Thanks a lot for the helpful advice and i am perfectly fine with it.
I can understand the difficulty for a single drive to do what you say.
Even if the idea can look intriguing in some ways.
This confirms what i have heard ... 4 ways are needed to cover the audio range
I was wondering about some kind of testing easy to perform to get at least an idea about the driver/speaker quality
I see the all issue is much much more complex of what i thought, as usual.
Thanks again and kind regards, gino
See, you are picking up hearsay that may or may not be correct in the instance you are thinking about. I find three ways to do just fine. The old decade rule is still about right. A driver does pretty well over about a decade. Speaker design is no where near that easy though. Please, hit the books. Learn a bit of the physics before you think you understand which answers are trustworthy. ( hint, some of what you quote is, some is not)
The OPs intention was to find a single measure to find the "best drivers" available. I understand this wish because we all try to find "the best drivers" with the least possible effort but unfortunately there is no such single measure that tells everything.
Just an example: Some excellent woofers with ceramic or metal cones would definitely fail such a test but nonetheless could they be used for a multiway project that has good square-wave response.
And we have not even considered any other properties yet ....
Regards
Charles
Just an example: Some excellent woofers with ceramic or metal cones would definitely fail such a test but nonetheless could they be used for a multiway project that has good square-wave response.
And we have not even considered any other properties yet ....
Regards
Charles
At a certain point you have to put down the book and switch on the scope
And i would much prefer to read a scope user manual than to go through pages and pages of formulas
In the end you have to measure to know what the real behaviour is
This is where i am aimed to ... measurements
I believe that the drivers selection must be based only on measurements
which kind of measurements we can discuss ... as i am doing here
You say decade ... i do not understand well.
I just say that if i take a woofer and this woofer reproduce a decent SW let's say only from 50 to 200 Hz i would not be comfortable in using it outside this range
Because outside this range it deforms the input signal shape. What i get does not resemble to what i send in.
Moreover if i have two woofers, one good at reproducing a decent SW and the other bad at that, i would prefer the first.
Am i wrong ? you tell me.
Regards, gino
Last edited:
Hi and thanks for the reply. You have pinpointed my basic idea
But we can talk of course of a set of measurements ... just one would be too good to be true
Just an addition: if a driver presents very good instrumental measurements only in a limited range of frequencies, there is its best field of application.
I guess it is impossible to have a single driver covering all the audio bandwidth with excellent performance at measurements
What i cannot accept easily it is a selection by ear ... by listening
I am not saying that the SW is the best signal around to test a driver
I am just asking if it can give some kind of information about the performance of a driver. I do not know.
Let's say that it is a basic signal for testing. A very though one also.
Ok ... it is not present in nature. But also a driver is not present in nature.
It is an invention of humans.
But yes you are almost right .... my intention still is to find a single measure useful to select drivers (i.e. you have 4 o 5 woofers with similar general characteristics and you want to pick the best of the bunch).
Of course i do not want to ditch the frequency response or distortion test for sure
These are the very fundamental and basic tests that we can find also in the drivers datasheets (not so often the distortion tests and this is for me very very sad)
I am not that mad
Thanks and regards, gino
Last edited:
Just to clarify and as far as I know:
100Hz - 200Hz is an octave ie a doubling of frequency;
100Hz - 1000Hz is a decade ie 10x the fundamental frequency.
So a decade would be slightly more than 3 octaves which I think is reasonable coverage to expect from one driver. It also means that a 3way should theoretically be enough to cover 20-20 000Hz successfully.
100Hz - 200Hz is an octave ie a doubling of frequency;
100Hz - 1000Hz is a decade ie 10x the fundamental frequency.
So a decade would be slightly more than 3 octaves which I think is reasonable coverage to expect from one driver. It also means that a 3way should theoretically be enough to cover 20-20 000Hz successfully.
Hi and thanks for the very valuable information
Of course i have nothing against theory, even if it is beyond my capacity to understand rightly.
But instruments do not lie and a lot of interesting measurements can be done that will be very telling
I like instruments a lot.
Because when you design maybe you make some assumptions on componets behaviour
With instruments we can check the real behaviour.
I am a supporter of measurements. They are beautiful.
On principle i think that in the future it would be possible to have a program with a set of test signals loades to check the drivers performances much much better than with listening tests, that i do not like at all.
But in the end i agree on the doubt about the opportunity to use a signal like the SW ... maybe another shape would be better ... i dont know
It was just an idea ... bad idea maybe.
Thanks a lot. Regards, gino
20Hz to 20480Hz is 10 octaves.
Three 3octaves capable drivers can't cover 10 octaves.
30Hz to 15360Hz is 9 octaves.
Four drivers each with AT LEAST 2.5octave capability can just reach 10octaves, if the overlap at each crossover is absolutely minimal.
It would seem that north of the border 'slightly more' means 'precisely the same'.
I have to bear that in mind next time I travel to Scotland. ;-)
Unless I am mistaken the only speakers capable of really reproducing square waves are quad electrostatics. One can of course argue that this ability is irrelevant to sound reproduction but it's hard to see why that might be the case given that a good sq w is not possible without excellent phase and frequency fidelity. Both desirable attributes in sound reproduction.
Last edited:
Hi and thanks. Very interesting.
I remember now an advert on HiFi News & Record Review, if i am not wrong, about a company building a passive multiway crossover able to pass a very decent SW.
I think that they presented this as a remarkable achievement
I do not remember the Company name by the way
Regards, gino
P.S. as i am also interested in coupling caps i wonder if a SW could be a good test signal to select the best caps for signal coupling
I remember now an advert on HiFi News & Record Review, if i am not wrong, about a company building a passive multiway crossover able to pass a very decent SW.
I think that they presented this as a remarkable achievement
I do not remember the Company name by the way
Regards, gino
P.S. as i am also interested in coupling caps i wonder if a SW could be a good test signal to select the best caps for signal coupling
Not so, some of the Danley Synergy horns (and some DIY versions, my own included) can do it as well. And something like the Manger, if used fullrange, can also.
One thing to keep in mind for most speaker measurements (including square wave reproduction capability) is that a speaker doesn't have just one output like an amplifier channel does. A speaker has many outputs that are almost always different toward every direction. Making a multi-driver, non-coaxial speaker reproduce a square wave only when measured from one direction (i.e. on some defined "listening axis") is difficult, but even when achieved probably still doesn't mean a heck of a lot. In a real room, you are going to hear all the sound outputs no matter what direction they are leaving the speaker at, as it will reflect around and get to you by some time (just turn a box speaker away from you and notice that the sound level really doesn't drop all that much, though the character of that sound probably will!).
Getting a speaker to reproduce a square wave (perhaps attenuated) over a wide range of output directions might mean something. But still I don't know of any documented evidence showing that that is necessary either.
Hi and thanks for the very interesting explanation
The main reason for the question is what i read about drivers selection.
If i understand well most of this process is done by ear, by listening to signal/music. If i am wrong i am here to know.
And sorry but i find this quite unscientific. Again tell me if i am wrong.
I would expect that after all these years some kind of instrumental procedure were available, with well defined test signals.
I do believe in difference between one driver and another, starting from the cone material for instance. I prefer paper cones almost always.
Thanks and regards, gino
P.S. and drivers selection is the first step of a project. The first step.
I should study more drivers.
Last edited:
Hi, just to say that i have found a champion ... at least at 1 kHz
1kHz squarewave response on listening axis at 1m
From Stereophile magazine.
I wonder if this behaviour is mantained also around that frequency.
Impressive.
I listened to these speakers and they amazed me for the realism of sounds, especially when reproducing special effects in movies.
I had an even scaring experience, never heard before.
We were watching a movie .. at a certain point a guy behind a glass wall knocked on it
I swear, i felt like he was inside the television knocking on the screen.
Unfortunately i do not remember the title of the movie but the effect was so real that both me and my friend, the owner of the speakers, were astonished.
Unbelievable. Was not a reproduced sound ... was the real sound.
Thanks and regards, gino
1kHz squarewave response on listening axis at 1m
From Stereophile magazine.
I wonder if this behaviour is mantained also around that frequency.
Impressive.
I listened to these speakers and they amazed me for the realism of sounds, especially when reproducing special effects in movies.
I had an even scaring experience, never heard before.
We were watching a movie .. at a certain point a guy behind a glass wall knocked on it
I swear, i felt like he was inside the television knocking on the screen.
Unfortunately i do not remember the title of the movie but the effect was so real that both me and my friend, the owner of the speakers, were astonished.
Unbelievable. Was not a reproduced sound ... was the real sound.
Thanks and regards, gino
Last edited:
Hi Gino, which speaker is it you call the champion?
Hi well champion maybe only on this specific issue.
It is the Quad esl 63
And there is only the test with a 1kHz SW.
Still it looks very good to me for a speaker.
As i said i listened extensively to these speakers because a friend owns a pair.
No real bass, limited max ouptut.
Someone call them the big headphones ...
I do not like at all their width ... in general i do not like panels
But the realism of some effects was completely unbelievable.
Like the crash of a glass ... never heard something more real really.
I do not know if this can be related to their behaviour in the SW response test
Thanks and regards, gino
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.