[usil]

I am an amateur DIY so maybe this observation and question is based on my ignorance but I have always wondered how a single speaker can produce accurately - different frequencies at the same time. When we listen to music there is no sequential horn followed by a drum followed by a violin note but rather all frequencies are all playing at the same time.


For a speaker - all these frequencies must compete 'at the same time' to be created and produced. From a logical standpoint, if a low drum note is playing at the same time as a mid-range string instrument and also at the same time as a high note from a flute - the end result from a speaker will be a compromise and certainly will not duplicate all three sound frequencies at the exact same time with equal accuracy. There must be conflicts within the speaker. Further, I just used an example using three instruments but in the real world, there are many more instruments and frequencies played at the same time which further compounds and compromises the speaker to be able to play all these frequencies accurately at the same time they are created by the orchestra.


So, my question is this - How does one speaker do this without compromising the accuracy of all frequencies played at the same time?


This leads to my second observation and question. When we design 2-way and 3-way speakers I understand it is to select the 'best' part of each speaker range for creating the sounds. But does this not 'also' address the compromise I described above when multiple frequencies compete to play from a single speaker at the same time? Two speakers or three speakers would have fewer compromises because one speaker can be playing a low note while a mid-range note is played by another speaker thus eliminating the compromise you would get from playing both notes at the same time from a single speaker.


If what I have surmised above has some truth in it then why not a 4-way, 5 way, 6-way or 7-way speaker system? If you have the span of frequencies being created by 6 or 7 different speakers would there not be even fewer compromises to be made? Would the internal conflict of creating the same frequencies at the same time not be lessened and as a result create more accurate music?


I know that it may be difficult to create a 7-way cross over or maybe not I am not an expert on this. The costs would certainly be expanded but if we eliminate the discussion of costs and consider the electronic difficulties as simply engineering issues that can be surmounted then - for an accurate and great sounding system - would not what I have suggested create a better sound system when compared to a 2-way or 3-way system?


And finally, has anyone else tried to design a 6-way or 7-way system?


I am interested in hearing a discussion on this from the experts.


Usil

 

[thylantyr]

Why not bypass all this and just go to a good concert

and listen to live instruments ?

else, you won't really

duplicate that experience, you can only find a compromise.


I prefer 4 way loudspeakers; tweeter, midrange, woofer, subwoofer

to alleviate the problems you mentioned. If you choose good drivers

you don't need a 5 way and beyond.


You can boost SPL levels and lower distortion by clever sound

system design.

 

[Scott Simonian]

A 7-way crossover. Wow. Just trying to imagine what a passive network like that would look like.

I think that something like that would do more damage than good. Whew, a 5+ way x-over would decimate the original signal. Better to use a full range driver and have a preamp filter out the bass to a subwoofer. If you want lots of SPL doing that you would have to build something like Thy mentioned. Possibly hornloaded or in an array.

 

[LHD21]



Quote:
Originally Posted by Scott Simonian /forum/post/15497858


A 7-way crossover. Wow. Just trying to imagine what a passive network like that would look like.

I think that something like that would do more damage than good. Whew, a 5+ way x-over would decimate the original signal. Better to use a full range driver and have a preamp filter out the bass to a subwoofer. If you want lots of SPL doing that you would have to build something like Thy mentioned. Possibly hornloaded or in an array.

What if you used some sort of active crossover?


Interesting topic.

 

[Scott Simonian]



Quote:
Originally Posted by LHD21 /forum/post/15498266


What if you used some sort of active crossover?


Interesting topic.

An active crossover would work. A bit confusing but it could work. Also, I've never ever seen an active Xover that does more than 4-way. You would need an array of at least one, two or more xovers per channel. Pretty wasteful, just think about how many amp channels you would need. If it could be done right I'd like to see (hear) it.

 

[LHD21]



Quote:
Originally Posted by Scott Simonian /forum/post/15498380


An active crossover would work. A bit confusing but it could work. Also, I've never ever seen an active Xover that does more than 4-way. You would need an array of at least one, two or more xovers per channel. Pretty wasteful, just think about how many amp channels you would need. If it could be done right I'd like to see (hear) it.

I'm thinking more of an abstract unobtanium crossover. Not something real. Basically eliminate complexity of crossover from his statements above.

 

[Scott Simonian]



Quote:
Originally Posted by LHD21 /forum/post/15498446


I'm thinking more of an abstract unobtanium crossover. Not something real. Basically eliminate complexity of crossover from his statements above.

Fair enough. All you have to worry about then is the HUGE baffle full of different drivers and trying to get them all to integrate at the listening position. I love the idea but I remember asking about something like this when I first got into this hobby yeeeaaarrsssss ago. Too complicated to work realistically.


3-way speakers are hard enough to work on.

 

[LHD21]

Would it not be the most badassed line array ever?

 

[Scott Simonian]



Quote:
Originally Posted by LHD21 /forum/post/15498592


Would it not be the most badassed line array ever?

Ahahaha! Definitely! I'd love to see something like this. I just hope it would sound good too. Sure would be HUGE. Just try and picture a 5.1 system like this. Ahahah! Better yet, try 7.1 with multiple subwoofers too.

Awesome.

 

[usil]

I see responses discussing the electronics issues and that is good as I would really like to see what such a system would actually need to exist. Are you sure that a passive crossover system could not be made to work?


And - back to the original question -

Quote:
...multiple frequencies compete to play from a single speaker at the same time...

Is this not the case and is the premise of designing more speakers to play fewer frequencies solid to get an optimal speaker design.


We often build speakers to save money and if you have the tools sometimes this is possible. Then we also build speakers that cost the same as good ones just because we just like DIY projects. The holy grail is going DIY to build something better than anything you can buy or may even be available. Thinking outside of the box - as I propose - may possibly do this. I would think that this has peaked some interest in the group.


As I said, not letting the cost factory hinder the thought process at this point in the discussion (we are brainstorming and will save cost containment as a later discussion) can we come up with a 6-way or 7-way speaker meeting these parameters? I am not talking about an array as this does no address the issue I have described at all.


If this is given serious consideration, my premise is sound and the costs are not exorbitant, I would think that someone might consider building this.


Side note: I do not see expensive speakers of 7 different types or designs being used in this. Instead, one speaker design, say the RS52AN Dayton Dome mid-range might be used 2 times at different mid-range frequency ranges. One tweeter design might be used twice at two upper frequency ranges and then add a woofer at 2 separate lower frequency ranges. The sub woofer would be separate based on its case design needs.


So the crossover is the key item to design. Can this be done with current software?


Usil

 

[Krawdad]

Interesting idea but I don't really see the point?


Quality drivers can play multi-frequency signals with equal amplitude( flat response) throughout the audible range already. So this seems like a solution without a problem.


But to play along.


If you divided the spectrum evenly from 20hz to 20Khz then each driver would need to do ~2800hz frequency range.


Well this might be OK on the top end it would probably not give the low end authority we would want from such a beast. so we could just use a subwoofer from 20hz to say 80hz. the remaining 6 drivers would need to reproduce from 80hz to 20khz.


This would leave the remaining drivers the responsibility of reproducing a frequency range of ~3300 hz each. With current drivers this may be an issue. as many woofers that reach 80hz have trouble at the next crossover frequency of ~4000hz. Maybe splitting the difference would work. But it would probably be better to have exponential increments all the way up to 20khz. If the crossover points were exponential we could go:


Driver 1- 20hz - 80hz

Driver2 - 80hz - 600hz

Driver3 - 600hz - 1.25hz

Driver4 - 1.25hz - 2.5khz

Driver5 - 2.5khz to 5Khz

Driver6 - 5khz - 10khz

Driver7 - 10Khz-20khz


It could be doable but it would not be pretty.

 

[LINEARX]

You have proposed a concept . . . . do your homework, gather the pieces, produce and refine the product and then show us a 6-7 way loudspeaker is better than the current 2 or 3 way products we have today.

Yes, a single driver can produce a multitude of frequencies at the same time. Is there distortion playing multiple frequencies at the same time? Yes. How much? Not much. When we listen to high end systems our perception tells us that what we hear is close to the original even though we know the original has been distorted by the reproduction process.

 

[catapult]



Quote:
For a speaker - all these frequencies must compete 'at the same time' to be created and produced. From a logical standpoint, if a low drum note is playing at the same time as a mid-range string instrument and also at the same time as a high note from a flute - the end result from a speaker will be a compromise and certainly will not duplicate all three sound frequencies at the exact same time with equal accuracy.

Your premise is based on a flawed mental image of how sounds combine. Maybe this pic will help. It's what you would see on an oscilloscope when playing some tones. Red is a low frequency tone. Blue is a high frequency tone. Magenta is what you get when the two tones combine. A speaker can play the combined magenta tone just as well as it can play either the red or blue tones.

There are limits of course. Drivers have a range in which they work best so we cross to other drivers if the frequency gets too high or too low for the driver to handle. But crossovers introduce their own problems. Without getting into a bunch of technical stuff, you'd generally like each driver to handle at least 2 octaves and 3 is better to minimize crossover interactions. That means something in the 3-4 drivers range to cover 20-20K is about optimum to keep each driver working in its best range and keep the crossovers from stepping all over each other.

 

[smokinghot]



Quote:
Originally Posted by catapult /forum/post/15504473


Your premise is based on a flawed mental image of how sounds combine. Maybe this pic will help. It's what you would see on an oscilloscope when playing some tones. Red is a low frequency tone. Blue is a high frequency tone. Magenta is what you get when the two tones combine. A speaker can play the combined magenta tone just as well as it can play either the red or blue tones.

I think you proved the OP's point instead of showing his flawed concept.


If I understand you pic, and what your saying. The magenta wave is what happens when both high and low frequencies are being produce at the same time, by the same speaker. The highs aren't as high as they should be, and the lows aren't as low. So the OP's concept of multiple drivers would allow a more accurate reproduction of the blue and red waves. The magenta, (or the combined frequencies) which isn't apart of the material being reproduced, is no longer there.


Seems to make sense.

 

[usil]

Linearx: You are right, in the end, an actual test would be needed to compare the results. A standard 3 way could be built and then a 6 or 7-way using the approach explained. Both could then be compared side by side.


Smokinghot: I think you are right and this would be part of the purpose of the design i.e., getting all of the highs and all of the lows with reduced interference. Granted, there is a limit to how far you want to go and a 6 or 7-way while it is an arbitrary selection - it is within build limits to do a comparison.


From all of our experiences - the cones of speakers have a remarkable ability to create multiple sounds at the same time. The question is - at what expense and what is lost or effected by the same cone competing to produce different frequencies at the same time.


What if the 7-way opens up the music, allows sounds you have not heard before with a clarity not previously detected. The brain plays a part in the experience. It has the ability to fill in the holes. It does this with sight as we watch a 24 frame per second movie but detect that the 60 frame per second imparts a smoother and 'better' visual experience. What if the brain also compensates for audio? While we are all used to listening to 2 and 3 way speakers which we consider normal - maybe we might detect a better and more accurate audio experience with a 7-way.


Some people spend tens of thousands of dollars to get a better audio experience. What if it turns out to be easier to do for a lot less money with this design concept?


Usil

 

[thylantyr]

Did your idea of a 6 or 7 way loudspeaker manifest because you haven't experienced a good 4 way? It's no

surprise here as commercial loudspeakers are all mediocre and the common DIY project is just that, a

common speaker to satisfy the common consumer. One could easily say these designs aren't adequate for

higher SPL, lower distortion playback that usually is one variable in getting that live performance experience.


If you erase your memory of these bad influences and

just think outside the box for a moment, why not seek

quality drivers ? Your 6/7 way loudspeaker would now

shrink down to a 4 way where the 3 way is mains and

the additional one is a great sub.

If you used these quality drivers in a 6/7 way, you are

doing more harm to the sound than good because

you assume the driver can't do it's passband when in

fact it does a great job.

I don't think you peaked anyones interest with this thread,

the idea is just dumb.

 

[Mark Seaton]

Hi usil,


There's no reason what you describe is not possible, and some have built such projects as DIY and exotic products for order. You could do it passively, but to say that makes a complicated project more difficult would be a gross understatement.


My first caution would be that you seem to be looking at this 4D problem of sound reproduction in only 2 dimensions. Of course some will argue that all they care about for music listening is one seating location, most all studies and experience suggests that the full power radiated and it's smoothness off axis vs. on axis matters. That is probably the biggest hurdle of going past a 5-6 way system.


For the record, there are plenty of active electronics which could enable such a product up past using one device for a mono, 10-12 way system.

There is some value to separating bandpasses further, although not for the sake of separating the reproduced frequencies. More benefit lies in making sure lower range dynamics (per bandpass) don't overload the drivers, and to optimize each bandwidth without the requirement of many octaves of operation. Catapult's depiction and explanation is right on. IMO, the hurdle to making such an effort beneficial lies more in the physical size of each device to achieve desired performance vs. keeping them close enough and crossover over such that they combine as a single entity. Including the subwoofer, I could see a 5-6 way being practical with real benefit.

 

[steve71]

Catapult is correct in his diagram. When two waves of differing frequency are present in a medium (air in our case) they intermodulate.

http://en.wikipedia.org/wiki/Intermodulation


Crossovers are a necessary evil in speaker design and there is absolutely nothing to be gained by using more than you need to work around your chosen drivers limitations.

 

[lennon_68]



Quote:
Originally Posted by smokinghot /forum/post/15504682


I think you proved the OP's point instead of showing his flawed concept.


If I understand you pic, and what your saying. The magenta wave is what happens when both high and low frequencies are being produce at the same time, by the same speaker. The highs aren't as high as they should be, and the lows aren't as low. So the OP's concept of multiple drivers would allow a more accurate reproduction of the blue and red waves. The magenta, (or the combined frequencies) which isn't apart of the material being reproduced, is no longer there.


Seems to make sense.

I think what you're missing is that if one speaker were playing the blue (perfectly) and one the red (perfectly) then what our ears would interpret would still be the magenta anyway, so what difference does it make whether one driver is producing the magenta or two are producing different frequencies that combine for the same sound anyway.

 

[mjg100]



Quote:
Originally Posted by Mark Seaton /forum/post/15507617


Hi usil,


There's no reason what you describe is not possible, and some have built such projects as DIY and exotic products for order. You could do it passively, but to say that makes a complicated project more difficult would be a gross understatement.


My first caution would be that you seem to be looking at this 4D problem of sound reproduction in only 2 dimensions. Of course some will argue that all they care about for music listening is one seating location, most all studies and experience suggests that the full power radiated and it's smoothness off axis vs. on axis matters. That is probably the biggest hurdle of going past a 5-6 way system.


For the record, there are plenty of active electronics which could enable such a product up past using one device for a mono, 10-12 way system.

There is some value to separating bandpasses further, although not for the sake of separating the reproduced frequencies. More benefit lies in making sure lower range dynamics (per bandpass) don't overload the drivers, and to optimize each bandwidth without the requirement of many octaves of operation. Catapult's depiction and explanation is right on. IMO, the hurdle to making such an effort beneficial lies more in the physical size of each device to achieve desired performance vs. keeping them close enough and crossover over such that they combine as a single entity. Including the subwoofer, I could see a 5-6 way being practical with real benefit.

Would it not require an enclosure for each speaker with the front baffles aligned and the speakers all tied into a crossover setting each bandwidth. I kind of picture it as a big enclosure with individual sections (optimized for each driver) for each speaker.