Dialing in main/sub crossover point using measurements -- what is the best approach?

All of that is mostly a waste of time. Actual musical performance is the acid test. I was an electronics teacher, engineer, and technician for over 40 years, and have been involved in hundreds of tests of audio gear using far more sophisticated test equipment than you have available, and it still is not as good for the final adjustments of the sound as your EARS.

Testing is a quick way to detect gross anomalies, but is not helpful in fine-tuning the system sound quality.

Your main speakers are rated to be -3db at 42 Hz, which means they start to roll off at around 50 Hz.

You should set the fronts to do everything they can do, not limit their frequency response in any way. That will give you the best sound by far.

The subwoofer should be set to an upper frequency in the 50-60 Hz range using only its built-in high-pass filter. It should only be doing what the main speakers CANNOT do.

Get some live acoustic recordings that have good string bass and bass drums and LISTEN; your ears will tell you when the balance between the bass and other instruments sounds realistic and natural. Keep adjusting the subwoofer rolloff control until you get the best balance between the front speakers and the sub.

Your ears are the best test instrument you have; use them.

One of the best test records I have is the "JAZZ" CD by Ry Cooder. It has bass drums, cymbalom, electric organ, and tuba notes that make a great bass test possible. It also has an excellent natural recorded sound quality, which helps a lot.

In my experience you should disable all receiver frequency controls and use only the rolloff filter on the subwoofer to make the adjustment. The receiver filters are damaging to the sound quality as they create undesirable phase problems.
Edited by commsysman - 7/14/12 at 8:03am

Ignores the fact that with woofers, bass extension is often significantly greater than clean bass extension. IOW just because a speaker will put out useful output at some low frequency doesn't mean that it should be used that way.

Consider the distortion and frequency response shown in this real-world example:



Notice that the mid band SPL is about 96 dB, so we must bear in mind that it is common for this speaker to be called upon to work at 10 dB higher SPLs which can cause a dramatic increase in distortion, particularly at low frequencies.

The - 3dB point of the speaker's frequency response is 60 Hz. Its distortion is about 36 dB down or about 1.5%.

If we raise the crossover point by just a fraction of an octave to 80 Hz, distortion is now at least 10 dB less or about 0.5% This shows how a relatively small increase in crossover point can make a substantial decrease in distortion, and that running speakers right up to their low frequency limit can cause a signficant increase in nonlinear distoriton.

your "experience" with this (as well as with most everything you comment upon), is, umm, wrong...

@the op...

run sweeps with them together, varying the xover point... see which provides the best splice...

it can also be useful to experiment with sub distances, depending on the rc/eq software being used...

The graph I showed of the frequency response and distortion of a speaker in post 3 was made using standard speaker measurement techniques - with a condensor microphone and standard FFT gear.

So, yes the distortion can be measured, and is not too subtle to detect with software like REW.

This is a severely misinformed post, and it only demonstrates that the poster is totally unfamiliar with modern Bass Management. The OP's measurement techniques are far better than the outdated techniques described above.

In addition to the steps the OP describes:

I would also suggest running the sweeps while adjusting the subwoofer Distance setting in the receiver. The subwoofer Distance can adjust for timing disparities and improve, (or hurt), the response around the crossover, whatever that crossover might be.

Craig

The THX crossover is only applicable for THX certified speakers and subs. The THX crossover is more than just the 80 Hz frequency; it also has specified slopes for each side of the crossover that match up with the specified rolloffs of THX speakers and subs. If your speakers and subs don't have those rolloffs, the THX crossover will be no better than any other potential crossover.

That said, it's likely, IMO, that an 80 Hz crossover will be optimal with your speakers. Your speakers are ported with a port tune in the mid-40's. I prefer to cross ported speakers about an octave above the port tune. An 80 Hz crossover fits that requirement. Optimize the response around the crossover with the subwoofer Distance setting, and that "should" be your best bet.

Craig

One of the more amazing things I've ever seen posted on AVS.

Care to clarify that?

I think we all know that THX is more about branding than technology.

I did a little research on what constitutes a THX crossover:

80Hz subwoofer -24dB/octave LPF
80Hz front LCR -12dB/octave HPF
125Hz surrounds -12dB/octave HPF

The high pass filters are completely generic in terms of slopes. Since many receivers allow setting the subwoofer crossover and the other crossover points independently over a range from at least as low as 80 Hz and at least as high as 120 Hz, there's nothing unique about the THX crossover points.

The only thing that is even a little different is the 24 dB slope on the sub LPF. It doesn't take a special sub to work with a higher slope filter. In fact, the major benefit of a high slope at this point is to give a break to subs that may not sound good st or above their crossover points. The high slope helps avoid stimulating them at frequencies where they aren't at their best.

I've just lately debunked the idea that we want to match the slopes of speakers and crossovers. One of the benefits of crossovers is that we use them to avoid driving speakers at frequencies that would otherwise be within their nominal frequency response range, but are not optimally used in that range.

For example, just about every 2-way speaker has a crossover in the 2-3 Khz range, but use a tweeter that goes down to more like 1 KHz, and a woofer that probably has some kind of response up to 5 KHz or beyond. Crossovers are easier to make work well if the speaker they are used with is clean at the crossover frequency and several octaves beyond.

In a THX system the high pass filter is 12db/octave but the main speakers are also spec'd to roll off at 12db/octave @ 80Hz. Room acoustics aside that forms a 24db/octave roll off which is symmetrical to the slope of the subwoofer's crossover. I think that is what CJ means by the THX crossover being only applicable with THX speakers. Or at least only applicable as gospel.

So the 2nd harmonic measurement is the sound produced by distortion? I'm not that familiar with measuring distortion but it seems to me that the 2nd harmonic could also be caused by reflections in the room and not only distortion of the speaker. Distortion causes noise at frequencies different from the frequency your speaker is intending to produce, so wouldn't distortion be the total noise level from all frequencies besides the fundamental frequency at each point? This would need to be done in an anechoic chamber to get rid of noise from reflections. I don't understand how a speaker would be expected to produce the exact 2nd harmonic of the fundamental frequency anytime it distorts; if it does, can you explain why?

Also, you call a 1% decrease in distortion a "substantial decrease", I don't know how you consider that substantial, I would consider it completely negligible. Assuming the measurements you are using are correct, even if someone is listening to that speaker at those levels, they would have an incredibly hard time hearing the distortion at 60 Hz at all; trying to hear the difference in amount of distortion from 60Hz to 80Hz would be impossible because they are going to be hearing the fundamental sound over the sound from distortion.
Edited by Kach22 - 7/17/12 at 11:54am

You're in good hands here. However, I thought I'd add some interesting and pertinant quotes from Dr. Geddes;


"I have always thought that if someone’s measurements do not “tell the whole story” then they are the wrong measurements. Technology has simply come too far to believe that “there are things that we cannot measure.”


" I have also never believed that all that matters is “how it sounds,” because this is such an unstable and personal opinion. Sound quality opinions can and will differ from person to person, system to system and most importantly even within the same person on different days (as I said before, I have personally witnessed this in well regarded “reviewers”). Personal preferences have such a low stability as to be an almost completely pointless thing to stake a claim to. “Hi-Fi” does not mean “pleasant” -- it means “accurate”; accuracy, as opposed to preference, is absolutely quantifiable and extremely stable – as stable as I care to control in my lab from day to day or test to test (but in any case its uncertainty is easy to quantify and understand). Decisions based on accuracy are therefore much more likely to be valid than decisions based on “how it sounds.” I do not see how one could ever support a position that “preference” trumps “accuracy.” That’s simply taking a giant step backwards in the evolution of Hi-Fi".

Dr Earl Geddes



Best of luck
.

^^^

sadly, many have not evolved along with it...

How does Geddes define "accuracy"?

http://www.gedlee.com/downloads/The%20Perception%20of%20Distortion.pdf

Thanx. I'm always curious how folks use that word (input signal, how it sounded during mixing, how the original event sounded, etc).

I did: What's really amazing is that this needs to be explained to you.
What you missed in your "research" is that speakers and subs designed to THX specifications are designed to work in conjunction with the slopes of a THX crossover:

A speaker designed to THX specification has a -3 dB point of 80 Hz and a acoustic roll off of 12 dB/octave. When COMBINED with the 12 dB/octave roll off of the crossover, you get a COMBINED 4th order, 24 dB/octave roll off.

A sub designed to THX specification is flat to 160 Hz, or 1 octave above the crossover. Apply the THX crossover to that sub and you get another 4th order, 24 dB/octave roll off

Combine the 4th order speaker roll off with the 4th order subwoofer roll off and you get the perfect 4th order/4th order Linkwitz-Riley crossover. Dr. Linkwitz has shown this crossover to be be essentially free of any phase anomalies: http://www.linkwitzlab.com/crossovers.htm (#6)

My point to the OP was that, if you're not using speakers and subs designed to the THX specifications, then the THX crossover is just another generic crossover... as you've recently discovered in your "research".
I'm sure THX will be calling you very soon for your advice.

Craig
Edited by craig john - 7/17/12 at 7:28pm

Which makes absolutely no sense when talking about loudspeakers (like in this thread), since the output isn't electrical, like the input is.

cheers,

AJ

It makes plenty of sense if your response shows some evidence of trying to understand it in context, and doesn't immediately jump to the conclusion that the writer is a total idiot.

Obviously, the "accurate facsimile" "at the listener" is acoustical. What else can it be?

I didn't miss it, it was right there in the OP. I referenced that very fact in my reply.

However, there needs to be little understanding of the technical point that I explained, but it was sadly hastily dismissed with apparent bluster and defensiveness.

Just because a certain sub is designed to work with a certain crossover, doesn't mean that the same sub can't work with some other crossover. Just because a certain crossover was designed for a certain sub, doesn't mean that the same crossover can't work with some other sub. I was trying to address that misapprehension with a little tutorial.

I tried to explain this point technically, and it appears the technical explanation was was ignored.

If you buy into the myth that THX is some hyper-advanced technology from outer space, that THX makes no sense in a general technical context, then of course THX subs can't work with anything but THX crossovers. I have a higher opinion of these THX components than that! I think that they are well-designed and that their technology has more general application.

You left out the part where this approach gets you half of a Linkwitz-Riley 24 dB/octave crossover, only partially implemented acoustically. I built speakers like this maybe 35 years ago when I had to design and build the crossover from scratch. I did it. It worked. It was and is a good idea.

Same answer as above and I was there and did that over 35 years ago.

BTW many subwoofers are reasonably flat to 160 Hz and above, not just those with the THX label on them. So right up front there was an obfuscated truth - an apparent claim that THX subwoofers are the only subwoofers that are reasonably flat to 160 Hz and above. The truth is that there are many subwoofers that are reasonably flat to 160 Hz and above. Sounds like a sales pitch for THX subwoofers.

The original JAES article that we read and followed over 30 years ago was: S. H. Linkwitz "Active Crossover Networks for Non-coincident Drivers," J. Audio Eng. Soc., vol. 24, pp. 2-8 (Jan/Feb 1976) You can read a good explanation of it here:

http://www.rane.com/note160.html

Actually, you seem to misunderstand your reference. The Linkwitz-Riley crossover is no way free of phase anomalies. It has a built-in gigantic phase anomaly (360 degree phase shift jump at the crossover point) which Linkwitz himself aptly calls "all pass". An all pass crossover trades off a gigantic phase anomaly (and the group delay that goes with it) at the crossover point for what Linkwitz claims (and I agree) is something more worthwhile than ideal phase response: To quote your cited source (Linkwitz): "The best electrical crossover filter is one that maintains the acoustic polar response of a loudspeaker throughout the crossover frequency range as output shifts from one driver to the next."

My claim is that the response was misleading and obfuscatory, as opposed to being enlightening, an unfortunate trend that the current response continues.

Take the phrase "just another generic crossover" in the context that somehow that is a bad thing. Sounds to me like part of a sales pitch, not a tutorial. Unless they are totally weird, all crossovers are generic. The THX crossover is not weird, it is generic and I mean that in a good way. However, by calling it generic, I might be stripping it of some hoped-for sales appeal. Sorry! ;-)


As I have pointed out before, nobody needs to call me for advice when they can get it from primary and good secondary sources (both cited here) that are well known to many. Just because some people don't seem to know the trade-offs in the THX subwoofer crossover (and every other 24/dB octave Linkwitz Riley crossover) doesn't mean that everybody is working at their level.
Edited by arnyk - 7/18/12 at 2:53am

Dismissive, self-aggrandizing tone noted. Effective procedures are often determined by the experience and learning level of the person implementing them. What works and is appropriate for a person with lots of experience can easily be a disaster for people who are starting out.

Agreed.

A truism that is unfortunately contingent on everything before the final adjustment working right.

As I will show, a procedure that depends on knowledge that the implementer does not have is doomed to failure. This is BTW something that a "electronics teacher, engineer, and technician for over 40 years" should know and appreciate.

Unless of course the 42 Hz spec is:

(1) Made out of whole cloth
(2) Depends on the room being a certain size
(3) Depends on room boundary conditions that may not apply.

All or some of the above may be true.


That has by now been effectively criticized, even debunked in many ways. So much for the "electronics teacher, engineer, and technician for over 40 years", part.

If you high pass the subwoofer at 50-60 Hz, why even have a subwoofer? Many subwoofers don't have high pass filters. Most active subs have low pass filters. Perchance you meant low pass filter?

Wrong. The subwoofer should be doing what it can do better than the main speakers. Just because they can do it, doesn't mean that they can do it best.

Here's the most serious problem with the above advice. I know for sure that many people particularly young people have not heard a great deal of " string bass and (acoustic) bass drums" in a live context. Many that have, did not hear it in what you and I would call a good live context, but rather in a crowded club or too-small or too-large venue. These days, if they heard it at all, they may have only heard it through a sound system. If you haven't noticed, almost nobody listens to live jazz any more.

Much of what might be called jazz is performed with electronic bass.

Unfortunately there are two controls on most active subs that relate to balance - one is rolloff, the other is gain. Some have a third for phase and of course there is a time delay for the sub in the AVR.

Test equipment of any kind, whether it is our ears, or some piece of gear is only as good as its calibration.



By all accounts Jazz was recorded in a studio in 1978. Its bass is therefore a fabrication of technology and not inherently natural.


In fact our ears tolerate many phase aberrations, especially in the bass range. Speaker setup involves many trade-offs and avoiding filters in a receiver makes no sense. A 12 dB/octave filter does not know whether it is in a receiver or a spearate component, , whether it is implemented in the analog or digital domain. If it has the laws-of-physics ordained phase and amplitude response, and low enough nonlinear distortion, it is what it is and I mean that in a good way.

IME good advice to ignore on many grounds. 30 years ago, it might have made more sense.

Where did I say that the THX crossover was either a good thing or a "bad thing", or that any other crossovers were either good or bad? All I said was that the THX crossover, WHEN USED WITH NON-THX SPEAKERS, will be no better or worse than any other crossover; hence it's just another "generic crossover" WHEN USED WITH NON-THX SPEAKERS. That is all I'm saying.

For example, if you use the 80 Hz, 12 dB/octave THX crossover and slope with a ported speaker with a -3 dB point of 40 Hz, you'll be down 12 dB at 40 Hz and 36 dB at 20 Hz, (12 dB from the electrical crossover plus 24 dB acoustical roll off from the speaker.) Can that "work"? Sure it can. But it won't work the same way a THX speaker will work with the same crossover. If you want the THX crossover to work the way it's intended to work, you need to use it with speakers designed to THX specifications. If you don't care about that, the THX crossover can certainly be used with other speakers. So can a 60 Hz or an 120 Hz crossover with different slopes than the THX crossover. However, unless you know those slopes, and the inherent roll offs of the speakers and subs, it's a crap shoot as to whether they'll blend ideally or not.

When using non-THX speakers I don't *necessarily* recommend an 80 Hz crossover, nor do I necessarily recommend against it. I recommend it IF that is the crossover that works best with the specific speakers and subs to which the user is applying the crossover. The chosen crossover needs to be optimized for a given set of criteria: speaker roll-offs, subwoofer extension, (on the upper frequency side), placement of the subs and speakers in the room, (which will impact the in-room roll-offs), and localization of the subwoofer. If you have speakers that roll-off at 150 Hz, and a sub that extends to 160 Hz, I would suggest placing the sub in the front of the room to minimize localization issues, and then cross the system at 150 Hz. If you have speakers that extend to 28 Hz in-room, with high output capability all the way to 28 Hz, Then a 40 Hz crossover could be the optimal choice, or something higher could be optimal. It all depends on the *system*! The THX crossover in a non-THX *system* may be the right choice or the wrong choice, depending on he components of the *system.*

Having said all of the above, it should be noted that many THX certified speaker don't actually hit the THX -3 dB point, or the intended roll-off. For example, the Klipsch THX U2 speakers are ported with a - 3 dB point of 48 Hz. http://www.klipsch.com/kl-650-thx-bookshelf-speaker That' doesn't meet the spec. The Atlantic Tecnology 8200e's are sealed, (12 dB/octave roll-off), but their - 3 dB point is spec'd at 60 Hz, not 80. That's closer, but still doesn't quite hit the mark. Crystal Acoustics THX U2 speakers are ported with a -35 dB point of 35 Hz, which is nowhere close. Some THX speakers do hit the mark; Snell and Jamo, for example, are sealed with an 80 Hz -3 dB point. Bottom line, even THX doesn't enforce their own spec all the time.

Craig
Edited by craig john - 7/18/12 at 10:50am

Your own ears are the best test equipment; the entire purpose of a sound system is for the listener to enjoy what is being played. You could use 1000 test intruments that are calibrated to perfection, but if the listener thinks it sounds terrible , it is terrible, regardless of what all the test instruments or any number of experts say.

OK, everything ever recorded is a "fabrication of technology", I don't understand what the purpose of your statement is besides attempting to belittle commsysman or trying to confuse people. Some recordings sound more natural than others, commsysman is saying that "JAZZ" sounds very natural. If you disagree with him, say that you don't think "JAZZ" sounds natural or something like "commsysman is an idiot" rather than saying something random about "fabrication of technology" and sounding like a crazy person.

I get the feeling that you, arnyk, are the kind of person that will just keep rambling on and repeating themselves when someone disagrees with you until the other persons stop talking and walks away; and then you think to yourself "I was right!". I could be wrong, but that's the way it seems reading these posts.

I requested feedback on the graph you provided and how/why you interpreted it the way you did, but you gave me no response and instead made lots of little comments on things not related to the discussion at hand. You are free to do as you please, but you can't expect anyone to take you seriously when you act like this.

EDIT: apparently you can't say "thinks it sounds like ****, it is ****".

In order to be called a "THX Crossover." the 80 Hz crossover, it must use the 12 dB/Octave High Pass Filter, (HPF), on the speakers and the 24 dB/octave Low Pass Filter, (LPF), on the sub. I don't know what the slopes are for the other crossover set points, but I expect they are the same. You could verify by measuring with REW.
A "filter" is either a High Pass Filter, (i.e.,it cuts the low frequencies and allows the highs to pass), or a Low Pass Filter, (i.e.,cuts the high frequencies and allows the lows to pass). These filters have a roll-ff that is classified by it's "order." A 1st order filter has a 6 dB/octave roll-off or slope. A 2nd order has a 12 dB/octave roll-off or slope. A 3rd order crossover has an 18 dB/octave roll-off; a 4th order crossver, a 24 dB/octave roll-off, etc.

A "crossover" is a combination of a HPF with a specific order and LPF with a specific order. None of them are "brick walls" that cut the frequencies completely off above or below their set points.

Craig

"Crossover" is just an audio term for one or more filters, they are the same thing. All filters and therefore crossovers have a slope, the slope is determined by the design of the filter and what order it is (a 2nd order filter is essentially like having two 1st order filters for the same frequency back-to-back and will have a steeper slope). The easiest way to create a 2-way speaker with a crossover of 2500Hz would be to directly run one wire from the speaker terminal to the woofer and run a second wire with a single high-pass filter set at 2500Hz in-line to the tweeter, you could add a low-pass filter at 2500Hz in-line to the woofer, but it is not required for the speaker to be considered a 2-way speaker with a crossover at 2500Hz.

For subwoofers with the adjustable crossover on the back it is suggested to turn the crossover to the max when connected to an AVR with an internal crossover setting for the subwoofer; if you set them both to the same frequency you would get a slope twice as steep. If you set them to different frequencies you would get a slope that gets steeper at some point and it would continue at the steeper slope the rest of the way down ( or I guess you could say "up" the frequency spectrum, since it is a low-pass filter, attenuating higher frequencies than the setpoint).

Unless otherwise stated your receiver very likely has a 12dB/octave crossover, that seems to be the standard.

There is no necessary connection between those two statements. Juxtapositioning them as above makes it seem like they are eternally joined at the hip.

That the entire purpose of a sound system is for the listener to enjoy is a truism. But that does not detract from the usefulness of relevant test equipment applied properly.

What you probably meant to say is that for the purpose of determining listening pleasure, the ears of a certain person comprise the best test equipment.

Unfortunately, the test equipment you place above all else lacks sensitivity, can be very time consuming to use, and is unstable - it does not give consistent results under the identical same circumstances.

So you want to put the fact that the listener might have a cold or be in a bad mood above everything else?

But some recordings are far more that way than others. A live recording made using minimal micing tends to be a more consistent and stable thing. If nothing else the recording staff has a ready reference - what the live music sounds like, to use to gauge and adjust their work.

Studio recordings are the exact opposite. They are anything that the artists and production staff want them to be. Often the various musical voices are never together in one place until the final mix. There is no reliable reference.

Trust me I'm trying to do neither. But I am trying to correct or at least provide an alternative viewpoint to a number of questionable statements that have been presented as being the only relevant facts.

The only true statement that anybody can make like this is that the recording sounds natural to them. Elevating that to a global truth seems unnecessarily adventurous.

That no two different speakers work the identical same way is another one of those unhelpful truisms. Every speaker works a little different, even different THX speakers.

The above is an example of cherry picking the details of an example to make a point as if it were a useful generality.

You seem to be trying to make it out that if a speaker doesn't work the identical same way as a THX speaker that there is an cataclysmic catastrophe that can't be altered, adapted or managed in any reasonable way.

Of course. That is another one of those truisms. But is the THX way the only reasonable way? Is it the best way in every case? Is it the most economical way in every case?

I actually got you to say that without payola or threat of violence? ;-)

Its a crap shoot anyway, because of room acoustics. And who says that THX knows better what will blend best in a given situation, than someone else who happens to have full access to that situation?