them to make the sweep go any deeper.

The main reason I was looking for them is to see how well the SVS and Hsu subs perform at below 20 Hz. It looks like the SVS starts to roll off at about 23 Hz but the Hsu is still staying pretty flat to that point and on down to 20 Hz which is the AVIA test disc limit...
It'd be nice to see if the SVS perked up a little at 20 Hz or if it continued to roll off at that point. I'd also like to see how low the Hsu can actually go. Looks like I will have to try and find some lower test signals elsewhere...
Suggestion for Guy: Seems to me that since many Movies and subwoofers are capable of frequencies lower than 20 Hz (the knock on the door in the Monster Inc. preview is 10 Hz for instance) that it might be worthwhile to include some lower frequency test tones... I also think it would be a nice addition to have some static test tones in the under 80 Hz region that you could stay on until you chose to move to the next increment, rather than just the sweep... Anyway, just a thought...

 

[Harold_C]

Why is it necessary to pre-compensated the LFE by -10 dB? Is it simply a result of all recievers base management boosting the signal by 10db or is it that most DVDs boost the LFE by 10db?

All Dolby Digital decoders boost the LFE channel on playback by 10 dB relative to any other channel. Thus, a recording engineer who wants equal output from the LFE channel and another channel has to record the LFE channel 10dB lower.

The 10dB change is handled in the bass management section of the Dolby decoder. Actually, with certain combinations of SMALL and LARGE speakers, the change in relative volume may be slighty different than 10 dB. It's done so that one LFE channel = the bass levels of the other channels combined.

 

[Harold_C]

Of what I can gather from many a fine post is that the x-over on the sub should be set to 40-50hz if possible. If I did this I'd have a gap between that 40-50hz and the 99hz. In this case I don't believe the "small" setting would work for my mains.

Somebody has really misled you.

100 Hz is a perfectly fine crossover point for a 5.1 channel system. The THX specified crossover is 80 Hz and there is virtually no difference between an 80 Hz and a 100 Hz crossover.

Setting the sub crossover to 50 Hz is too low for a 5.1 channel system. That's not removing sufficient bass energy from the main speakers/amplifiers to really get much benefit from bi-amping. With the characteristics of a Dolby Digital crossover, your low-pass filter has already rolled off the bass -6dB at 50 Hz. This means that the rolloff has started well below 50 Hz. The rolloff at the extreme bottom end of the subwoofer has usually begun by 35 Hz, so you are using the sub for very little information. Likewise, your main speakers will only be 3 dB down at 50 Hz from the crossover. That's not very far at all from just having them play full range -- in fact, on most bookshelf sized speakers, the speaker itself will have already rolled off before the crossover even kicks in.

With very few exceptions, you want to set all your speakers to SMALL and use the 80 or 100 Hz crossover on your receiver. Most of the expceptions would be in the other direction: satellites with 3.5 or 4 inch bass drivers should probably be crossed over higher than 100 Hz because they are going to struggle with bass output and excursion below that.

 

[Stephen Houdek]

Guy, or whomever,

I keeping seeing phase mentioned in this thread and it has me a bit concerned. I'm not a newbie but have been out of audio for a while so please bear with me as I'm still learning the the 5.1 world. BTW: I bought Avia and find it extremely useful for Audio and Video calibration.

Here's the phase concern......Is it NOT just as simple as ensuring you have proper polarity between your amp and all of your speakers? IE: +to+ and -to-??? Am I missing something? I checked, re-checked and am very careful when making these connections. I suppose it would be possible for the factory to miswire the internals but would hope that is far from a frequent occurance. Am I missing something with all of this concern over phase?

Thanks for your indulgence.

 

[Barry Barnes]

Is it NOT just as simple as ensuring you have proper polarity between your amp and all of your speakers?

For the most part yes. Unfortunately, "stuff happens" and wires do get crossed. Especially if they are "in wall" and you are counting on wall plates and the like. The AVIA disc will make it immediately obvious if one of your speakers is out of phase. They should all do essentially the "same thing" with the alternating tone. The language on the Disc is somewhat confusing as I found my speakers were easier to "localize" when they were out of phase, but hey it's just semantics. I knew what he meant and it was clear that all of them were wired the same.

Where phase really becomes more of an issue is when you add an external subwoofer. But again, with a sound level meter it is very easy to see which phase setting offers the most bass (besides the fact that it's a fairly audible difference as well). If the sub is out of phase from the mains you can easily lose 3 dB off the level.

 

[Harold_C]

Here's the phase concern......Is it NOT just as simple as ensuring you have proper polarity between your amp and all of your speakers? IE: +to+ and -to-??? Am I missing something? I checked, re-checked and am very careful when making these connections. I suppose it would be possible for the factory to miswire the internals but would hope that is far from a frequent occurance. Am I missing something with all of this concern over phase?

No. You are not really missing anything. You would probably be surprised at the number of people who don't check and recheck their speaker connections for proper polarity.

Due to the fact that subwoofers are sometimes placed some distance away from the main speakers, there are rare cases where the subwoofer might be electrically "in-phase" with the main speakers, but acoustically "out of phase". This will not impact the overall bass response (where the subwoofer is working alone), but will change the frequency response in the crossover range where both speakers are reproducing the same notes. Most of the "phase" tests on the Avia are aimed at identifying this (pretty rare) problem.

 

[Vince Maskeeper]

Setting the sub crossover to 50 Hz is too low for a 5.1 channel system. That's not removing sufficient bass energy from the main speakers/amplifiers to really get much benefit from bi-amping. With the characteristics of a Dolby Digital crossover, your low-pass filter has already rolled off the bass -6dB at 50 Hz. This means that the rolloff has started well below 50 Hz. The rolloff at the extreme bottom end of the subwoofer has usually begun by 35 Hz, so you are using the sub for very little information. Likewise, your main speakers will only be 3 dB down at 50 Hz from the crossover. That's not very far at all from just having them play full range -- in fact, on most bookshelf sized speakers, the speaker itself will have already rolled off before the crossover even kicks in.

In the case of using bookshelf speaker- maybe- but even most mid towers can handle a xover point of 60 without issue. If you review my past points, I'm one of the minority here who absolutely would never trust a full range/single amp speaker to handle full-on 40hz- but I think you're pushing it suggesting that 80 is perfect.

I've notice that you've often made ref to Dolby Spec'd decoder chips in your posts (in terms of ref tones, and now in terms of x-over slopes)- and I'm curious as to where you found this info. I'm certainly not questioning it- however your post alludes that the crossover slope and type are dictated by Dolby in all processors- and I have found the opposite to be true. Many processors offer different x-over points (obviously)- and it seems sonically obvious that different processors have featured different slope types and order factors.

Bottom line is, if you have a solid tower speaker, or even a mid tower- I would say that the box specs are lies (don't believe that they do 28hz)- but I would temper that by saying that I think claiming all x-over points should be 80 or higher is equally false.

-Vince

 

[BruceD]

100 Hz is a perfectly fine crossover point for a 5.1 channel system. The THX specified crossover is 80 Hz and there is virtually no difference between an 80 Hz and a 100 Hz crossover.

I agree with Vince on this one Harold, and totally disagree with your viewpoint.

Many early HT receivers took what I will call a shortcut and offered less than flexible bass management (like a 100Hz xover) that sounded awful with obvious localizable bass.

I'm a strong beliver in flexible bass management and prefer a 60Hz xover for my -3dB 31Hz main speakers. I think 80Hz is a good option, but no higher.

 

[Stephen Houdek]

They should all do essentially the "same thing" with the alternating tone. The language on the Disc is somewhat confusing as I found my speakers were easier to "localize" when they were out of phase, but hey it's just semantics.

Heh, and I thought I was the only one that found the opposite easier to find! Ran all around checking connections over and over thinking I was all messed up!

Thanks for the responses.

I'm curious if anyone here has discovered a speaker that was internally wired incorrectly?

 

[Harold_C]

Quote:

Most consumer grade Dolby processors are using the Cirrus Logic family of integrated Dolby/DTS chipsets and the associated post-processing digital audio chips. My information about the way these chipsets work comes from the enigineering application guides for these chipset products.
You can find many of the application guides on this webpage:
Cirrus Logic chipset info
The most useful document is found on the right side of the page: a PDF file called:
AN163 Outboard Decoder Systems: Application Code User's Guide for the CS4932X Family
This one covers most of the register settings for the various surround sound processors (Dolby Digital, DTS, etc.) and basic audio manager functions.
The other document that is useful is Appendix G:
AN163G Crystal Digital Post Processing User's Guide for the CS49330
I have a couple of more appendices that are not available on website covering some of the newer 6.1 and 7.1 channel features like Dolby Surround EX processing.
NOTE FOR THE FAINT OF HEART: These documents are guaranteed to make your head spin and will make absolutely no sense until you study them and study them and study them like a physics textbook. They deal with binary chipset register settings to activate and deactivate and configure various features. For example, the bass management information is contained in page after page of sample configurations, each containing 20+ register settings. I first started using these as reference material about a year ago and it's taken that long to really get comfortable with the way they are presented to the point where, now when I get a new one (like the Appendix on Dolby Surround EX), I can pick up on what they are saying pretty quickly.

 

[Harold_C]

Many early HT receivers took what I will call a shortcut and offered less than flexible bass management (like a 100Hz xover) that sounded awful with obvious localizable bass.

I can't say what was in the older receivers. However, I can say that the Cirrus Logic chipset that is used in the vast majority of units on the shelves today has a digital implementation of a standard THX crossover.

The lo-pass is a cascaded configuration of two second order filters resulting in an overall 4th order slope. The high-pass is a single second order filter. These are hard coded in the chipset.

The factory supplied crossover point is 80 Hz -- because these chipsets are used in THX approved processors and this is the THX required crossover. The way it works is that the Cirrus Logic chipset handles every available option: Dolby Digital, DTS, Surround EX, DTS Neo:6, DTS-ES, THX, etc. As a receiver manufacturer, you decide what options you want to pay the license fee for and Cirrus Logic provides the software to enable those features. Thus, it just makes sense to design the standard crossover to meet THX's requirments since some of the customers for the chipset will be paying the THX license fee and, therefore, MUST have that crossover.

The standard 80 Hz high-pass crossover is -3dB at 80 Hz with a 12 dB per octave rolloff, so it will be -15 dB at 40 hz. The low pass section is -6dB at 80 Hz with a 24 db per octave slope.

Now, what I don't know is what "name" each receiver uses for this crossover setting. I think it's obvious that the receiver manufacturers play fast and loose with their nomenclature (look at the various names they give to Dolby's dynamic range settings.). If you want me to guess, I bet that Pioneer's "100 Hz" crossover is really this standard configuration Cirrus Logic 80 Hz THX crossover, but I have no way of knowing that for sure without measuring it.

The crossover points are set by loading different "coefficients" in binary registers. It is not terribly difficult for the receiver designer to offer a choice of user selectable crossover points. In my opinion, a receiver manufacturer should offer at least a couple higher than 100Hz to deal with these increasingly common el cheapo speaker packages with 3.5 and 4 inch mid bass drivers. These speakers simply won't handle power down to 100 Hz and a little bass localization is preferable to a bunch of blown drivers.

I think 100 Hz to 120 Hz is a pretty good choice for 5 inch midbass drivers. 80 Hz to 100 Hz is good choice for 6 inch midbass drivers.

I agree with you that, purely from a speaker/biamping standpoint, many larger floorstanding speaker could work fine with a 60 Hz crossover. If this were the 1980s and we were talking about stereo systems, your recommendation of a 60 Hz crossover would get no argument from me (for very capable speaker/amplifier combos).

However, the way the crossover and bass management is implemented in Dolby Digital, you will be throwing away the upper portion of the LFE channel with a 60 Hz crossover unless you have a system configured with all of your speakers as LARGE and no low-pass or hi-pass filter in the circuit at all. It is for this reason that I state that the THX-spec 80 Hz crossover is the lowest setting that you should use in a Dolby Digital 5.1 system. In my opinion, there are really only two viable options for system configuration: 5 LARGE speakers and a sub or 5 SMALL speakers and a sub. I don't believe that the 5 LARGE speaker setup is a sensible choice for 99.9% of the systems and therefore the ALL SMALL configuration is the way to go. With an ALL SMALL system, you shouldn't use anything less than the 80 Hz THX crossover and, depending on the capability of your satellites, you may want to use something a little higher.

It may be theortetically possible, through register settings, for the receiver designer to alter the slopes of the crossovers. But, I think it would be highly unlikely that any receiver engineer, trying to get a product to market with this year's latest "whiz bang new feature" would bother with that.

 

[Roger Kint]

I think that's probably quite true -- when you are listening to the subwoofer by itself and when you are testing with pure bass tones and no harmonic overtones. Neither of these conditions exists very often in a home theater situation. As long as you do not have the subwoofer level set so high that it totally overpowers the upper harmonics, the harmonic struture played by the front/center/surround speakers will always provide the localization cues that convince your ears that's where the bass is coming from. That is the whole premise behind a mono subwoofer. In practice, on music or soundtrack materials, you won't start to hear any significant localization of the sub until you get up around 125 to 150 Hz.

I've studied physics of acoustics and can vouch for Harold's statement. He pretty much nailed it right on. ITD and IID localization cues are primarily taken from the sound from the main speakers to trick our ears into thinking that the sound is not coming from the subwoofer.

 

[Harold_C]

Roger,

The only time I've heard bass from my system that is NOT perceived as coming from one of the five main speakers is on something like the rumble of the tanks in the closing scenes of Private Ryan. Here, the producer has used essentially pure bass tones with little or no associated upper harmonic content. However, even in these scenes, you can't tell where the subwoofer is. Rather the bass simply shakes the entire house.

On any normal program material (such as music), my system sounds like bass is coming from the small speakers -- and that is with the sub in the back of the room and what is labeled a "100 hz" crossover.

My hunch is that, if someone is localizing the sound of their subwoofer, it's a sure sign the sub levels are set too high.

 

[Ellen]

Harold,

I don't think I'm pushing it at all. Apparently Tomlinson Holman, Lucasfilm, and THX share my opinion. The reaon that 80 Hz 4th order low-pass and 2nd order high-pass crossovers are so common in surround sound processors is because that is the crossover configuration that THX requires in every THX installation. They require that all main speakers be high-passed fitered with a 2nd order 80 Hz crossover. They require that all subs be lo-pass filtered at 80 Hz with 4th order slopes.

Just out of curiosity, do you know if the 2nd order high-pass filter is Bessel or Butterworth?

 

[Steve_Ma]

I suspect that the audible difference between 80 and 100 hz x-overs MAY BE way overstated when it comes to practical implimentation in a HT environment. My mains are clean down to about 47/48hz before they start rolling off, so I was anxious to try to coax more out of them for 2ch listening (actually 2.1). After a great deal of experimenting with connections and x-overs, I noticed very little audible difference between an 80 and 100hz x-over. In the end, I was able to measure a significant improvement in the "in room response" (flatness) using the fixed 100hz x-over on my Marantz AVR, but as far as an audible difference goes....lets just say it was dissapointing. :frowning:
Just my 2 cents.
--Steve

 

[Harold_C]

Just out of curiosity, do you know if the 2nd order high-pass filter is Bessel or Butterworth?

They are digital implementations of Butterworth filters as required by THX.

The hi-pass is a 2nd order Butterworth, -3dB at the crossover point, rolling off at 12 dB per octave.

The lo-pass is two cascaded Butterworths, resulting in a Linkwitz-Riley characteristic. -6 dB at the crossover point, rolling off at 24 dB per octave.

Not being a digital engineer, I don't know how closely you can approximate an analog Butterworth filter in the digital domain. There may be some slight variations, although my understanding is that you can come pretty close.

 

[Arnel]

This means that if you set your crossover point lower than 80 Hz, you are throwing away part of the LFE channel. Nothing will be reproducing the LFE signal between your crossover point and 80+ Hz.

Harold, I'm just a newbee on bass management and I'm trying to understand every post on this thread. May you please explain further, as quoted above, how would a processor's crossover affects the frequency spectrum assigned in the LFE considering it is a discreet channel. As I understand, the only thing that would affect its frequency reproduction is its own crossover at the subwoofer. Thanks in advance.

 

[Harold_C]

OK, I'll try. No guarantees, though!
First, lets forget about the decoder for a minute and look at what is actually recorded in the LFE channel. By definition in the Dolby Digital specification, the signal being recorded in the LFE channel cannot have ANY output above 120 Hz. In the mixing studio, the Dolby encoder has a brickwall filter that absolutely throws away anything above 120 Hz going into the encoder. Recording engineers don't want that kind of brickwall filter affecting the signal -- because it would sound like crap. So, in practice, they send the LFE signal through a THX style 24 dB per octave Linkwitz-Riley 80-Hz low-pass filter. This smoothly (and steeply rolls off the top end of the LFE signal so that it never "bangs" into that brickwall filter.
Now, lets turn to the playback end of things and look at two configurations.
First, lets look at a 5.1 channel system with five LARGE speakers and a powered sub. In the Cirrus Logic implementation, there is NO crossover in this configuration at all. Each of the five LARGE speakers handles the full-range signal recorded in its respective channel. The subwoofer handles the LFE channel and nothing else. There is NO low-pass crossover for the LFE channel -- if the LFE channel had a full-range signal, a full range signal would be output to the subwoofer. But, that doesn't happen, because it's illegal to record anything above 120 Hz in the LFE channel. So what is recorded in the LFE channel is already rolled off at 24 dB per octave at 80 Hz. There's no need for a crossover because the recorded signal is already filtered properly.
Now, let's turn to any configuration with all SMALL speakers. The high-pass (above 80 Hz) information is pretty straightforward. In any channel set to SMALL, the full-range signal recorded in that channel is sent through a high-pass filter (12dB per octave at 80 Hz) and sent to the main speaker.
However, the bass is handled completely differently than before. All of the channels (5 main channels plus the LFE channel) go into a "subwoofer summing module". Here their levels are adjusted (this is where the 10 dB LFE boost comes in). The ONLY way any bass signals can get to the subwoofer output is through the low-pass crossover.
That's fine if you use the standard 80 Hz (or higher) crossover point in the receiver. But, if you set your receiver to, say, a 60 Hz crossover, there's a problem. You have a signal recorded in the LFE channel up to 80 Hz, yet your receiver's crossover is chopping off the signal at 60 Hz. That means that any information recorded above 60 Hz in the LFE channel is simply thrown away. It has no path to get to the subwoofer without going through the receiver's lo-pass crossover. It's not like any LFE information above your crossover point is going to magically get routed to the main speakers -- there is no mechanism for that to happen. This means that you must set the receiver's crossover high enough to accomodate the entire LFE signal - 80 Hz or higher.
Think of it the same way as setting your receiver's subwoofer crossover at 80 Hz, but the crossover on the back of your subwoofer at 60 Hz. Something is going to get thrown away.
Just as a sidenote: after studying the block diagrams for the bass management system, I am really quite impressed with how well it is implemented. That's why I'm a stickler for just hooking things up in a standard configuration and letting the Dolby processor work it's magic. Why fight something that is designed to work pretty well?

 

[Arnel]

Harold, thanks, now I understand what are you trying to say but actually, I haven't heard that before. Is that true only for DD, how about for DTS?

Now, please consider the following setup and tell me if a low crossover (say 60hz) could still chop-off those freaquencies:

SETTING:

Main L&R Speakers = Large

Center and Surrounds = Small

SWFR = no

SUB-CONNECTION :

a) via speaker wire, Main speakers connected to the sub and using the crossover of the sub.

b) connected to the main out (L&R) via RCA of the receiver. This means that the sub will also be amplifying or duplicating the data from main L&F speakers.