Hi folks-
I bet many of us have systems with a crossover frequency right smack dab in the Schroeder region. Let's chew on that specific problem. Do rules of thumb exist?
Crossover slopes is my main question - do we like a particular "taper" or "feathering" from ray behavior to wave behavior?
I'm thinking more about how you slope the woofers than the mains. I think the mains lowpass is dictated by your power bandwidth needs, and you typically take all the lo end you can get.
On the other hand, you can noodle the subwoofer highpass all day and call it art. But what do the great artists do?
Put another way: Assuming ideal speakers & ideal subs but non-ideal (real) rooms, does an optimum Schroeder transition philosophy reveal itself? Thanks.
I bet many of us have systems with a crossover frequency right smack dab in the Schroeder region. Let's chew on that specific problem. Do rules of thumb exist?
Crossover slopes is my main question - do we like a particular "taper" or "feathering" from ray behavior to wave behavior?
I'm thinking more about how you slope the woofers than the mains. I think the mains lowpass is dictated by your power bandwidth needs, and you typically take all the lo end you can get.
On the other hand, you can noodle the subwoofer highpass all day and call it art. But what do the great artists do?
Put another way: Assuming ideal speakers & ideal subs but non-ideal (real) rooms, does an optimum Schroeder transition philosophy reveal itself? Thanks.
If your drivers and their horns (if any) that are being crossed have relatively similar phase response into their crossover interference stop bands, and neither driver (or drivers) are running into problems with maximum excursion or power handling in the interference band at maximum SPL output conditions anticipated, then I don't see any reason not to use the lowest order crossover filter that you can (nominally, that's first order, but it could be zeroth order). In fact, it will probably sound better to use lower order IIR filters than higher order IIR filters, in order to reduce or eliminate induced phase growth on the lower frequency drivers (ref. the audibility of bass phase distortion).
The only problem is that the definition of Schroeder frequency is defined by the longest room dimension and its reverberation time (supposedly RT60, but in home hi-fi rooms, that standard is too deep--more like RT30). If you're using the loudspeakers only in relatively small home-sized hi-fi rooms, then there's no problem. If you put the loudspeaker into a larger room, one that's heavily damped (i.e., RT30 <0.2 s or so), or use it outside, then you may be able to hear the effects of any directivity mismatch issues in the 100-200 Hz band (the "boominess" band).
Chris
The only problem is that the definition of Schroeder frequency is defined by the longest room dimension and its reverberation time (supposedly RT60, but in home hi-fi rooms, that standard is too deep--more like RT30). If you're using the loudspeakers only in relatively small home-sized hi-fi rooms, then there's no problem. If you put the loudspeaker into a larger room, one that's heavily damped (i.e., RT30 <0.2 s or so), or use it outside, then you may be able to hear the effects of any directivity mismatch issues in the 100-200 Hz band (the "boominess" band).
Chris
Excellent topic!
IMO a Schoeder frequency crossover should be considered with 3 way designs where the mid to low crossover point can be around that frequency depending on the room.
The last time I had a system with the crossover dialed into the rooms Schoeder frequency I used a 24DB active. With modern DSP a person can tweak slopes and responses until the cows come home. I saw no short comings.
When designing the woofer box make sure the internal nodes are above the Schroder frequency. This will give best response below but limits box size since the larger the box the lower the first node.
Another point to consider is the baffle step response, this can now be applied to the mid\high subsystem or by simply adjusting the level via that subsystems amp.
Furthermore, distance between woofer and mid should still be considered. With large or multiple woofers beaming can still occure.
IMO a Schoeder frequency crossover should be considered with 3 way designs where the mid to low crossover point can be around that frequency depending on the room.
The last time I had a system with the crossover dialed into the rooms Schoeder frequency I used a 24DB active. With modern DSP a person can tweak slopes and responses until the cows come home. I saw no short comings.
When designing the woofer box make sure the internal nodes are above the Schroder frequency. This will give best response below but limits box size since the larger the box the lower the first node.
Another point to consider is the baffle step response, this can now be applied to the mid\high subsystem or by simply adjusting the level via that subsystems amp.
Furthermore, distance between woofer and mid should still be considered. With large or multiple woofers beaming can still occure.
A quarter wavelength at 100 Hz is about 34 inches (86 cm). Lobing at these frequencies is probably not an issue between woofers in a loudspeaker box, and by definition ceases to be an issue when the 1/2 wavelength of sound is greater than the larger of the width or height of the listening room (i.e., the room end can no longer support a half wavelength).
This is why some corner horns (e.g., bifurcated horn bass bins) can be located away from the room's walls at a distance less than or equal to 1/4 wavelength at the frequency/wavelength where the first 1/4 wavelength of sound in the horn begins to exit the horn's mouth, even though they possess exponential or hypex horn expansions that have a sharp lower frequency cutoff.
Chris
Last edited:
Wow 300Hz is really up there, I didn't realize anybody's room went that high. I'm definitely thinking lower down, in the traditional sub crossover range.
Once down there, we find an additional complication: your ears are going monophonic as frequencies go down. All of this is happenning in pretty much the same region of the spectrum.
And it's a region, not a single frequency. I bet for many rooms it's more than an octave wide and of course quite vague. That's why I'm pounding on slopes here - you gotta slide your way thru this region, it ain't like flipping a switch.
In a multi-woofer setup you also have some thinking to do about the 'shape' of the sound you can make as you blend between wildly different radiation patterns either side of crossover.
I've done enough playing to know this frequency region powerfully controls the stereo soundstage and the solidity of the central image. You can widen & narrow the 'hole in the middle' effect by playing around here.
Once down there, we find an additional complication: your ears are going monophonic as frequencies go down. All of this is happenning in pretty much the same region of the spectrum.
And it's a region, not a single frequency. I bet for many rooms it's more than an octave wide and of course quite vague. That's why I'm pounding on slopes here - you gotta slide your way thru this region, it ain't like flipping a switch.
In a multi-woofer setup you also have some thinking to do about the 'shape' of the sound you can make as you blend between wildly different radiation patterns either side of crossover.
I've done enough playing to know this frequency region powerfully controls the stereo soundstage and the solidity of the central image. You can widen & narrow the 'hole in the middle' effect by playing around here.
Last edited:
I'm not sure I understand where you're coming from. This kind of looks like the result of overthinking.
Have you considered running them overlapping?
Member
Joined 2009
Paid Member
good question!
Seems to me it's not about sound waves but pressure modes within the room once the wavelengths are long enough. I'm not sure any fixed rules will apply because each room can be so different - being a 3 dimensional thing with rooms have different shapes and openings into other spaces to contend with.
Seems to me it's not about sound waves but pressure modes within the room once the wavelengths are long enough. I'm not sure any fixed rules will apply because each room can be so different - being a 3 dimensional thing with rooms have different shapes and openings into other spaces to contend with.
If you accept the Schroeder region as a vaguely defined area of significant bandwidth, and not a single frequency, you are REQUIRED to overlap but with a (hopefully) elegant and managed taper as you transition across this region.
I suspect the key unknown here is "how many octaves wide is a real Schroeder region anyway?". What are the real-world stats for residential listening rooms? If anyone can cite a reference for further reading I'd be grateful.
In thinking further, I can see rooms where the Schroeder band & the transition to mono hearing band do not overlap neatly but are instead offset. This needs further thought.
Here's a cool calculator, my room comes out to 200hz. I crossed around 250hz from a 15 to a 6.5 midrange. I have an easily measured box node at 300hz.
https://www.acoustic.ua/forms/room-modes-eng.html
This also gives other regions that are new to me..
https://www.acoustic.ua/forms/room-modes-eng.html
This also gives other regions that are new to me..
I've heard a rule of thumb ( from pro acoustician in studio field) of 4 octave with center freq behing the so called 'Schroeder freq' ( 2 octave above/below S.Freq).
This is kind of similar to Baffle step behavior of boxed loudspeaker.
Could you clarify this, i don't get your thoughts about this. What do you call 'transition to mono hearing band'?
Last edited:
GM, thanks for posting the Akutek paper, exactly the kind of stuff I'm on about. One of the footnotes out of that paper is this paper which is excellent & on topic https://akutek.info/articles_files/stochastics.htm
AllenB, understood, and those Akutek links are more or less my rant here.
Krivium, four octaves is the widest I've heard so far, a 3-way design might have 2 crossover frequencies in play over an area that wide.
Mono comments - As frequency goes down your hearing transitions from directional to non-directional, from stereo to mono. Widely reported to occur ~80Hz. Also, your ROOM transitions from stereo to mono, according to Schroeder. Above Schroeder sound is a lighthouse; below Schroeder sound is the surf.
AllenB, understood, and those Akutek links are more or less my rant here.
Krivium, four octaves is the widest I've heard so far, a 3-way design might have 2 crossover frequencies in play over an area that wide.
Mono comments - As frequency goes down your hearing transitions from directional to non-directional, from stereo to mono. Widely reported to occur ~80Hz. Also, your ROOM transitions from stereo to mono, according to Schroeder. Above Schroeder sound is a lighthouse; below Schroeder sound is the surf.
Ok thank you for clarification,
I was confused because the way our brain behave and the way the room does are two different things.
80 hz as where the brain loose ability to focus on directionality is an average. I experienced this differently but there was no room involved...
Anyway i observed that in a room the ability to discern directionality is linked to room volume/dimension: the smaller the room, the higher in freq ( it's related to Schroeder frequency we agree).
But i see no offset there: the room dominate, our brain set a lower limit.
It's not an issue in practice: i was ( still is) involved into bass heavy electronic music in 90's/00's where the only real usable media for djing was vinyl. To have consistency in low end in the different playback's environments (clubs, oudoor P.A, cars,... and because of technical limitation in cuting with lathe and head) we often (most always) applied an eliptical filter at mastering stage ( monoing low end). The range of freq at which the elliptical filter was applied ranged from circa 300hz to 80hz.
I bet you won't (or had not ever) notice(d) it*, except their would be a real difference with a not optimal calibrated P.A. or with same track with phase issue in low end ( non monoed)...
In what the fact multiple xover could be involved in such a 'wide' range effect could be an issue? It's not with BSC which often happen higher in freq than Schroeder frequency and span over same bandwidth.
I use 48db/octave FIR xover on my threeways and it doesn't really comes into play with Schroeder freq in my case.
* if you heard rap/hip hop tracks in same period you were exposed to same thing as it was a way to have a 'solid' low end. Tracks from Dre, Snoop,... things like that.
I was confused because the way our brain behave and the way the room does are two different things.
80 hz as where the brain loose ability to focus on directionality is an average. I experienced this differently but there was no room involved...
Anyway i observed that in a room the ability to discern directionality is linked to room volume/dimension: the smaller the room, the higher in freq ( it's related to Schroeder frequency we agree).
But i see no offset there: the room dominate, our brain set a lower limit.
It's not an issue in practice: i was ( still is) involved into bass heavy electronic music in 90's/00's where the only real usable media for djing was vinyl. To have consistency in low end in the different playback's environments (clubs, oudoor P.A, cars,... and because of technical limitation in cuting with lathe and head) we often (most always) applied an eliptical filter at mastering stage ( monoing low end). The range of freq at which the elliptical filter was applied ranged from circa 300hz to 80hz.
I bet you won't (or had not ever) notice(d) it*, except their would be a real difference with a not optimal calibrated P.A. or with same track with phase issue in low end ( non monoed)...
In what the fact multiple xover could be involved in such a 'wide' range effect could be an issue? It's not with BSC which often happen higher in freq than Schroeder frequency and span over same bandwidth.
I use 48db/octave FIR xover on my threeways and it doesn't really comes into play with Schroeder freq in my case.
* if you heard rap/hip hop tracks in same period you were exposed to same thing as it was a way to have a 'solid' low end. Tracks from Dre, Snoop,... things like that.
Last edited:
I'm glad that someone else has surfaced the problems we might see with recordings of the phonograph era (which apparently is still with us). I'm reminded of the transfer function characteristics of elliptical filters used in mono-ing the stereo tracks of music tracks originally mastered before 1983--the year of the Compact Disc--CD--introduction.
In my many experiences undoing some of mastering effects on music tracks (about 25K--30K tracks to date) I've run into many issues around 80-200 Hz to undo, from the standpoint of magnitude response (EQ) and apparent issues with phase response in this area--which is still audible. Below is a screen capture of the full transfer function response of a low-pass elliptical filter for reference:
The magnitude response looks great, but the phase response of elliptical filters really sucks. The phase growth of the passband increases in this 5th order example by ~360 degrees by point that the frequency is achieved that the filter kicks-in. It's not a mystery to me you can hear this sort of artifact in the music, and when trying to smooth this frequency band of obvious notches/dips in magnitude (SPL) response, it becomes clear that this sort of "feature" effectively can't be undone once it has been performed during mastering.
The real point I'd like to make is that, of all the issues that I try to correct in my music tracks via "demastering" using Audacity, etc, the frequency band that we call the Schroeder transition band (about 80-210 Hz for home hifi-sized rooms) is the most affected by mastering practices (outside of the really terrible bass attenuation often found below 100 Hz on stereo CD and phonograph tracks).
Chris
Last edited:
Reading through the two linked papers. I'm now wondering if speaker designs would benefit by having a single device crossed at the low and high ends of the Schroeder region a subwoofer below and tweeter (or mid and tweeter) above. It would be interesting to look back at some JBL monitor designs to see if this matches up.
I greatly appreciate all your thoughtful replies. Here's a bit more on bass really being mono for more than one reason, whether we like it or not.
As frequency goes down, three different systems fade from directional to non-directional. All of us have the first 2 items in play, the 3rd affects sub users only.
1) Your hearing is mono at low frequencies.
Widely reported to occur about 80Hz.
2) Your room is mono at low frequencies.
Above Schroeder sound is a lighthouse; below Schroeder sound is the surf. Endless debate as to where in the spectrum this occurs "at my place".
3) Your stereo system is mono at low frequencies.
I suspect the overwhelming majority of us run 2+ subs in mono? The real-world tunings here are driven by the specifics of your hardware.
As frequency goes down, three different systems fade from directional to non-directional. All of us have the first 2 items in play, the 3rd affects sub users only.
1) Your hearing is mono at low frequencies.
Widely reported to occur about 80Hz.
2) Your room is mono at low frequencies.
Above Schroeder sound is a lighthouse; below Schroeder sound is the surf. Endless debate as to where in the spectrum this occurs "at my place".
3) Your stereo system is mono at low frequencies.
I suspect the overwhelming majority of us run 2+ subs in mono? The real-world tunings here are driven by the specifics of your hardware.