2 way vs 3 way

[almikel 1,650]

4 hours ago, Cloth Ears said:

So a good speaker has nothing to do with anything, except it's directivity?

of course not - low distortion at the SPL required of the driver is also critical

 

4 hours ago, Cloth Ears said:

BTW. I wasn't worried about the crossover as such - it's the movement of the sound from one driver to another that is more the issue than the crossover itself.

agreed - which is where the directivity of the drivers at the crossover frequency (and throughout the crossover region) is important - ie how they acoustically combine on and off axis.

Active crossovers are generally steeper than passives, so with a steeper active crossover you can reduce the frequency band of interaction between drivers, but matching driver directivity is still important for a speaker to provide a smooth "on and off axis" frequency response.

 

A smooth "on and off axis" frequency response with low distortion are the metrics put forward by Toole/Olive for "good sounding speakers" and validated by their tests with both trained and un-trained listeners - ie both trained and un-trained listeners prefer speakers with a smooth "on and off" axis response with low distortion.

 

cheers

Mike

[davewantsmoore 6,723]

8 hours ago, Cloth Ears said:

So a good speaker has nothing to do with anything, except it's directivity?

🙄

 

The biggest (by a country mile and then some) difference for those speakers is their directivity.

It's the major reason why they sound different to other speakers  (even when EQed to identical axial response)

 

[davewantsmoore 6,723]

4 hours ago, almikel said:

low distortion at the SPL required of the driver is also critical

I don't think so.   The distortion cannot be "too high".... but low distortion is not critical.     For a relatively well designed driver not being asked to play unfair SPL .... distortion can basically be ignored.

 

4 hours ago, almikel said:

Active crossovers are generally steeper than passives, so with a steeper active crossover you can reduce the frequency band of interaction between drivers, but matching driver directivity is still important for a speaker to provide a smooth "on and off axis" frequency response.

Indeed.  Using very steep slopes can exacerbate the "step" in directivity that might exist between two drivers.

 

4 hours ago, almikel said:

A smooth "on and off axis" frequency response with low distortion are the metrics put forward by Toole/Olive for "good sounding speakers" and validated by their tests with both trained and un-trained listeners -

Check point 1 and 3 in his blog post here:

http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html

 

The NLD distortion got done.... and it says mostly not audible (with caveats).   People just don't want to believe it as it contradicts what audiophiles have been dogwhisting about for decades (both the "pleasant distortion" camp, and the "super low distortion" camps..... both are wrong.  It just not audible for most intents and purposes).

[RoHo 461]

I've wondered about the audibility of distortion in the situation where a woofer is operating in areas of cone break-up.  This is probably more likely to occur in a 2-way speaker where the woofer is forced to cover more of the spectrum.  Is the issue more about frequency response aberrations than distortion?  So with judicious use of EQ you can operate the woofer into this area? Or will it still sound bad, no matter how well designed the speaker is?

[almikel 1,650]

On 24/12/2020 at 10:00 AM, RoHo said:

I've wondered about the audibility of distortion in the situation where a woofer is operating in areas of cone break-up.  This is probably more likely to occur in a 2-way speaker where the woofer is forced to cover more of the spectrum.  Is the issue more about frequency response aberrations than distortion?  So with judicious use of EQ you can operate the woofer into this area? Or will it still sound bad, no matter how well designed the speaker is?

I would expect cone breakup distortion would be quite “high” order , so I would expect it to be more audible than other non linear driver distortion which is typically “low” order

[almikel 1,650]

On 23/12/2020 at 7:49 PM, davewantsmoore said:

I don't think so.   The distortion cannot be "too high".... but low distortion is not critical.     For a relatively well designed driver not being asked to play unfair SPL .... distortion can basically be ignored.

agreed - thanks for the reminder and the links to refresh my understanding of what Toole/Olive actually say...

[henry218 1,359]

On 24/12/2020 at 8:00 AM, RoHo said:

I've wondered about the audibility of distortion in the situation where a woofer is operating in areas of cone break-up.  This is probably more likely to occur in a 2-way speaker where the woofer is forced to cover more of the spectrum.  Is the issue more about frequency response aberrations than distortion?  So with judicious use of EQ you can operate the woofer into this area? Or will it still sound bad, no matter how well designed the speaker is?

A well designed speakers wont let the cone breakup interfere the music repro on the speakers.

On the other hand designing speakers that can be very expressive, controlled and enjoyable is the very art itself.

[henry218 1,359]

On 23/12/2020 at 5:39 PM, davewantsmoore said:

🙄

 

The biggest (by a country mile and then some) difference for those speakers is their directivity.

It's the major reason why they sound different to other speakers  (even when EQed to identical axial response)

 

Its also one of the major contributing factor for the sound in the room.

[almikel 1,650]

On 24/12/2020 at 10:00 AM, RoHo said:

I've wondered about the audibility of distortion in the situation where a woofer is operating in areas of cone break-up.  This is probably more likely to occur in a 2-way speaker where the woofer is forced to cover more of the spectrum.  Is the issue more about frequency response aberrations than distortion?  So with judicious use of EQ you can operate the woofer into this area? Or will it still sound bad, no matter how well designed the speaker is?

 

On 25/12/2020 at 3:35 PM, almikel said:

I would expect cone breakup distortion would be quite “high” order , so I would expect it to be more audible than other non linear driver distortion which is typically “low” order

Geddes, Toole, Olive all discuss the inadequacies of current non-linear distortion measurements (eg THD, IMD) that don't correlate well to actual listening test results.

 

Geddes proposed a new "distortion measurement" http://www.gedlee.com/Papers/Distortion_AES_I.pdf that includes:

1. It should be more sensitive to higher order nonlinearities than lower order ones.
2. It should be weighted towards greater values for nonlinearities at lower signal levels
3. It must be immune to changes in offset and gain (first order slope) since these are inaudible effects (I'm not sure what this means???)

Mike

[frozenpod 1]

Regarding directionality and off axis reaponse the article with rear facing drivers whilst the article makes them look amazing it had me thinking of the Bose 901s.

 

If I remember correctly they had a large number of drivers in the back of the speaker pointed towards the wall (maybe 9) and one forward facing driver.

 

I remember the 901s being highly room dependant as to how they would sound.

 

 

[aussievintage 4,070]

30 minutes ago, frozenpod said:

I remember the 901s being highly room dependant as to how they would sound.

I remember them sounding bad  

[frozenpod 1]

12 hours ago, aussievintage said:

I remember them sounding bad  

 

Yes but was that due to the lack of a tweeter or the drivers firing towards the wall?

 

[aussievintage 4,070]

1 hour ago, frozenpod said:

 

Yes but was that due to the lack of a tweeter or the drivers firing towards the wall?

 

 

They were Bose, nuff said      I honestly never heard a Bose system I liked - even including their PA systems

[Cloth Ears 1,238]

On 23/12/2020 at 8:39 PM, davewantsmoore said:

🙄

 

The biggest (by a country mile and then some) difference for those speakers is their directivity.

It's the major reason why they sound different to other speakers  (even when EQed to identical axial response)

 

So Henricksen's paper is wrong? Where he says:

"Fortunately, all single speakers in enclosures behave in an identical manner as far as directivity is concerned. Furthermore, this behavior is independent of the cone geometry (dome, deep-dish, shallow cone, etc) as pointed out by Kates and experiments at Altec on various woofers."

[davewantsmoore 6,723]

On 29/12/2020 at 7:52 AM, Cloth Ears said:

So Henricksen's paper is wrong? Where he says:

 

No.  He is not wrong..... (you just don't understand what he is saying)  🤣

 

He is not saying that all single spakers have the same directivity......  but that all (circular) transducers have dirctivity which adeheres to the same formula (ie. it is proportional to the transducer radius)

 

What this says is that drivers which ar th sam size(s) (and with the same crossovers in th case of a multi-way speaker) have the same directivity.... fairly independently (with a few caveats) of cabinet size.

 

.... but as this differs between most speakrs, they have diffrent angular frequncy respons..... and this is the thing which you hear most dominantly (so much so, it is almost important in exclusion to almost all else).

 

[davewantsmoore 6,723]

On 27/12/2020 at 5:14 PM, almikel said:

Geddes proposed a new "distortion measurement" http://www.gedlee.com/Papers/Distortion_AES_I.pdf that includes:

Which he later abandoned, replaced by the belief that non-linear distortion is totally irrlevant.... beyond a few caveats, that are easy to naviagte using well designed speaker drivers and elctronics.

[davewantsmoore 6,723]

On 25/12/2020 at 4:35 PM, almikel said:

I would expect cone breakup distortion would be quite “high” order

Not really, no.

 

.... it is just (usually) extremely significant in level (and time domain properties).... looing at it by it's THD (or IMD) is not a very hlpful way to charactrise it.

[davewantsmoore 6,723]

On 27/12/2020 at 4:44 PM, almikel said:

agreed - thanks for the reminder and the links to refresh my understanding of what Toole/Olive actually say...

 

..... the industry is having a very very hard time of letting go of the whole "non-linear distortion in %s" thing.   

[Cloth Ears 1,238]

15 minutes ago, davewantsmoore said:

 

No.  He is not wrong..... (you just don't understand what he is saying)  🤣

. . .

(so much so, it is almost important in exclusion to almost all else).

 

I do, but your short responses, in a topic that is unrelated, meant that I did not think that YOU understood what he meant. Why do you take topics away from their original question and steer them to your own generalised ideas? 🤣

 

I guess I don't need to spend my money on expensive drivers, as all I need is to is design the baffle correctly and any speakers will produce perfect sound?

 

Surely it deserves its own topic, if you are so vehement in your defense of it.

[Steffen 1,194]

22 minutes ago, davewantsmoore said:

 

What this says is that drivers which ar th sam size(s) (and with the same crossovers in th case of a multi-way speaker) have the same directivity.... fairly independently (with a few caveats) of cabinet size.

 

.... but as this differs between most speakrs, they have diffrent angular frequncy respons..... and this is the thing which you hear most dominantly (so much so, it is almost important in exclusion to almost all else).

 

 

But surely it’s not just driver size and crossover design? Otherwise it would be relatively straightforward to hone in on the optimal set of driver sizes and crossover schematic, and by extension, make all speakers with the same number of drivers sound alike. What about things like driver arrangement (both in the x-y plane of the front baffle and in the z-direction relative to the baffle), or the baffle size, or things that modify diffraction like wave guides, felt strips, etc?

[davewantsmoore 6,723]

3 hours ago, Steffen said:

But surely it’s not just driver size and crossover design?

 

Surely what?
Nothing else affects the directivity profoundly.    The quoted paper is right that all (circular) drivers (of the same radius) have larglely similar direcitity.

 

The caveats are at higher frequencies....where cone shape and geometry and the cabinet, contribute..... but the paper was focussed on low frequencies.

 

 

3 hours ago, Steffen said:

Otherwise it would be relatively straightforward to hone in on the optimal set of driver sizes and crossover schematic, and by extension, make all speakers with the same number of drivers sound alike.

Well...

 

If you have a speaker with the same drivers/layout.... and the same crossover..... and you EQ them to the same axial response..... then they will sound "alike".

 

 

 

3 hours ago, Steffen said:

What about things like driver arrangement

Yes, this can profoundly affect directivity.

The quoted paper was discussing LF where this isn't an issue. 

 

Unfortunately a strawman was errctd where I could say that paper was wrong, or.....   

 

3 hours ago, Steffen said:

(both in the x-y plane of the front baffle and in the z-direction relative to the baffle), or the baffle size, or things that modify diffraction like wave guides, felt strips, etc?

Wavguides have a profound effect ..... although like drivers, their dirctivity is bound by the mouth size.

[davewantsmoore 6,723]

3 hours ago, Cloth Ears said:

I guess I don't need to spend my money on expensive drivers, as all I need is to is design the baffle correctly and any speakers will produce perfect sound?

After testing a metric shedload of drivers, I think this is much more true than people think.... given a few cavats.

 

 

.... BUT back to THIS topic.   Directivity is the BIGGEST difference between ANY speakers....  it's hard to generalise a lot about 2-ways vs 3-ways without taking this into account.    The danger is that we say "all 2-ways this" ... or "all 3ways that" .... which might lead people astray.

 

3 hours ago, Cloth Ears said:

 

Why do you take topics away from their original question and steer them to your own generalised ideas? 🤣

I'm actually really not trying to do that (which may surprise you).... I'm just trying to help people understadn the difference between 2 and 3 ways  

[davewantsmoore 6,723]

3 hours ago, Steffen said:

 

But surely it’s not just driver size and crossover design? Otherwise it would be relatively straightforward to hone in on the optimal set of driver sizes and crossover schematic, and by extension, make all speakers with the same number of drivers sound alike. What about things like driver arrangement (both in the x-y plane of the front baffle and in the z-direction relative to the baffle), or the baffle size, or things that modify diffraction like wave guides, felt strips, etc?

 

Indeed this is precicely why it is hard to generalise about 2way vs 3way, etc.

 

 

You could make a 3way with a lumpy or perfct response....  and narrow or wide dirctivity ..... and the same for a 2-way.    .... and these things will sound completely (and utterly) different from each other.

 

[TerryO 2,849]

On 21/11/2020 at 10:33 AM, Ian Lyons said:

I  am currently still listening to a number of different speakers.My question  at the moment  is about 2 way and 3 way speakers.I dont if I am fooling myself under the guise that more is better,but I feel the 3 way speakers that I have listened to  have more presence than 2 way.One manufacturer said 2 way is most definitely better.I appreciate cost would play a part in  this,but  for an equal value what do members prefer.

So Ian, are you any the wiser after four pages of comments? ... 

 

 

cheers,

Terry

[Hazzzy 388]

On 27/12/2020 at 9:30 PM, aussievintage said:

I remember them sounding bad

They  impressed me big time in 1980 at a party. Might have been influenced by other substances at the time but the memory is there.

I said at the time I will get a pair of these speakers but went for marantz hd 600, which were great for many years. 

Edited April 3 by Hazzzy
model number