I feel that some may find the question controversial but the aim is benign, really.
I think we all will agree that no loudspeaker system is perfect and there are countless sources of defects in reproduction. Each loudspeaker designer aims to address and eliminate or reduce the effects of some, many or most of them but it's obvious that not all can be solved or equally addressed.
I will list some of the problems that a designer has to face (at least some that I know of): Low frequency extension, flatness of frequency response, irregularities at crossover points, lobbing, dispersion, baffle step, enclosure resonances, port puffing, transducer distortion, cone breakup, etc.
Now, since solving a problem may introduce others (like going from a full range speaker to extend the frequency response, to a 2-way or 3-way introduces the maladies and problems of multiple drivers and crossover points) and making a large enclosure increases the issues of panel resonances, each design decision prioritizes problems and in one way or another, makes solving some of the problems more important than others.
Not limited by the list I provided (from my limited knowledge and poor memory), what is the one or two things you believe a good loudspeaker should eliminate or reduce substantially to sound good?
I think we all will agree that no loudspeaker system is perfect and there are countless sources of defects in reproduction. Each loudspeaker designer aims to address and eliminate or reduce the effects of some, many or most of them but it's obvious that not all can be solved or equally addressed.
I will list some of the problems that a designer has to face (at least some that I know of): Low frequency extension, flatness of frequency response, irregularities at crossover points, lobbing, dispersion, baffle step, enclosure resonances, port puffing, transducer distortion, cone breakup, etc.
Now, since solving a problem may introduce others (like going from a full range speaker to extend the frequency response, to a 2-way or 3-way introduces the maladies and problems of multiple drivers and crossover points) and making a large enclosure increases the issues of panel resonances, each design decision prioritizes problems and in one way or another, makes solving some of the problems more important than others.
Not limited by the list I provided (from my limited knowledge and poor memory), what is the one or two things you believe a good loudspeaker should eliminate or reduce substantially to sound good?
Assuming that the response needs to be flat (smooth is still good).
I agree, you should not eat loudspeakers
Assuming that the response needs to be flat (smooth is still good).
I agree, you should not eat loudspeakers
Why not ?
Music is food for the soul.
You should eat it, digest it and evacuate.
Use whatever "paper" to clean yourself
This is not correct.
The on-axis response should ideally be flat but given what the combination of the off-axis response and the room may be doing improvements in perceived sound can often be found by moving away from a flat on-axis response. Dipoles are a common example.
The off-axis response should be smooth but claims for the correctness of constant directivity by it's many supporters tend not to be supported by listening tests. What seems to be the best compromise, perhaps not surprisingly, is to mimic in an averaged sense what musical instruments do and narrow the beamwidth with rising frequency. Both the width and the degree of narrowing are obviously influenced by the room, the recording and the listeners preference in terms of spaciousness.
There is a range of off-axis responses that can work well without there being a correct one. To have a correct one we would need the recording industry to take an interest in genuine sound quality and specify the degree and type of indirect sound to be included in the recording and what is expected from the speaker's interaction with the room.
Read it, twice. Good book. Tells you that two ears and a brain work wonders to fix almost whatever fault there is in the speakers and the room.
But why not make the illusion a bit better?
Am I totally off shifting this towards "should have relatively flat power response"?
In my opinion it is placement relative to the user. The importance of the equilateral triangle for the stereo effect cannot be overstated.
I have have been listening to music for many years including a full rack system in the 80s and good headphones. It is only when I stated my modest speaker builds with crossover less full range speakers pulled from old computer speakers that I saw the light. Digital source CD, You Tube, Deezer these have elevate the source quality to levels unimaginable with tape.
In fact, I am in the planning process for speakers for my living room because the large box Sony speakers from the 80s system I am using cannot be placed at ear level due to their size. Smaller speakers placed where I am listening more or less on-axis should give me a memorable listening experience.
Listening in the car I regret the fact that the door speakers are aimed at my shin mainly, not the best thing for listening.
I have have been listening to music for many years including a full rack system in the 80s and good headphones. It is only when I stated my modest speaker builds with crossover less full range speakers pulled from old computer speakers that I saw the light. Digital source CD, You Tube, Deezer these have elevate the source quality to levels unimaginable with tape.
In fact, I am in the planning process for speakers for my living room because the large box Sony speakers from the 80s system I am using cannot be placed at ear level due to their size. Smaller speakers placed where I am listening more or less on-axis should give me a memorable listening experience.
Listening in the car I regret the fact that the door speakers are aimed at my shin mainly, not the best thing for listening.
A speaker doesn't distort
A speaker driven beyond its capability distorts
You are talking about crossover distortion, same as amplifiers with shared swinging output devices?
But that happens because of the artifacts created during the emission.
When more than one voice is used, all, more or less, contribute to the total of what arrives at your ear(s).
Crossover distortion is mostly phase related. See? As speakers are minimum phase devices ( now I don't remember the exact name), they replicate exactly ( more or less > see also phase changes in the unwrapped phase curve graphic of any LS ) what's at the input.
A speaker driven beyond its capability distorts
You are talking about crossover distortion, same as amplifiers with shared swinging output devices?
But that happens because of the artifacts created during the emission.
When more than one voice is used, all, more or less, contribute to the total of what arrives at your ear(s).
Crossover distortion is mostly phase related. See? As speakers are minimum phase devices ( now I don't remember the exact name), they replicate exactly ( more or less > see also phase changes in the unwrapped phase curve graphic of any LS ) what's at the input.
A speaker doesn't distort
A speaker driven beyond its capability distorts
So far there are no speakers which do not distort. None. What you mean is 'does not distort to a level it impacts the sound reproduction too much.
That's simply not true. A phase error does not automatically result distortion. And there are a lot of other distortions it can add, ie. a saturated core coil or cross induction of different coils or eddy currents, just to name something.
Again, a phase error does not equal distortion. And they are not 'minimum phase devices' either. You can build such speakers, the vast majority are not though.
Or save $56 and the time it takes to read 500 pages by posting a thread here at DIYaudio.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.