Crossover Levels at 80 or full range? Are you bi-wiring?

[Boomzilla]

What you propose creates a frequency response dip at or below the crossover point, but if you like the effect, go for it.

[RichGuy]

Dec 11, 2016 10:59:02 GMT -5 Boomzilla said:

What you propose creates a frequency response dip at or below the crossover point, but if you like the effect, go for it.

Not at all, it creates a better bass from the other speakers, as well as a better blend with the subwoofer into the system. At 80 hz it is too apparent that the bass sound is coming from the subwoofers location, while a lower crossover to the subwoofer helps the subwoofer drop into the range where low frequencies are not able to be localized by human ear. The range where low frequency cannot be localized by human ear varies among people but at 80 hz it is often apparent to many people and definitely is to me. The bass from the subwoofer should blend into the system not noticeably stand apart from it.

[Boomzilla]

Hi RichGuy -

If you can hear location @ 80 Hz, I believe you. Most others, though, can't.

Wikipedia: For frequencies below 800 Hz, the dimensions of the head (ear distance 21.5 cm, corresponding to an interaural time delay of 625 µs), are smaller than the half wavelength of the sound waves. So the auditory system can determine phase delays between both ears without confusion. . . . As the frequency drops below 80 Hz it becomes difficult or impossible to use either time difference or level difference to determine a sound's lateral source, because the phase difference between the ears becomes too small for a directional evaluation.

Cordially - Boomzilla

[RichGuy]

True many can't hear localization at 80 hz, but others can, I think more likely around 60 hz is a general max for peoples hearing ability but it may be lower.

I prefer to hear more bass at each speaker location as it is directed and not hear (being able to tell) when the bass is coming from the subwoofer, keeping the subwoofer sounding great but invisible to detection.

I know as for 2 channel (actually 2.1 channel) on my system with my XSP-1 my sub blends flawlessly with crossovers set at 50 hz as though the sound was only coming from my front speakers but making them appear to sound better and go lower, when I set the crossover higher I can notice the difference and be able to tell where some bass is coming from the subwoofer and not the front speakers.

[yves]

If the internal crossovers of a speaker are separate from eachother (you can do the 9V battery test to find out... if the woofer moves, you are out of luck...) then bi-wiring can, and does, make some kind of measurable difference to the signal. In some cases this difference is an improvement, but in some other cases it is a deterioration, and, in some other cases still, this difference is not in any way audible.

That being said, personally, I, am not using two regular speaker cable runs to bi-wire, but instead am using inakustik Reference LS-1002 bi-wiring cables to bi-wire. FYI, there is currently another discussion going on about bi-wiring, here: emotivalounge.proboards.com/thread/48795/chuck-hawks-bi-amping-wiring

P.S., setting the crossover point below 80Hz can help to reduce audible distortions, and, while it is true that humans can't localize sounds below 80Hz, the crossover does not stop at 80Hz, but instead rolls off gradually above. Further, sound localization in human listeners is still affected by the loudness of sounds below 80Hz because loud tones can mask out softer ones regardless of frequency, but human hearing also uses timbre recognition for sound localization, and timbre of tones above 80Hz can also perceptually be altered by frequencies below 80Hz so basically things are not that simple.

[Wideawake]

Subwoofers are used to augment the bass frequencies of your audio system. The crossover point, from your main speakers to the subs, is determined by the capability of your main speakers. The xo point will vary for music and HT applications. Then comes phase alignment and room corrections, which I shall not touch upon here.

For music, I would set the xo at about 10dBs higher than the lowest frequency that your mains can produce, if they are not full range (e.g. roll off at 50 - 120 Hz). But, what if your mains are closer to full range and they are capable of playing down to 35Hz (+/- 0db)? Even in such a case, I would set the xo point at about 60 - 70Hz to begin with and adjust that lower based on what sounds best. The reason is that even though your mains may play down to 35Hz, most speakers will sound better in the upper bass and midrange when they are not commissioned to play at their limits in the low bass region.

HT bass management is a completely different animal. You have the bass from all channels and you also have the LFE channel to deal with. The LFE channel may contain frequencies all the way up to 120Hz, at which point you hit a brick wall. That is, you don't have 121Hz coming from the LFE. The sound engineer may roll off the LFE from 80Hz up. To make matters worse you have a 10dB increase in the LFE channel, above the reference of 95dBs. This introduces S/N issues, but since our ears are not so sensitive in this audio region we don't notice the deterioration in S/N too much. This means that the subwoofer needs to produce low bass at 105dBs with peaks of 115dbs. Yikes! Not many subs can do that cleanly and without bottoming out. You can lower the reference volume in your setup to 75dB, which is loud enough for most of us. So, if subs have a difficult time with this then your mains certainly will not be up to snuff to accommodate the LFE channel. That is presuming the the processor allows you to send LFE output to your mains.

For HT, use 80Hz as the xo point and send bass from all channels as well as the LFE to your subs. Then adjust downward as you see fit. Just because the THX standard states an xo at 80Hz, it does not mean that we should follow that exactly. Each room is different and we have speakers with varying capabilities. All of that will need to be taken into account when setting the xo point. Make sure to send bass from all channels to your subs, not just the LFE channel so that your subs are not just sitting idle when no LFE output is present and your speakers will not need to work too hard. If your processor allows it, then let your speakers run full range (if capable) while your sub comes in at 80Hz for bass from all channels as well as the LFE. This is the ideal situation. Note that if you adjust the xo to your sub lower than 80Hz then you could be missing out on some of the LFE output. For that reason, I suggest you keep it at 80 Hz if other variables allow you to do so. If your processor is handling the xo setting then switch off the xo on the sub or set it to max frequency.

Regarding bi-wiring, try it. If you find it makes a difference use it. It doesn't do anything in my system so I don't use it.

[Cogito]

I find myself bypassing the subwoofer completely for my music listening, only turning it on with movie watching. In which case, I set the low-pass filter to 50 Hz while still running the mains at full range for best results. Simply experiment a bit. No two systems/rooms are going to behave the same.

As for bi-wiring... I couldn't point to a single definitive advantage to doing so. But again, experiment.

[leonski]

Dec 11, 2016 10:59:02 GMT -5 Boomzilla said:

What you propose creates a frequency response dip at or below the crossover point, but if you like the effect, go for it.

Not necessarilly true.

I LOW CUT my main speakers at about 50hz.   They are good to 40 or perhaps a Little lower.   Mid-30s is certainly IT for them.
I use the subs crossover at a LOWER frequency.  
Everyone agrees, I hope, that crossover frequencies are NOT a brick wall.   Response exists past the crossover frequency and diminishes at whatever the 
slope is PLUS any natural rolloff of the driver. 

Somebody might say 'you have a dip between speakers'.   But this isn't true.   While ELECTRICALLY there is a dip, the idea is for the speakers to SUM FLAT
thru the crossover region.  

Just my opinion, and I'm NOT a HT guy, but I'd figure some way to have only the sub below a certain cutoff, and lower than 80hz, at that.    Sub and mains at the 
same time will, IMO have the potential to 'muddy' the output, cause phase problems and even effect image, if that's important in an effects laden movie.  
Allow the sub to roll-off some before the mains take over.   Allow the mains to roll-off some before the sub takes over.    Simple.   

[leonski]

Dec 11, 2016 10:59:02 GMT -5 Boomzilla said:

What you propose creates a frequency response dip at or below the crossover point, but if you like the effect, go for it.

IF, (biggie) I understood the previous posts, what was said was to cross over and leave a gap between sub and mains.
I do exactly that.    I cross mains (12db/oct) at maybe 50 to 55hz high pass.    I cross sub (24db/oct) at 40hz or so.
The gap is nicely filled in since of course, crossovers are NOT brickwall.   
The idea is to SUM FLAT thru the crossover region.  

[Boomzilla]

Sum flat is EXACTLY the goal.

With 6dB / octave crossovers, called "first-order" crossovers (not at all common in AVRs or in typical electronic crossovers), - Having the SAME crossover frequency AND the top/bottom speakers in-phase with each other produces (in each driver) a -3dB dip at EXACTLY the crossover point. The top/bottom then acoustically sum to flat frequency response AT the crossover frequency.

With 12dB / octave crossovers, called "second-order" crossovers (common in most AVRs and in typical electronic crossovers), - Having the SAME crossover frequency AND the top/bottom speakers OUT-of-phase with each other produces a flat acoustic output at the crossover frequency.

With 18dB / octave crossovers, called "third-order" crossovers (not common anywhere), - Having the SAME crossover frequency AND the top/bottom speakers in-phase with each other produces a flat acoustic output at the crossover frequency.

Now the above statements apply in theory. If either the high or low pass drivers have an inherent ACOUSTICAL roll-off within one to one-and-a-half octaves of the actual crossover point, then the acoustic and the electrical attenuations sum to produce a steeper slope than expected. Then other adjustments must be made to achieve flat response at the crossover frequency. Common solutions include:

1. Adjusting the crossover order for one of the two drivers
2. Adjusting the phase of one of the two drivers
3. Modifying the crossover frequencies for the drivers individually
4. Allowing either a "gap" or an "overlap" as needed between the drivers to achieve flat response at the crossover frequencies

So in short, crossover design is a science, but in the real world, also an art. The most skilled designers use measurement tools to "tweak" their crossovers to achieve the "voicing" that they desire from the drivers.

Although this SEEMS like a difficult process, listening can substitute for the measurements. In fact, EVERY ONE of those "skilled designers" I mentioned tweaks their final design by ear. If there's a discrepancy between what the theories/measurements predict, and what the designer hears, the "by ear" tuning is what gets selected (not some of the time - not most of the time - but ALWAYS). Keep this in mind when some internet gomer tells you that acoustics is pure science, and that your ears lie to you.

So in summary - No, crossovers are NEVER "brick-wall." All involve slopes, and the goal IS for the drivers to sum to ACOUSTICALLY flat through the crossover region (whether or not the electrical measurements seem flat).

Cheers - Boomzilla

[novisnick]

Feb 18, 2017 11:30:13 GMT -5 Boomzilla said:

Sum flat is EXACTLY the goal.

With 6dB / octave crossovers, called "first-order" crossovers (not at all common in AVRs or in typical electronic crossovers), - Having the SAME crossover frequency AND the top/bottom speakers in-phase with each other produces (in each driver) a -3dB dip at EXACTLY the crossover point. The top/bottom then acoustically sum to flat frequency response AT the crossover frequency.

With 12dB / octave crossovers, called "second-order" crossovers (common in most AVRs and in typical electronic crossovers), - Having the SAME crossover frequency AND the top/bottom speakers OUT-of-phase with each other produces a flat acoustic output at the crossover frequency.

With 18dB / octave crossovers, called "third-order" crossovers (not common anywhere), - Having the SAME crossover frequency AND the top/bottom speakers in-phase with each other produces a flat acoustic output at the crossover frequency.

Now the above statements apply in theory. If either the high or low pass drivers have an inherent ACOUSTICAL roll-off within one to one-and-a-half octaves of the actual crossover point, then the acoustic and the electrical attenuations sum to produce a steeper slope than expected. Then other adjustments must be made to achieve flat response at the crossover frequency. Common solutions include:

1. Adjusting the crossover order for one of the two drivers
2. Adjusting the phase of one of the two drivers
3. Modifying the crossover frequencies for the drivers individually
4. Allowing either a "gap" or an "overlap" as needed between the drivers to achieve flat response at the crossover frequencies

So in short, crossover design is a science, but in the real world, also an art. The most skilled designers use measurement tools to "tweak" their crossovers to achieve the "voicing" that they desire from the drivers.

Although this SEEMS like a difficult process, listening can substitute for the measurements. In fact, EVERY ONE of those "skilled designers" I mentioned tweaks their final design by ear. If there's a discrepancy between what the theories/measurements predict, and what the designer hears, the "by ear" tuning is what gets selected (not some of the time - not most of the time - but ALWAYS). Keep this in mind when some internet gomer tells you that acoustics is pure science, and that your ears lie to you.

So in summary - No, crossovers are NEVER "brick-wall." All involve slopes, and the goal IS for the drivers to sum to ACOUSTICALLY flat through the crossover region (whether or not the electrical measurements seem flat).

Cheers - Boomzilla

Very well written and thought out but in short, hearing IS believing!   
Thanks for the lesson.

[Boomzilla]

My "lessons" are worth every penny you pay for them! LOL

[leonski]

Feb 18, 2017 11:30:13 GMT -5 Boomzilla said:

Sum flat is EXACTLY the goal.

With 6dB / octave crossovers, called "first-order" crossovers (not at all common in AVRs or in typical electronic crossovers), - Having the SAME crossover frequency AND the top/bottom speakers in-phase with each other produces (in each driver) a -3dB dip at EXACTLY the crossover point. The top/bottom then acoustically sum to flat frequency response AT the crossover frequency.

With 12dB / octave crossovers, called "second-order" crossovers (common in most AVRs and in typical electronic crossovers), - Having the SAME crossover frequency AND the top/bottom speakers OUT-of-phase with each other produces a flat acoustic output at the crossover frequency.

With 18dB / octave crossovers, called "third-order" crossovers (not common anywhere), - Having the SAME crossover frequency AND the top/bottom speakers in-phase with each other produces a flat acoustic output at the crossover frequency.

Now the above statements apply in theory. If either the high or low pass drivers have an inherent ACOUSTICAL roll-off within one to one-and-a-half octaves of the actual crossover point, then the acoustic and the electrical attenuations sum to produce a steeper slope than expected. Then other adjustments must be made to achieve flat response at the crossover frequency. Common solutions include:

1. Adjusting the crossover order for one of the two drivers
2. Adjusting the phase of one of the two drivers
3. Modifying the crossover frequencies for the drivers individually
4. Allowing either a "gap" or an "overlap" as needed between the drivers to achieve flat response at the crossover frequencies

So in short, crossover design is a science, but in the real world, also an art. The most skilled designers use measurement tools to "tweak" their crossovers to achieve the "voicing" that they desire from the drivers.

Although this SEEMS like a difficult process, listening can substitute for the measurements. In fact, EVERY ONE of those "skilled designers" I mentioned tweaks their final design by ear. If there's a discrepancy between what the theories/measurements predict, and what the designer hears, the "by ear" tuning is what gets selected (not some of the time - not most of the time - but ALWAYS). Keep this in mind when some internet gomer tells you that acoustics is pure science, and that your ears lie to you.

So in summary - No, crossovers are NEVER "brick-wall." All involve slopes, and the goal IS for the drivers to sum to ACOUSTICALLY flat through the crossover region (whether or not the electrical measurements seem flat).

Cheers - Boomzilla

Even my panels use this idea.     While the HF/LF 'curve' crosses at 600hz, the KNEE frequencies are above And below that point.   LF is a 12db/oct (2nd order) while the HF is a simple cap in series or '1st order'.   
The 2 sections of the panel are 90' out of phase and wired 'backwards' electrically.    The net effect is a 'lobe' effect which renderes the panel Not perfectly symetrical front-to-back.

I'm not going to the trouble, but when Stereophile review this speaker, they noted the flat response Thru the crossover region.  

It must be noted that FIR Filters have NO phase shift thru the passband and can be easily (sure!) implimented using DSP.

dspguru.com/dsp/faqs/fir/basics

I'm not sure the linked article will help.