Jump to content

Subwoofer Low Pass Cutoff Question - General Theory


RoboKlipsch

Recommended Posts

I've read probably a hundred reviews of subwoofers, and they always seem to make a point to talk about how high up they can play flat....i.e. this one is flat to 200hz.

 

When it comes to crossing over the mains, there has been a lot of discussion that by crossing them off higher than their lowest frequencies, they gain more headroom i.e. if the speaker could go down to 35hz, but you cross it at 80, it provides "headroom"...if I am using the terminology correctly.

 

When reading reviews of the subs, they make a point to talk about how this sub can go this low, it also can go up this high and remain flat.  My question is, do subwoofers also have the headroom issue like other speakers?  

 

For example, if your sub can go up to 200hz, but you cross it at 80hz or 100 or 120....are you gaining "headroom" with your sub at the lower frequencies, or does this not apply?

 

Always trying to learn.  If I asked the question incorrectly, please feel free to correct it.

 

Thank you,

RK

Link to comment
Share on other sites

I don't think crossover frequency has anything to do with headroom. The capability of your mains decide what frequency you should crossover at. If you mains are capable then 80hz should be the max crossover value. I use 80hz as my rf7iis supposedly have an impedance dip at around 70hz. 

Link to comment
Share on other sites

The headroom gained is in the amplifier. If it doesn't have to reproduce lower frequencies it uses less of its available power from its power supply and will have more available energy for the mid and high frequencies. But some amplifier has to make up for this, either the built in amp in your sub or a separate amp if your sub is passive(requires a outboard amp). All that said if you crossover your mains too high it may not sound good anyway.

Link to comment
Share on other sites

When it comes to crossing over the mains, there has been a lot of discussion that by crossing them off higher than their lowest frequencies, they gain more headroom i.e. if the speaker could go down to 35hz, but you cross it at 80, it provides "headroom"...if I am using the terminology correctly.

 

Your front left/right loudspeakers use direct radiating woofers, so one of the concerns is always "how much should the low frequency pass band be broken up into subwoofer (mono) vs. front left-right woofers (stereo) in order to reduce AM distortion and compression distortion.  If you don't play your system very loud, then you can basically cross them over at the lowest point available and retain stereo bass, moving the crossover frequencies from the more audible midbass to the less audible low bass frequencies. 

 

Also, the physical distance separation of the subwoofer from the woofers begins to become audible at some frequency.  Most people choose 80 Hz as the highest point.  I use 40 Hz for my setup since I've got horn-loaded bass bins and horn-loaded (TH) subwoofers directly behind the bass bins on each side, and this sounds better.

 

There are many room acoustics considerations that usually get wrapped up with any choosing of the crossover point, but in any case, I personally recommend not going above 80 Hz, and dealing with the room acoustics, i.e., loudspeaker positions relative to the room corners, subwoofer position(s) relative to the room modes, and any acoustic reflectors in the midbass region out of the near field of the loudspeakers or your main listening position.

 

Chris

Edited by Chris A
Link to comment
Share on other sites

they always seem to make a point to talk about how high up they can play flat....i.e. this one is flat to 200hz.

Upper extension with subs has to do with the blending of the mains. Usually subs with good upper extension have a low "Le" which is the voice coil inductance. Look at Acoustic Elegance subs if this is what you're after, they're the kings of this realm. Although, you see more benefits of this while in a two way design with it being used as a midwoofer than with a home theater sub.

Bragging about flatness up high could be talking about the opposite as well though, some subs can be boomy up high, possibly due to an enclosure that isn't optimal. Flat is good. Boomy and colored, not so much.

Edited by MetropolisLakeOutfitters
Link to comment
Share on other sites

The headroom gained is in the amplifier. If it doesn't have to reproduce lower frequencies it uses less of its available power from its power supply and will have more available energy for the mid and high frequencies.

I imagine it's both speakers and amps to some extent. Consider the big mains in a commercial theater. Those big 15" woofers may only have like 6mm xmax. You don't want to be sending 20 hz signals strong enough to fill up a commercial theater to those guys no matter what kind of amp you have. You're going to have to turn the volume down if you did. If you want to hit reference level, you're going to have to offload the heavy lifting to the subs.

Edited by MetropolisLakeOutfitters
  • Like 1
Link to comment
Share on other sites

Example:

 

Mains 40 Hz to 20 kHz

Center 60 Hz to 20 kHz

Surrounds 70 Hz to 20 kHz

Sub 20 Hz to 300 Hz

 

The LPF should be set 10-15 Hz above the weakest speaker.  In this case the Surrounds.  So, 80 Hz is good but, the range could be anything between 80 Hz and 300 Hz.  Sub localization happens around 100 Hz so, to prevent that most people will not set the LPF higher than 100 Hz.

 

2 ch guys love to set low XO and some HT like to set a higher XO.  Keeping the low stuff out of the Mains will net 5-6 db of headroom in the avr or amp.  You are shifting the heavy stuff to the sub or robbing Peter to pay Paul.  Isn't this just as bad? Not really, most subs have there own dedicated amp.  In most cases, the sub amp is much stronger than the avr or amp driving the mains.  Running the Mains full range in HT can eat into headroom.

Edited by derrickdj1
Link to comment
Share on other sites

I cant explain it well at all but there is also supposedly a certain type of distortion, modulation maybe, that occurs when the mains do too much work down low, basically imagine a strong and low test tone being sent to the mains so they're moving real good, then you play somebody's voice through the same speaker. The tone of the voice will actually change due to the cone moving violently from the test tone, trying to reproduce music while it is moving pretty hard.

This situation also produces distortion due to eddy currents. Acoustic elegance is good about this aspect though.

Offloading that material to the subs keeps everything else cleaner regardless of the amp. I would think that when talking about headroom, you would want it to be clean.

Link to comment
Share on other sites

but there is also supposedly a certain type of distortion, modulation maybe, that occurs when the mains do too much work down low, basically imagine a strong and low test tone being sent to the mains so they're moving real good, then you play somebody's voice through the same speaker. The tone of the voice will actually change due to the cone moving violently from the test tone, trying to reproduce music while it is moving pretty hard.

 

See https://community.klipsch.com/index.php?/topic/151087-why-horn-loaded-sounds-better-than-direct-radiating-faq/ for more information and a visualization of the difference between modulation distortion (AM and FM) and harmonic distortion. You can easily measure the levels of modulation distortion by looking at the sidebands a two-tone test signal played on the woofers of your fronts or subwoofer.  Just use REW to produce the two-tone signal and look at the side band amplitudes on either side of the higher frequency signal.  It's not hard to see them.

 

Horn-loaded woofers exhibit much lower side band amplitudes than direct radiating woofers, typically 25 dB lower, i.e., amplitudes about 1/20th to 1/25th the amplitude of the same woofers used in direct radiating mode to produce the same SPL output.

 

That's why horn-loaded bass bins sound so much better that direct radiating ones, especially vented boxes, which also experience a big rise in group delay at port resonance frequencies. 

 

The trick of choosing a crossover point between your direct radiating subwoofer(s) and your other direct-radiating woofer surrounding loudspeakers in-room is a trade-off between the factors listed in post #4, above.  The OP is free to ask questions on any of these various factors.

 

I find two factors usually control where to cross over:

 

1) room acoustics (subject mainly to the dimensions of the room, the placement of the loudspeakers, and the placement of the listening position), and

 

2) the inherent sound quality of the bass bins/subwoofers, usually controlled by the diameter of the bass drivers, the expansion efficiency and rate of the associated bass horns (if any), and the linearity of the drivers in "large signal" (Richard Small's terminology) amplitudes.

 

I find that having to cross over at frequencies above 80 Hz is usually driven by poor sound reproduction system choices for bass reproduction for a given room.  Educating yourself on the physics of bass modulation distortion and room acoustics usually leads to different choices in bass configurations, and in room choice--loudspeaker/listening positions.

 

Chris

Link to comment
Share on other sites

 

but there is also supposedly a certain type of distortion, modulation maybe, that occurs when the mains do too much work down low, basically imagine a strong and low test tone being sent to the mains so they're moving real good, then you play somebody's voice through the same speaker. The tone of the voice will actually change due to the cone moving violently from the test tone, trying to reproduce music while it is moving pretty hard.

 

See https://community.klipsch.com/index.php?/topic/151087-why-horn-loaded-sounds-better-than-direct-radiating-faq/ for more information and a visualization of the difference between modulation distortion (AM and FM) and harmonic distortion. You can easily measure the levels of modulation distortion by looking at the sidebands a two-tone test signal played on the woofers of your fronts or subwoofer.  Just use REW to produce the two-tone signal and look at the side band amplitudes on either side of the higher frequency signal.  It's not hard to see them.

 

Horn-loaded woofers exhibit much lower side band amplitudes than direct radiating woofers, typically 25 dB lower, i.e., amplitudes about 1/20th to 1/25th the amplitude of the same woofers used in direct radiating mode to produce the same SPL output.

 

That's why horn-loaded bass bins sound so much better that direct radiating ones, especially vented boxes, which also experience a big rise in group delay at port resonance frequencies. 

 

The trick of choosing a crossover point between your direct radiating subwoofer(s) and your other direct-radiating woofer surrounding loudspeakers in-room is a trade-off between the factors listed in post #4, above.  The OP is free to ask questions on any of these various factors.

 

I find two factors usually control where to cross over:

 

1) room acoustics (subject mainly to the dimensions of the room, the placement of the loudspeakers, and the placement of the listening position), and

 

2) the inherent sound quality of the bass bins/subwoofers, usually controlled by the diameter of the bass drivers, the expansion efficiency and rate of the associated bass horns (if any), and the linearity of the drivers in "large signal" (Richard Small's terminology) amplitudes.

 

I find that having to cross over at frequencies above 80 Hz is usually driven by poor sound reproduction system choices for bass reproduction for a given room.  Educating yourself on the physics of bass modulation distortion and room acoustics usually leads to different choices in bass configurations, and in room choice--loudspeaker/listening positions.

 

Chris

 

Wow.  That was quite a side-trip into distortion!

 

Most interesting is that if I am understanding this better, what I thought was headroom is actually lower FM Distortion.  When a sub is given it's entire range of frequencies to play, say from 20-200, the fact that it plays such a large range of lower frequencies causes higher FM Distortion.

 

When pushing a sub to it's mechanical limits, AM Distortion comes into play.  Its no longer as linear and this difference translates to our ears as distortion.  

 

When pushing a sub to it's thermal limits, compression comes into play, and the frequency response is no longer the same, another form of distortion.  (FM distortion again?

 

Chris am I on the right track?

Link to comment
Share on other sites

Example:

 

Mains 40 Hz to 20 kHz 

Center 60 Hz to 20 kHz

Surrounds 70 Hz to 20 kHz

Sub 20 Hz to 300 Hz

 

The LPF should be set 10-15 Hz above the weakest speaker.  In this case the Surrounds.  So, 80 Hz is good but, the range could be anything between 80 Hz and 300 Hz.  Sub localization happens around 100 Hz so, to prevent that most people will not set the LPF higher than 100 Hz.

 

2 ch guys love to set low XO and some HT like to set a higher XO.  Keeping the low stuff out of the Mains will net 5-6 db of headroom in the avr or amp.  You are shifting the heavy stuff to the sub or robbing Peter to pay Paul.  Isn't this just as bad? Not really, most subs have there own dedicated amp.  In most cases, the sub amp is much stronger than the avr or amp driving the mains.  Running the Mains full range in HT can eat into headroom.

 

I'm quoting Derrick here, but Chris I would like your opinion on Derrick's crossover discussion above.

Is he "right", generally speaking, about where the crossover should be set?

 

Generally speaking, do you think all the speakers should be crossed at the same frequency?

 

Thank you,

RK

Link to comment
Share on other sites

Chris gave some great info.  Using multiple subs means that the cones are not moving as much which  lowers the systems distortion.  Also, running the mains as small means their cones are not movie and much and the sound is cleaner for the same reason.  Most people are not running the system on full blast so, with good linear drivers in direct radiator subs, a very clean sound is possible.  For the UM 18 sub, the distortion is very low, i.e. under 3% for most of it's passband.  Ricci, at Database.com set a low of 10% for subwoofers.

Link to comment
Share on other sites

Rob, I will let Chris answer your question in the above post.  XO for different speaker is a feature on some avr's.  XO a center over at 60 Hz and mains at 50 Hz is not going to be audible IMHO.  I am very comfortable with Pioneer avr's global XO setting.  As Earl Geddes, a prominent acoustic researcher pointed out, when using subwoofers, there is not reason to run the Mains full range or use the lowest XO point.

 

Different avr's use different schemes of bass management,  Some actually will provide bass management while running the Mains set to Large.  This may help the FR for some less capable speakers systems that don't produce enough bass but, it has it's drawbacks.

Link to comment
Share on other sites

Most interesting is that if I am understanding this better, what I thought was headroom is actually lower FM Distortion. When a sub is given it's entire range of frequencies to play, say from 20-200, the fact that it plays such a large range of lower frequencies causes higher FM Distortion...When pushing a sub to it's mechanical limits, AM Distortion comes into play. Its no longer as linear and this difference translates to our ears as distortion.

 

 

Actually, AM distortion comes into play immediately at low frequencies, and will be the most significant contributor to bass distortion.  The nonlinear effects of having non-overhung voice coils in the woofers immediately come into play and be audible even at reasonably low SPL.  (Woofers with overhung voice coils have heavier moving mass than typical zero-overhung woofers--which exhibit BI modulation effects when used as direct radiators.  Overhung VC woofers also are typically extremely expensive.)

 

FM distortion shows up as the dominant distortion type on drivers handling midrange-tweeter frequency bands.

 

 

I'm quoting Derrick here, but Chris I would like your opinion on Derrick's crossover discussion above. Is he "right", generally speaking, about where the crossover should be set? Generally speaking, do you think all the speakers should be crossed at the same frequency?

 

 

What Derrick is saying is what I'd do to handle all the surrounding loudspeakers' limitations (with the exception of crossing above ~80 Hz...I'd find bigger center and surround loudspeakers to have lower LF cutoffs instead.)  Center and surround loudspeakers typically have much higher bass cutoff frequencies than the front left/right loudspeakers, sometimes as much as two octaves higher.  So AVRs/AVPs typically have the ability to handle different bass crossover frequencies because of surround loudspeaker limitations.

 

It would be cleanest sounding to have each surrounding loudspeaker able to handle the full bass spectrum without having to combine their LF channel output into a single "subwoofer" channel.  However, if you cross them low enough, bass localization in small rooms (like all of our home listening spaces) will be difficult to localize if shared with the subwoofer.  However, that puts extra resulting FM and AM distortion loads on the subwoofer.  It's usually a reasonable compromise, however, given surround loudspeaker limitations.  [You wouldn't really want to do that in a commercial movie theater, however, with their much larger distances between surround loudspeakers and much better human hearing localization at bass frequencies.]

 

Chris

Link to comment
Share on other sites

There is some interesting new hypothesis on multiple subs which hasn't been discussed here as far as I know. With a sub in front and a sub in the back or the HT, the cutoff is different in this hypothesis. The front sub overlaps the low frequencies of the L/C/R, and the rear sub is crossed much lower to increase modal density and be 100% non-directional as it only provides very very deep bass frequencies for the overall room LFE. I think the phase is set on the front sub to the mains first and then the rear phase is adjusted.

 

I will dig around till I find that study, but I think it is at least thought provoking.

Link to comment
Share on other sites

One of the things that I find interesting is how low frequencies are treated differently than the rest of the audible spectrum by most people.  The reality is that it's a continuum with no differences in how we should look at their reproduction.  When you think about it from a strictly acoustical perspective, you'd really like to have all channels able to reproduce down to the limit of acoustic hearing--usually pinned at somewhere between 20-30 Hz (below which the human hearing system gives way to other human sensory systems).  When I has a kid, that's how everyone looked at reproducing low frequencies.  I still view the issue this way to a great degree.

 

In the early-to-mid-1960s, the first "subwoofer" patent was granted, culminating in 1974 with the film Earthquake using a patented system called "Sensurround" (developed by Don Keele, et al.).  Since then, a "subwoofer" mentality has risen.  I find that the way that most others have developed their views toward LF reproduction--"subwoofers" in particular--has changed that view into a new set of "truisms" (actually they're more like memes).  I find this view to be less useful.  It's a continuum.  Loudspeakers reproduce sound - all frequencies that we can hear and sense.  Breaking loudspeakers up into separate pieces isn't really something that I view as useful.  Others seem to gloss over the problems in viewing LF reproduction this way--as separate devices. 

 

There are really just two things that are different about LF reproduction:

 

1) the wavelengths are so long that in small rooms like our home listening rooms, we're always sitting in the "nearfield"--less than 2 wavelengths--so that we have a difficult time judging the direction of arrival of LF wavefronts (unlike in commercial movie theaters with their much larger internal distances between surfaces).  In small rooms, the room itself experiences a "sparse mode region" at these frequencies since resonances and cancellations of LF waves occur at much fewer frequencies due to the required size of the re-radiating surfaces of the room: entire walls, the whole ceiling and the floor of small rooms, instead of smaller objects like tables, equipment racks, and loudspeaker enclosures. There are just not that many modes of how the sound gets reflected around the room, resulting in a "Schroeder frequency" for the room.  The Schroeder frequency is typically below 200 Hz: in my room it's about 110 Hz.  This is the approximate crossover frequency between the "dense mode region" and the "sparse mode region" of the room.

 

2) the human hearing system gets less and less sensitive in terms of SPL as the wavelengths get longer and longer (i.e., relative to the length of the human ear canal which serves as a horn antenna to the eardrums), which means that the SPL of these wavelengths must remain very high for us to hear them at all. [Double bass and tuba players never play their instruments as softly as violas, violins, and trumpets, etc--they always play fairly loud in comparison.]  Also, the human hearing system gets less able to judge direction of arrival as the wavelengths get larger than the distance between the two eardrums.

 

Otherwise, sound is the same.  Looking at low frequencies as somehow different than all the other frequencies that your loudspeakers must reproduce is not really very useful or insightful to this view. 

 

I find it's more useful to think of subwoofers as separate boxes that can be separately located from the loudspeakers, but not advisable to locate them very far away in reality: less than 1/4 wavelength at the crossover frequency.  Any longer than that, and you start to experience coherence issues in the crossover region. At 80 Hz (at room temperature) that's about 3.5 feet, and at 40 Hz that's about 7 feet.  You're beginning to catch my point.  Only point #2 above lets you get away with separating the subwoofer to a greater distance than the rest of the loudspeaker and your ears to not cry "foul". It's not advisable to separate the subwoofer any further if you're trying to cross at a higher frequency. 

 

Also, there is the problem of converting to mono from stereo if you're trying to cross over higher than 80 Hz.  The same sort of limitations apply to the human hearing system allowing you to do this without objections being lodged to your conscious hearing processes.

 

Chris

Edited by Chris A
Link to comment
Share on other sites

For example, if your sub can go up to 200hz, but you cross it at 80hz or 100 or 120....are you gaining "headroom" with your sub at the lower frequencies, or does this not apply?

 

Short answer is yes.

 

However, you need to think of it in terms of bandwidth. Covering 20Hz to 80Hz is only 2 octaves. Extending that to cover 80Hz to 160Hz is 3 octaves. Your main speakers on the other hand are covering 80Hz to 20kHz. Usually your speakers are 2-way or 3-way, so really the next region of bandwidth is probably something like 80Hz to 1280Hz, which is 4 octaves. If you let them go down to 40Hz, then it's 5 octaves.

 

So we have 2 octaves on the sub and 4 octaves for the LF section of your mains. Headroom is a bigger issue for your mains than it is for the subs.

 

 

Now all that said - headroom should not be the dominant deciding factor for choosing a xover frequency. There are more variables to consider, but that's a huge topic when you start getting into it. At the end of the day, I'd recommend sticking to 80Hz for the subwoofer no matter what speakers and subs you have. It's a rare scenario where deviating from this is an advantage - and even then it's usually going to be a subtle difference. Definitely don't try going higher than 80Hz on the xover frequency unless you have multiple subs colocated with your mains.

  • Like 1
Link to comment
Share on other sites

but not advisable to locate them very far away in reality: less than 1/4 wavelength at the crossover frequency.  Any longer than that, and you start to experience coherence issues in the crossover region.

 

 

Once you're in the modal region of a room, then that kind of analysis doesn't apply. The modal region usually covers the 50Hz to 400Hz range in most rooms. We have to treat all of the reflective surfaces as multiple sources - in which case we want to approach the problem more like a line array, rather than trying to achieve a point source. The xover frequency is only adding one more additional source, and only over a very narrow range of frequencies.

Edited by DrWho
Link to comment
Share on other sites

but not advisable to locate them very far away in reality: less than 1/4 wavelength at the crossover frequency.  Any longer than that, and you start to experience coherence issues in the crossover region.

 

Once you're in the modal region of a room, then that kind of analysis doesn't apply. The modal region usually covers the 50Hz to 400Hz range in most rooms. We have to treat all of the reflective surfaces as multiple sources - in which case we want to approach the problem more like a line array, rather than trying to achieve a point source. The xover frequency is only adding one more additional source, and only over a very narrow range of frequencies.

 

I believe that you read my statement differently than the intent: I'm referring to sound localization, which is audible at least down to 80 Hz.

 

The value of 400 Hz that you stated above is almost to tuning fork frequency (A=440 Hz, the "A" above middle C on the piano).  Human localization of sound is extremely audible in this region.

 

Below 80 Hz (more than two octaves lower than 400 Hz), the point generally agreed to represent the limit of human localization in mixed environments, is where the ability to discern interaural time delays becomes problematic. 

 

The minimum audible angle (MAA) vs. frequency is shown in the following figure:

 

MAudA.jpg

 

Being within a 1/4 wavelength of the crossover frequency to the sub is a rule-of-thumb since the actual angle difference between the loudspeaker and the subwoofer is also dependent on distance from the listener to the loudspeakers, as well as the lateral offset of the subwoofer to the surround loudspeaker.  A quarter wavelength maximum separation distance at the crossover frequency is more than a generous limit...probably too generous.  Many people have their subs located 10s of degrees outside the location of each surround loudspeaker. 

 

The 80 Hz upper limit of crossing over to a sub is also "pushing the envelope" a bit in terms of its generosity, IME.

 

Chris

Edited by Chris A
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...