Subwoofer crossover, towers, bookshelves, and midbass slam.

[yepimonfire]

The thread I recently posted regarding whether or not towers were needed with a good sub got me thinking about getting impactful midbass (about 50hz-150hz range) with a subwoofer integration.

Frequencies between 50-70hz are generally responsible for most of the “kick in the chest” tactile response from heavy bass (think of the impact from a loud kick drum).

I’ve always preferred a crossover at 60hz or lower when possible, most large subs tuned low just don’t seem to reproduce midbass frequencies with the same fullness and impact in my experience as speakers capable of reproducing at least 50hz, which is why I prefer lower xover points.

The real question, is how much speaker is needed to reproduce this range with movies or highly dynamic movies? Based on my own measurements, my Klipsch RP 150ms with a single 5.25” ceramic/aluminum woofer and box tuning of around 60hz are capable of reproducing 50hz-200 at low distortion (>3% THD) with the with the volume control set to -10dB with -6dBfs pure sine waves, which have a crest factor of 6dB, which translates to about 95dB at 10 1/2 away at the mlp. The RP-250c, which utilizes dual 5.25 inch woofer and an identical box tuning, is capable of reproducing 50hz @97dB at the same distance.

Since I don’t exceed -10dB volumes due to the fact that 1. My subwoofer bottoms out at 35hz above 105dB @10.5’, making this the max safe volume, and 2. My wife can’t tolerate full reference level (0dB), I would assume this is plenty sufficient. The max headroom in the midbass for the front three speakers seems to be about -7dBon the volume knob, at this point, significant distortion sets in (likely due to exceeding xmax causing the vc to extend outside of the gap). It seems the cerametallic woofers do an excellent job remaining pistonic even at high excursion, and the tractrix flared port is chuff free. Would there really be any benefit to larger midbass drivers if my current speakers are capable of comfortably handling my spl requirements?

I wonder if a dedicated midbass sub would be useful. A 12” behringer pa sub is tuned to about 50hz (since most music doesn’t really require much below 40hz) and would be capable of dishing out stupidly high levels of midbass, a few members over at avsforum have used it with a separate crossover like this, low passing the regular sub at ~50hz with an external xover and setting the avrs xover to 150hz-200hz. Hsu also make an MBM, as does diysg.

Secondly, has anybody else tried a lower xover point at 60hz and below and noticed better midbass impact than offered by crossing over at 80hz? IME most bigger subs seem to really lack in performance above 60hz, especially at producing good tactile response.


Sent from my iPhone using Tapatalk

 

[lovinthehd]

What sub experience do you have aside from the Dayton subs you have?

 

[yepimonfire]

What sub experience do you have aside from the Dayton subs you have?

Used to have a really nice kef sub that was flat to 20hz. I’ve heard a few other subs, nothing like a big svs or hsu.


Sent from my iPhone using Tapatalk

 

[yepimonfire]

Also forgot, a buddy of mine used to have a monster diy sub that he claimed was flat to about 14hz.

Just going by response graphs, it’s obvious low tuned subs don’t perform as well above 60hz and especially 80hz. Obviously, there are exceptions, but more often than not, large subs just don’t have a high extension. The SVS PB-16 for example, starts taking a nosedive at ~70hz https://www.soundandvision.com/content/svs-pb16-ultra-and-sb16-ultra-subwoofers-review-test-bench


Sent from my iPhone using Tapatalk

 

[everettT]

I don't think there is any issue eqing the PB16 flat out to 120hz at all. As for large Subs not having extension out past 100hz, you should do some investigation on it as there a many that do. The 15s in my subs have no problem going out to 250hz with plenty of headroom.

Here is a 21" driver with those parameters

http://data-bass.com/data?page=system&id=127&mset=138

Also forgot, a buddy of mine used to have a monster diy sub that he claimed was flat to about 14hz.

Just going by response graphs, it’s obvious low tuned subs don’t perform as well above 60hz and especially 80hz. Obviously, there are exceptions, but more often than not, large subs just don’t have a high extension. The SVS PB-16 for example, starts taking a nosedive at ~70hz https://www.soundandvision.com/content/svs-pb16-ultra-and-sb16-ultra-subwoofers-review-test-bench


Sent from my iPhone using Tapatalk

 

[lovinthehd]

Used to have a really nice kef sub that was flat to 20hz. I’ve heard a few other subs, nothing like a big svs or hsu.


Sent from my iPhone using Tapatalk because I'm lazy and like annoying you

I wouldn't call that much experience, especially as regards your statement as to response range. Go spend some time on data-bass perhaps like Everett points out....

ps I have several subs that have good response well beyond 200hz

 

[lukesky518]

Physics being what they are a 6 - 6.5" driver in a bookshelf enclosure is not a woofer . Couple that with a 12" sub woofer and there will be some problems with integration , especially on music . A 6-6.5" driver in a tower can be a woofer if tuned correctly which makes it easier to integrate . If I had bookshelf speakers I would use an 8" or maybe a 10" sub to integrate better and give the upper bass more presence . BTW ...just because a sub can go to 200hz doesn't mean you should . For movies , maybe . For music , I wouldn't .

 

[TLS Guy]

There is so much misunderstanding of how you achieve what you are aiming for.

As usual you are looking at frequency response and power bandwidth.

Yep I think you have been round here long enough to know what I'm going to tell you. Time, time, time and time!

Now this is something that professionals do not get their heads around.

The reason is that we make frequency response the be all and end all. I could very easily make a speaker with a flat frequency response, that was totally useless as a speaker where with speech you would not catch a word.

Actually the importance of time is neglected in many fields of science and engineering. After all time is the fourth dimension of the universe.

We pay lip service to phase. As I keep pointing out, because we measure it in the degrees of a circle. So we miss its significance. Stop thinking of phase in the parlance of the circle. From now on express it as time. This way the reality of what you are doing will hit you full face forward.

Now you will see adding another driver with a narrow bandwidth is a really terrible idea. Yes, now you get it, you will add more time shift, and band pass gain for good measure.

I know you are really interested in all this. The first thing to become inside your head, if not entirely in practice is a full range driver enthusiast. If I played you some drums with single full range Lowther driver expertly horn loaded and played you a well recorded drum track it would just blow you away. Yes, the slam would be there aplenty and the realism astonishing. The reason being that the hit, crack, bang and decay would all be in time. Time is the key here.

Once you start adding drivers with varying pass bands, you have set yourself a lofty task.

An absolute prerequisite of doing what you want is to preserve the timing of the entire drum hit as precisely as you can.

We got into this territory the other day, but I did not feel up to responding. Speakers that have good timing also have good depth of field and reproduce the original venue more accurately.
This is especially true if minimalist miking is used. Unfortunately phase difference which means time difference mic techniques have been the order of the day, rather than intensity difference techniques. The trouble is that we have the old proverbial chicken and egg scenario.

Speakers that do a good job of preserving time relationships do a superb job of reproducing minimalist intensity difference recordings. I just love to demonstrate some of those recordings I have made. Unfortunately speakers that have significant time aberrations, which is most, do a very poor job of reproducing them.

 

[shadyJ]

In my opinion, 'slam' is a matter of sufficient SPL that produces a tactile sensation. It doesn't make a difference what speaker can produce it so long as it is produced. Obviously satellite speakers are going to have a lot tougher time doing that than mains that have 15" midrange woofers. I can't help but to wonder if there is something about near-field acoustics that makes tactile sensations more sensible even for the same SPL. There has been speculation that experiencing the direct sound wave before its coherence is lost in scattered room reflections is the key because you get hit with all that acoustic energy at the same time as opposed to sitting in the far-field where the indirect-to-direct energy ratio is much lower.

 

[killdozzer]

Actually the importance of time is neglected in many fields of science and engineering. After all time is the fourth dimension of the universe.

Word!

 

[yepimonfire]

There is so much misunderstanding of how you achieve what you are aiming for.

As usual you are looking at frequency response and power bandwidth.

Yep I think you have been round here long enough to know what I'm going to tell you. Time, time, time and time!

Now this is something that professionals do not get their heads around.

The reason is that we make frequency response the be all and end all. I could very easily make a speaker with a flat frequency response, that was totally useless as a speaker where with speech you would not catch a word.

Actually the importance of time is neglected in many fields of science and engineering. After all time is the fourth dimension of the universe.

We pay lip service to phase. As I keep pointing out, because we measure it in the degrees of a circle. So we miss its significance. Stop thinking of phase in the parlance of the circle. From now on express it as time. This way the reality of what you are doing will hit you full face forward.

Now you will see adding another driver with a narrow bandwidth is a really terrible idea. Yes, now you get it, you will add more time shift, and band pass gain for good measure.

I know you are really interested in all this. The first thing to become inside your head, if not entirely in practice is a full range driver enthusiast. If I played you some drums with single full range Lowther driver expertly horn loaded and played you a well recorded drum track it would just blow you away. Yes, the slam would be there aplenty and the realism astonishing. The reason being that the hit, crack, bang and decay would all be in time. Time is the key here.

Once you start adding drivers with varying pass bands, you have set yourself a lofty task.

An absolute prerequisite of doing what you want is to preserve the timing of the entire drum hit as precisely as you can.

We got into this territory the other day, but I did not feel up to responding. Speakers that have good timing also have good depth of field and reproduce the original venue more accurately.
This is especially true if minimalist miking is used. Unfortunately phase difference which means time difference mic techniques have been the order of the day, rather than intensity difference techniques. The trouble is that we have the old proverbial chicken and egg scenario.

Speakers that do a good job of preserving time relationships do a superb job of reproducing minimalist intensity difference recordings. I just love to demonstrate some of those recordings I have made. Unfortunately speakers that have significant time aberrations, which is most, do a very poor job of reproducing them.

You and I are probably the only two here that have noticed the effects of temporal differences even when frequency response is identical. My system measured the same no matter what xover I use, but sounds different regardless.

I’m not sure a full range drivers time coherence is worth the other issues that come about, like collapsing polars and increased IMD. To my ears, nothing sounds worse than IMD, and I have become a huge fan of constant directivity designs due to their excellent off axis coverage and immunity from timbre aberrations introduces by early reflections. Good time coherence can be achieved with multiple driver systems if proper xover points are utilized. Mutual coupling (ie the drivers behaving as a single radiating source) between drivers is achieved when the spacing between them is less than 1/2 wavelength.

My guess is that a lower crossover point has better summing in the time domain than a higher one. My sub is 7’ from the left channel (the farthest speaker) a crossover of 50hz places this well below the half wavelength distance, and nearly within 1/4 wavelength, while an 80hz or greater puts it just outside a half wavelength. Since the distance between the sub and the left and right speakers differs, it’s impossible to properly time align them for both speakers. Perhaps I should experiment with placing the sub directly in the middle and seeing how it sounds at 80hz.

In addition, the driver spacing of my mid/tweeters is just under half a wavelength as well, resulting in perfect summing in the far field, and no lobing at the crossover point.


Sent from my iPhone using Tapatalk

 

[everettT]

You and I are probably the only two here that have noticed the effects of temporal differences even when frequency response is identical.
Sent from my iPhone using Tapatalk

I needed that, thanks

 

[everettT]

Oh and how's this for impulse response on that 21" driver

 

[TLS Guy]

In my opinion, 'slam' is a matter of sufficient SPL that produces a tactile sensation. It doesn't make a difference what speaker can produce it so long as it is produced. Obviously satellite speakers are going to have a lot tougher time doing that than mains that have 15" midrange woofers. I can't help but to wonder if there is something about near-field acoustics that makes tactile sensations more sensible even for the same SPL. There has been speculation that experiencing the direct sound wave before its coherence is lost in scattered room reflections is the key because you get hit with all that acoustic energy at the same time as opposed to sitting in the far-field where the indirect-to-direct energy ratio is much lower.

I could not disagree more. It is not just a matter of spl.

A drum crack is complex, there is the strike with a lot of HF content the bang and decay.

Timing of all this is crucial.

I have an instructor at a college. He has Genelec speakers in his studio. He vows my speakers are the only ones he has heard that reproduce drums realistically.

 

[KEW]

Oh and how's this for impulse response on that 21" driver

That is good, specially compared to most ported subs, but this one is about as good as I have seen!

Maybe I am misinterpreting, but the PSA gets up to full amplitude almost immediately, and while it is not quite as quick to settle back down, it does get back within 10dB very quickly.

 

[shadyJ]

That is good, specially compared to most ported subs, but this one is about as good as I have seen!

Maybe I am misinterpreting, but the PSA gets up to full amplitude almost immediately, and while it is not quite as quick to settle back down, it does get back within 10dB very quickly.

Again, both you of guys are misreading those graphs. That is not a measure of transient response. Neither measurement is especially good or bad.

 

[yepimonfire]

Again, both you of guys are misreading those graphs. That is not a measure of transient response. Neither measurement is especially good or bad.

Better to look at spectral decay or waterfalls in an anechoic setting. I believe kew measured the Dayton subs and found they were “quicker” than the pb 1000. I don’t think it’s an issue of driver transient response, more time alignment with the mains. I’m sure with some good measurements I could get it closer, but since my front three speakers have no problem reproducing 50-200hz at 95dB @ under 3% THD, and ~1.5% IM distortion when a 50hz & 1500hz tone is played at a 4:1 ratio, also at 95dB at the seating area 10 1/2 feet away, there’s no good reason to force a higher xover. They don’t begin to break up until around 98dB for the bookshelves and 102dB for the mtm center.


Sent from my iPhone using Tapatalk

 

[shadyJ]

Better to look at spectral decay or waterfalls in an anechoic setting. I believe kew measured the Dayton subs and found they were “quicker” than the pb 1000.

spectral decay in an anechoic setting is a tall order. I would be looking at group delay.

I have rarely seen impulse response correctly interpreted.

 

[yepimonfire]

spectral decay in an anechoic setting is a tall order. I would be looking at group delay.

I have rarely seen impulse response correctly interpreted.

Okay, my speakers measure pretty flat down low on GD, no spikes of any sort, the sub also measures well by itself with no peaks until about 17hz, which is way below its response anyways. When using an 80hz xover, measuring both the subs and speakers, a large spike at 80hz occurs. Even making adjustments to the delay in the avr, anywhere from 4’-20’ makes no difference. I do not see GD when measuring the combo of the R speaker (about 2’ from the sub) and sub, only the left speaker and sub(about 9’ from the sub) or combined response. Adjusting the delay in the avr above or below the measured distance of 9.6’ simply causes large cancellations in the FR measurements, without solving the issue. At 60hz and below, there is no more GD spikes. While it measures flat regardless (as flat as in room an be anyways), the bass sounds “slow” when an 80hz xover is used.


Sent from my iPhone using Tapatalk

 

[Andrein]

I personally prefer 60hz xo because i think speakers cones, especially if they are small enough but capble of reproducing 50-60hz, returning back to initial position quicker than subs large drivers. This makes to me sound cleaner and arguably punches stronger where they actually are.