Hi,
Apologies, I cant really link to where I read this because I cant find it anymore, I think it was in a thread about advantages of MTM (2.5-way?) over 3-way on this site somewhere. It stuck in my mind though.
Be gentle, this is from memory, it went along the lines of a 3-way woofer and midwoofer will have different "voices", membrane size, cone material...so when you XO right in the middle of a female vocal range part of the singers range will sound "different" as it crosses drivers, meaning you have to take care with driver choices so they match?
My question is how big of an issue is this really, can it be measured, is it audible to non golden ears, can it be "fixed" with DSP crossover (I doubt I'll ever build or buy a passive crossover), reason I ask is I feel "soured" against 3-way?
Manufacturers are reinforcing this because they seem to have moved to sexier MTMs and coaxials (sometimes with added big bass woofers).
I know this is somewhat subjective.
Thanks!
Apologies, I cant really link to where I read this because I cant find it anymore, I think it was in a thread about advantages of MTM (2.5-way?) over 3-way on this site somewhere. It stuck in my mind though.
Be gentle, this is from memory, it went along the lines of a 3-way woofer and midwoofer will have different "voices", membrane size, cone material...so when you XO right in the middle of a female vocal range part of the singers range will sound "different" as it crosses drivers, meaning you have to take care with driver choices so they match?
My question is how big of an issue is this really, can it be measured, is it audible to non golden ears, can it be "fixed" with DSP crossover (I doubt I'll ever build or buy a passive crossover), reason I ask is I feel "soured" against 3-way?
Manufacturers are reinforcing this because they seem to have moved to sexier MTMs and coaxials (sometimes with added big bass woofers).
I know this is somewhat subjective.
Thanks!
You'll get different opinions on this, and here's one.
The issue of different cones sounding different tends to be minimised by using them within their intended range and watching for resonance or breakup issues.
As far as crossing is concerned, a crossover can be made inaudible but it's not easy. Usually best to presume it will be slightly imperfect unless you know otherwise.
The issue of different cones sounding different tends to be minimised by using them within their intended range and watching for resonance or breakup issues.
As far as crossing is concerned, a crossover can be made inaudible but it's not easy. Usually best to presume it will be slightly imperfect unless you know otherwise.
Thanks, much appreciated and it makes perfect sense to me.
On crossing, if you've measured your drivers, and are within your measured good FR (no resonance or breakup) can it essentially be made inaudible with a software based DSP XO?
While I'm at it... coaxial "single apparent source", is that something a non golden ears is going to be "floored" by?
Thanks again!
On crossing, if you've measured your drivers, and are within your measured good FR (no resonance or breakup) can it essentially be made inaudible with a software based DSP XO?
While I'm at it... coaxial "single apparent source", is that something a non golden ears is going to be "floored" by?
Thanks again!
I don't know of any evidence but my guess is that this effect shows up usually from jumps in directivity from driver to driver, sometimes from a jump in edge diffraction (changes stereo image), and sometimes from resonant issues. Point being all of those can be minimized if not eliminated between pretty different drivers.
Indeed you will get a variety of opinions.
Tweeters range in size but crossover points will be typical and require to not be too low
to reduce distortion.
Mystical magical wild guessing about vocal blah blah is just that Blah blah.
It either sums flat or doesnt, its either crossed to low and distorts or it doesnt.
MTM will increase center to center spacing and the crossover point becomes very low
to " fix it" which will basically mean a bigger tweet, Making the spacing worse.
Or a low crossover point causing distortion at higher listening levels.
It looks " cool" and balanced and all kinds of techno logic to support it.
And is the biggest waste of time. And yields absolute garbage vertical.
Dead center drivers for bad diffraction and non flat response.
30 ohm junk peaks in the crossover to fix the baffle step yada yada.
Pure junk, waste of time
Tweeters range in size but crossover points will be typical and require to not be too low
to reduce distortion.
Mystical magical wild guessing about vocal blah blah is just that Blah blah.
It either sums flat or doesnt, its either crossed to low and distorts or it doesnt.
MTM will increase center to center spacing and the crossover point becomes very low
to " fix it" which will basically mean a bigger tweet, Making the spacing worse.
Or a low crossover point causing distortion at higher listening levels.
It looks " cool" and balanced and all kinds of techno logic to support it.
And is the biggest waste of time. And yields absolute garbage vertical.
Dead center drivers for bad diffraction and non flat response.
30 ohm junk peaks in the crossover to fix the baffle step yada yada.
Pure junk, waste of time
It's possible to cross and not hear that the drivers aren't coaxial. The idea that they should be coaxial is more theoretical, it's also pleasing to ones sensibilities even if it isn't required.
Coaxial does potentially offer to do away with lobing although it can also create problems in the practical implementation. In short, they can be easier to work with and to cross.
DSP can't change the direction of radiated sound.
Coaxial does potentially offer to do away with lobing although it can also create problems in the practical implementation. In short, they can be easier to work with and to cross.
Producing an effectively ideal cross involves considering all the sound in all the directions and how the room sends that back to you.
DSP can't change the direction of radiated sound.
Thanks, any good links for a good explanation on this, are you saying that an ideal crossover is only ideal in the one room it was designed in?
^^ In some cases even the entire speaker is, however that's an advanced topic. It really is difficult to suggest where to start but perhaps if you want a practical goal to work towards you could look to do a three dimensional crossover. Now this alone is only part of the equation but you want to lead into understanding the acoustics behind speakers.
If you want the easy way into 3D crossovers you might begin learning vituixcad and trying to use it to it's potential.
As far as building is concerned you might look at waveguides, and use what you learn there to approach other parts of the speaker as well.
If you want the easy way into 3D crossovers you might begin learning vituixcad and trying to use it to it's potential.
As far as building is concerned you might look at waveguides, and use what you learn there to approach other parts of the speaker as well.
Not coaxial , Drivers mounted or Centered on the baffle for less ideal response but " looks" better
Thanks, are you talking about using vituixcad to output a 3D representation of the optimal sound of the inputted DIY speaker parameters, data is then exported to winpcd to give you a passive crossover design?
Wow, I'm not sure if I'll ever go to those lengths.
It's similar to doing a regular crossover, you just have multiple measurements to work with at the same time. You take those angular measurements and the crossover simulator makes it easier to see them as a whole.
Also keep in mind that the characteristic sound of a voice is mostry related to it's harmonics and those are probably coming from another driver anyhow.
The DSP wouldn't change the direction of radiated sound if multiple independently driven woofers were not used.
Subcardioid pattern control down to low frequencies with a single woofer is possible, as seen in Fulcrum Acoustic's CCX series of 8" to 15" coaxial loudspeakers. DSP is used in the CCX to flatten frequency and phase response, but has no effect on the polar response.
Indeed.
Of course the point remains that DSP can't change sound which has been radiated, which makes the original point that when you have a speaker which is radiating sound in any manner of the designers choosing, and thereby creating a set of room interactions, those cannot be altered by DSP.
The reason can be explained simply by saying that the DSP signal path is itself not three dimensional. There are methods that are used as a last resort, convolution for example, which compound the problem in an attempt to reduce audibility by introducing extra signals to counter reflections in intensity, but smearing time and not affecting space.
Of course the point remains that DSP can't change sound which has been radiated, which makes the original point that when you have a speaker which is radiating sound in any manner of the designers choosing, and thereby creating a set of room interactions, those cannot be altered by DSP.
The reason can be explained simply by saying that the DSP signal path is itself not three dimensional. There are methods that are used as a last resort, convolution for example, which compound the problem in an attempt to reduce audibility by introducing extra signals to counter reflections in intensity, but smearing time and not affecting space.
And thus we got into four dimensions in no time 😆
But back to the original question, with dynamic loudspeakers distortion rises with cone excursion. For this reason alone three-way systems offer more potential in terms of sound quality. Directivity also is an issue, but that has been discussed. Lastly, it’s not that easy for a cone to perform great over a big range (say 30-2500Hz) and produce adequate SPL.
So I’d say such claims are moot.
But back to the original question, with dynamic loudspeakers distortion rises with cone excursion. For this reason alone three-way systems offer more potential in terms of sound quality. Directivity also is an issue, but that has been discussed. Lastly, it’s not that easy for a cone to perform great over a big range (say 30-2500Hz) and produce adequate SPL.
So I’d say such claims are moot.