Crossover with different slopes lowpass and highpass.

[Musician]

If I could wish - I would like switchable lowpass and highpass crossover between 36 dB butterworth, 24 linkwitz riley, 18dB butterworth, 12 dB butterworth, 12 dB linkwitz . I also wish that lowpass and highpass crossover could have different slopes to combine , so a closed box main speaker can be crossed 12dB/oct high pass ( making it a acoustical 24 dB/oct ) and combined with 24dB/oct low pass for the subwoofers. This would sum be a correct 24/24 dB acoustical crossover LP and HP, - if one uses closed box main speakers. This is called a THX crossover as you all know, and all THX AVR:s have this option.

Best integration for two channel music would infact be 30dB/oct low pass combined with 18dB/oct high pass for closed mains - so I would want that option to. A 30/30dB acoustical slope crossover. This gives even better power response in the room than the THX crossover. Not many manufacturers know this.

No streamer on earth has those options, and if the ultra would have this, it would be a sensation.

 

[harkpabst]

Your figures assume that the high pass filter is always aligned with the "natural" roll-off of the main speakers.

The biggest advantage of a bass management that controls both, the low pass for the sub and the high pass for the mains is that we no longer have to chase after this arrangement. Instead we can use a high crossover frequency, which in most cases is beneficial, as long as the slope is high enough.

The in-room response of a speaker is hardly ever the same as theory or anechoic measurements may suggest. The lower corner frequency will be lower in-room and there won't be a nice and smooth 12 dB/Oct slope for closed speakers.

Havin said that, there's nothing wrong with having full control over independent crossover frequencies and filter slopes, once everything else is sorted.

A 30/30dB acoustical slope crossover. This gives even better power response in the room than the THX crossover. Not many manufacturers know this.

It's not only manufacturers who don't know this. I'm sure you're more than prepared to link to some sources.

 

[Smartplug]

If I could wish - I would like switchable lowpass and highpass crossover between 36 dB butterworth, 24 linkwitz riley, 18dB butterworth, 12 dB butterworth, 12 dB linkwitz . I also wish that lowpass and highpass crossover could have different slopes to combine , so a closed box main speaker can be crossed 12dB/oct high pass ( making it a acoustical 24 dB/oct ) and combined with 24dB/oct low pass for the subwoofers. This would sum be a correct 24/24 dB acoustical crossover LP and HP, - if one uses closed box main speakers. This is called a THX crossover as you all know, and all THX AVR:s have this option.

Best integration for two channel music would infact be 30dB/oct low pass combined with 18dB/oct high pass for closed mains - so I would want that option to. A 30/30dB acoustical slope crossover. This gives even better power response in the room than the THX crossover. Not many manufacturers know this.

No streamer on earth has those options, and if the ultra would have this, it would be a sensation.

That was the feature I always wanted to have. I had mini dsp and the amount of customization was incredible. The integration between main and sub was top class. Now, if wiim can make such feature similar to mini dsp it would be ground breaking feature.

 

[Musician]

Your figures assume that the high pass filter is always aligned with the "natural" roll-off of the main speakers.

The biggest advantage of a bass management that controls both, the low pass for the sub and the high pass for the mains is that we no longer have to chase after this arrangement. Instead we can use a high crossover frequency, which in most cases is beneficial, as long as the slope is high enough.

The in-room response of a speaker is hardly ever the same as theory or anechoic measurements may suggest. The lower corner frequency will be lower in-room and there won't be a nice and smooth 12 dB/Oct slope for closed speakers.

Havin said that, there's nothing wrong with having full control over independent crossover frequencies and filter slopes, once everything else is sorted.


It's not only manufacturers who don't know this. I'm sure you're more than prepared to link to some sources.

Here you are :

faktiskt.io • Visa tråd - Hur brant ska man dela topp / baslådor

Its the genious constructor of Guru audio that has come to this conclusion, read his posts ( Ingoehman ).He also sells a crossover with 30/18dB slopes, Ino cr80.
This is backed up by science , an odd order crossover has better power response in the room than an even order, such as THX 24/12dB, meaning better bass impact.

Kaffekoppen skrevet finns väl ingen direkt fördel med udda ordningens filter....enda diffen är att de möts vid -6db istället för -3db....

Njae...

För det första möts udda ordningens funktioner helst med -3 dB, medan det är jämna som möts vid - 6 dB (om man eftersträvar en tonkurverak addition) men framförallt så kan det vara en stor fördel med det udda ordningens filtret, eftersom det möjliggör två saker:

1. Man kan få BÅDE hyfsat rak tonkurvan för direktljudet och för energikurvan. Med jämna ordningens övergångsfunktioner går det bara om basmodulen står i toppsystemets absoluta närhet (vilket betyder max 0,5 meter mellan dem bas-bas).

2. Det är en tillgäng således att det kan användas för att kmpensera stereosystemfelen i lågrfrekvensområdet, således att man kan nyttja de loober som upptår till att hjälpa den bortre och stjälpa den närmare kanalen, vid snedplacerad lyssnare.

Både de ovanstående poängerna kräver dock att man tar hjälp av både noggranna beräkningar och mätningar om det skall finna något hopp att hitta rätt. Att arbeta i blindo är nästan omöjligt när det gäller så här komplicerade saker - vilket är ett argument för att man väljer den lättare lösningen - THX-delning!

Kaffekoppen skrev:18db är för flackt (dät hör jag 160hz/-18db, och 30db kan orsaka andra problem. Att lösa dem är dyrt.

1. Vilka problem?

2. Vad är det som kostar för att lösa dem?

3. Hur går det till att lösa dem?

Kaffekoppen skrev:Mig veterliggen känner jag inte till många som kan höra 160hz/ -24db ... skulle vilja påstå attt jag inte känner någon om det samtidigt spelas musik dynamiskt från huvudhögtalarna i de frekvensområdet.

Det beror väl på en förfärlig massa förutsättningar?

Kaffekoppen skrev:Märk väl att NHT 3.3 hade en 12" bas delad vid 120hz/12db spelandes åt sidan. Det innebär att den faktiskt spelade med upp till 240hz/-24 db rakt åt sidan... utan att det reflekterades på det. Jorå, det var järnkärnespole

Ja, där har du just en sådan där förutsättning som sätter tumregelsvaren ur funktion. I den högtalaren satt sakerna ihop. Riktningen blev i stort sett identisk, och dessutom var fasvillkoren under kontroll tack vare att det var en färdig konstruktion, til skillnad från två lösa delar vars samarbete kan bli väldigt besvärligt (kombinernade av slumpvis vald högtalare med slumpvs vald subwoofer).

I en dedicerad konstruktion ser jag inga hinder för att använda filter som skär 6 dB/oktav, om effekttålighet/distorsion inte är något problem och elementens natruliga frekvensgångsegenskaper tillåter det.

Som sagt - det är komplicerat, och tumregelsvar förslår inte. Skall man sammanfatta problemet att dela mellan ett par toppsystem och basmoduler och förutsättningarna är att de inte står i gemensam punkt i rummet och delningsfrekvensen är vald till 80 Hz, är det för det mesta så att man landar på en slutgiltig akustisk delning med en branthet om 30 dB/oktav. Ungefär alltså. Man bör dock minnas att det är mötningspunktens nivåer och fasegenskaper, samt fasutvecklingen utanför delningsfrekvensen som är av störst betydelse, inte filterbrantheten som sådan, så länge funktionen är rimligt brant.


Vh, iö

 

[Musician]

In English :

Kaffekoppen wrote:
There is no direct advantage with odd-order filters....the only difference is that they meet at -6db instead of -3db....
———-
Ingoehman: Well...

First of all, odd-order functions preferably meet at -3 dB, while even ones meet at -6 dB (if you are aiming for a straight tone curve addition) but above all, there can be a great advantage with the odd-order filter, since it enables two things:

1. You can get BOTH a fairly straight tone curve for the direct sound and for the energy curve. With even-order transition functions, this is only possible if the bass module is in the absolute vicinity of the top system (which means a maximum of 0.5 meters between them bass-bass).

2. It is an advantage that it can be used to compensate for stereo system errors in the low frequency range, so that you can use the loops that occur to help the far and overturn the closer channel, when the listener is placed obliquely.

However, both of the above points require the help of both careful calculations and measurements if there is to be any hope of finding the right one. Working blindly is almost impossible when it comes to such complicated things - which is an argument for choosing the easier solution - THX sharing!
————-

Kaffekoppen wrote:
18db is too flat (I hear 160hz/-18db, and 30db can cause other problems. Solving them is expensive.

Ingoehman:
1. What problems?

2. What does it cost to solve them?

3. How do you solve them?
————-

Kaffekoppen wrote:
As far as I know, I don't know many people who can hear 160hz/-24db ... I would like to say that I don't know anyone if music is simultaneously played dynamically from the main speakers in that frequency range.

Ingoehman:
It depends on an awful lot of conditions, right?
———

Kaffekoppen wrote:
Note that the NHT 3.3 had a 12" bass divided at 120hz/12db playing sideways. This means that it actually played up to 240hz/-24 db straight sideways... without it being reflected. Oh my, it was an iron core coil

Ingoehman: Yes, there you have it exactly the kind of condition that puts the rule of thumb responses out of action. In that speaker, things were put together. The directionality was largely identical, and in addition the phase conditions were under control thanks to the fact that it was a finished construction, unlike two loose parts whose cooperation can be very difficult (combined by a randomly selected speaker with a randomly selected subwoofer).

In a dedicated construction, I see no obstacle to using filters that cut 6 dB/octave, if power tolerance/distortion is not a problem and the natural frequency response characteristics of the elements allow it.

As I said - it is complicated, and rule of thumb responses are not enough. If one were to summarize the problem of dividing between a pair of top systems and bass modules and the conditions are that they are not at the same point in the room and the crossover frequency is chosen to 80 Hz, it is most of the time that one ends up with a final acoustic crossover with a steepness of 30 dB/octave. Approximately so. However, it should be remembered that it is the levels and phase characteristics of the intersection point, as well as the phase evolution outside the crossover frequency that are of greatest importance, not the filter steepness as such, as long as the function is reasonably steep.

Vh, iö

 

[Musician]

Your figures assume that the high pass filter is always aligned with the "natural" roll-off of the main speakers.

The biggest advantage of a bass management that controls both, the low pass for the sub and the high pass for the mains is that we no longer have to chase after this arrangement. Instead we can use a high crossover frequency, which in most cases is beneficial, as long as the slope is high enough.

The in-room response of a speaker is hardly ever the same as theory or anechoic measurements may suggest. The lower corner frequency will be lower in-room and there won't be a nice and smooth 12 dB/Oct slope for closed speakers.

Havin said that, there's nothing wrong with having full control over independent crossover frequencies and filter slopes, once everything else is sorted.


It's not only manufacturers who don't know this. I'm sure you're more than prepared to link to some sources.

Agree, its always the acoustical crossover slopes that matters, thats why THX crossovers 24/12 (electrical) is a rather good option because a closed box falls 12/dB oct below resonanse , without any crossover, and much less if the room has gain .

But the best sounding crossover slope below 100 Hz is 30/30 , acoustical . Often 30/18 electrical. This gives slightly better impact than an even order crossover like THX. 18/6dB electrical slope is also often better than THX 24/12dB if one uses two subwoofers in stereo near each speaker.
Why?

Odd order crossovers ( acoustical ) gives slightly better impact in the bass because of better power response in the room.

 

[harkpabst]

I'll have a closer look into it at some later point. These short excerpts don't give much reasoning behind the statements.

My point remains: The biggest advantage to subwoofer integration stems from being able to set the crossover frequency and slope in a way that keeps it mostly unaffected by the speakers' acoustical slope.

For this to work the three basic guidelines are:

• Keep the crossover frequency well above the speakers' lower border.
• Keep the crossover frequency well below the subwoofer's upper border.
• Use steepsymmeteical filters (like LR4).

Again, more flexibility is fine, but it requires quite some expertise to identify how and when to make use of it for (sometimes) even better results.

 

[jed1]

All these filters / crossover types are IIR filter and can be easily implemented with combination of biquads. Depending on the order of a filter, number of biquads are determined. Existing peak and shelving filter take 1 set of biquad. Second order filter needs two set of biquads. So, limit on number of these filters depends on WiiM's DSP capacity. For example, if I need a third order filter and a second order filter, I have 5 peak / shelving filters left.

 

[canard]

I'll have a closer look into it at some later point. These short excerpts don't give much reasoning behind the statements.

My point remains: The biggest advantage to subwoofer integration stems from being able to set the crossover frequency and slope in a way that keeps it mostly unaffected by the speakers' acoustical slope.

For this to work the three basic guidelines are:

• Keep the crossover frequency well above the speakers' lower border.
• Keep the crossover frequency well below the subwoofer's upper border.
• Use steepsymmeteical filters (like LR4).

Again, more flexibility is fine, but it requires quite some expertise to identify how and when to make use of it for (sometimes) even better results.

For "wiim" users who are not necessarily experts, the use of a rather simple but reliable solution... your approach seems appropriate to me... but there remains the time /phase adjustment etc... and that seems to have progressed little (?)

(and for futur..two sub..l/r ;-) )

 

[Musician]

I'll have a closer look into it at some later point. These short excerpts don't give much reasoning behind the statements.

My point remains: The biggest advantage to subwoofer integration stems from being able to set the crossover frequency and slope in a way that keeps it mostly unaffected by the speakers' acoustical slope.

For this to work the three basic guidelines are:

• Keep the crossover frequency well above the speakers' lower border.
• Keep the crossover frequency well below the subwoofer's upper border.
• Use steepsymmeteical filters (like LR4).

Again, more flexibility is fine, but it requires quite some expertise to identify how and when to make use of it for (sometimes) even better results.

LR 24 HP and LP is a very good option especially with crossovers above 300 Hz. When I made my Hybrid active dsp speaker I kept coming back to fourth order crossover between tweeter and midbass, as the best sounding option.

 

[harkpabst]

For "wiim" users who are not necessarily experts, the use of a rather simple but reliable solution... your approach seems appropriate to me... but there remains the time /phase adjustment etc... and that seems to have progressed little (?)

That's a problem, indeed. Of course, users can (and currently have to) reverse the polarity manually and then adjust the delay between mains and sub in 1 ms steps. Doing this purely by listening is both, cumbersome and error prone. Fully automatic procedures have their own troubles. It would be nice if we could get a general frequency response measurement option.

and for futur..two sub..l/r ;-)

I won't vote against that.

LR 24 HP and LP is a very good option especially with crossovers above 300 Hz. When I made my Hybrid active dsp speaker I kept coming back to fourth order crossover between tweeter and midbass, as the best sounding option.

View attachment 20998

Nice looking.

 

[canard]

That's a problem, indeed. Of course, users can (and currently have to) reverse the polarity manually and then adjust the delay between mains and sub in 1 ms steps. Doing this purely by listening is both, cumbersome and error prone. Fully automatic procedures have their own troubles. It would be nice if we could get a general frequency ..

being able to observe a raw measurement (with or without peq upstream) precisly, is really expected, without rc mode etc.. it has really been expected for a while... (see being able to export it in formats like rew)...

 

[MrA]

In my system (open baffle speakers with sealed subs) I have the option to use digital Linear Phase filters all the way to 12th order (72dB slope). Although 8th or 10th order filters do tend to result in a slightly 'warmer' overall bass, using 12th order crossed at 100hz (sub) and 80hz (mains) provides the tightest and cleanest overall bass with very realistic transient attack.

 

[harkpabst]

Although 8th or 10th order filters do tend to result in a slightly 'warmer' overall bass, using 12th order crossed at 100hz (sub) and 80hz (mains) provides the tightest and cleanest overall bass with very realistic transient attack.

You don't advertise this exact setup as a general recommendation when integration subs and mains, irrespective of speakers and rooms? Or do you?

 

[MrA]

You don't advertise this exact setup as a general recommendation when integration subs and mains, irrespective of speakers and rooms? Or do you?

No, just my own system & this was after a lot of measurements and listening. Once the crossovers & phase alignment were finalised, time alignment between the various speakers and subs and a small amount of bass eq were the last adjustments.

 

[canard]

rappel
small deviation...
in relation to the discussions on the possibility of low cuts, without relation to the integration of a subwoofer... proposing in the peq mode or outside peq ( in the main audio settings) a strong high pass num, adjustment, on the entire wiim range( or vibelink?), would be welcome
(This is a very "general public" testing possibility... ;-) )
*

;-)

(ps from the first days of the arrival of the crossover possibility was asked the possibility of enriching with some additional types of classic slopes ... but clearly it was not to offer overly complicated approaches aimed at "experts", who are not really the target of wiim products ... thinking and especially achieving general public solutions concerning audio functions is already a small challenge .. achieving in a qualitative way especially is a serious one)

*" si son usage ne s entend pas..c est qu il est utile... rigolo non? ;-)"