Thank you very much again I have to read those papers very carefully before asking silly things
bad news indeed Very bad actually i do not see any nearfield studio monitor using horns This could be the result of what you say
Instead many use waveguides or lenses I should study those ?
this surprise me a lot I was sure that distortion decreases when SPL decreases This makes horns unsuitable for low level listening in the near field ?
bad news indeed Very bad actually i do not see any nearfield studio monitor using horns This could be the result of what you say
Instead many use waveguides or lenses I should study those ?
No, it doesn't make horns unsuitable for use at low levels. Distortion will increase as SPL increases in any speaker. What I meant is something like the following, though these numbers are made up:
Many people consider the distortion values in horns to be negligible. But they are typically higher than those seen in high quality direct radiators when used within the low distortion SPL region of the direct radiator.
Nearfield can be more problematic for horns because you normally need a minimum distance for crossovers/drivers to integrate properly. This minimum distance is larger for large speakers.
There is no perfect speaker. They are always a large series of compromises. Which compromises are made comes down to the designer's opinions about which things matter more for the particular use case.
| Direct radiator | Horn |
| 80 dB | 0.1% distortion | 0.5% distortion |
| 110 dB | 3% distortion | 1% distortion |
Many people consider the distortion values in horns to be negligible. But they are typically higher than those seen in high quality direct radiators when used within the low distortion SPL region of the direct radiator.
Nearfield can be more problematic for horns because you normally need a minimum distance for crossovers/drivers to integrate properly. This minimum distance is larger for large speakers.
There is no perfect speaker. They are always a large series of compromises. Which compromises are made comes down to the designer's opinions about which things matter more for the particular use case.
Last edited:
If you have a chance, compare the usual 2 way PA speakers with something suitable from Danley Sound Labs.
I see they have a presence in Italy.
https://m.facebook.com/profile.php?id=100063546399388
I built a pair of SH50 (3 way horns);
Compression driver, 4x5" mids and 2 x 12" LF in each horn / cabinet.
I use them with 2 X sub woofers (tapped horns)
Best thing I've ever done. Replaced a 4 way horn system + the same subs with ease!
They are more of a challenge to build and get right than some but... The results🙂
I see they have a presence in Italy.
https://m.facebook.com/profile.php?id=100063546399388
I built a pair of SH50 (3 way horns);
Compression driver, 4x5" mids and 2 x 12" LF in each horn / cabinet.
I use them with 2 X sub woofers (tapped horns)
Best thing I've ever done. Replaced a 4 way horn system + the same subs with ease!
They are more of a challenge to build and get right than some but... The results🙂
Hi thank you very much to All of you for the really kind and precious advice
After a lot of thinking my best wanted solution considering that I listen from only 3 meters could be a small full range horn loaded above a bass box
Problem is that most of these little fr have a nasty 10dB peak around 10kHz that could sound offensive
If not I would use them immediately cut a 300Hz or so
After a lot of thinking my best wanted solution considering that I listen from only 3 meters could be a small full range horn loaded above a bass box
Problem is that most of these little fr have a nasty 10dB peak around 10kHz that could sound offensive
If not I would use them immediately cut a 300Hz or so
An example of the model and frequency response you are talking about would help. Properly designed horn speakers can have mostly flat response through the treble, but they also typically need a creative crossover or equalization to achieve it. The raw/unfiltered response of a compression driver/horn is often not pretty.
Many start rolling off around 15 kHz. A high frequency resonance is also relatively common, but it's not always large or at a frequency that's likely to be intrusive. In something like a strictly pro audio application these features are often irrelevant for the use case, so those may have more ragged response.
As stated earlier, the horn and compression driver work together, so ultimately you have to look at them as a unit. Compression driver frequency response is often given based on a particular horn or a plane wave tube. That does not tell you exactly how that compression driver will behave on a different horn.
This model from the Celestion pages is a good example of top octave performance that's pretty smooth by horn standards.
https://celestion.com/wp-content/uploads/2019/10/186.pdf
Many start rolling off around 15 kHz. A high frequency resonance is also relatively common, but it's not always large or at a frequency that's likely to be intrusive. In something like a strictly pro audio application these features are often irrelevant for the use case, so those may have more ragged response.
As stated earlier, the horn and compression driver work together, so ultimately you have to look at them as a unit. Compression driver frequency response is often given based on a particular horn or a plane wave tube. That does not tell you exactly how that compression driver will behave on a different horn.
This model from the Celestion pages is a good example of top octave performance that's pretty smooth by horn standards.
https://celestion.com/wp-content/uploads/2019/10/186.pdf
Thank you very much indeed I see It is a matter to find the best combination
Just one question You say
i saw an upgraded passive xover with unbelievably chunky coils and beer can size caps mounted on a board as big as a small kitchen table
ok people are free to do what they like
anyway ... what is your opinion abot EQ ? i have seen that some units have even a automatic function that can flatten almost any curves
i just wonder if this kind of units can be very detrimental for sound If not trying one seems almost mandatory to me
all those breakup peaks could be flattened out just pressing a button ? 😵
this would be wonderful
Just one question You say
i do not want sound lazy but this sounds like a show stopper to me I cannot stand complex xovers They look naive to me
i saw an upgraded passive xover with unbelievably chunky coils and beer can size caps mounted on a board as big as a small kitchen table
ok people are free to do what they like
anyway ... what is your opinion abot EQ ? i have seen that some units have even a automatic function that can flatten almost any curves
i just wonder if this kind of units can be very detrimental for sound If not trying one seems almost mandatory to me
all those breakup peaks could be flattened out just pressing a button ? 😵
this would be wonderful
The crossovers don't have to be complex to achieve good response, but they are often adjusted so they are most effective. It's not a case where a textbook crossover is likely to give flat response, since the raw frequency response of the driver is not flat. Simulating in software is the fastest way to get to a simple passive crossover that does what it needs to when working with drivers that have uneven frequency response. If you are following someone else's build instructions, these details should already be worked out though. The Celestion crossover is pretty straightforward.
Some people seem inherently drawn to complexity. I've seen this repeatedly at work, so I think it's a thing. Some are also pursuing the flattest frequency response no matter the cost. It's another case where the designer gets to decide what is important to them and how they want to solve the problem.
EQ seems to be gaining popularity again. Some of this is because it's often done in the digital domain now, and if your signal is already digital at some point it's minimally invasive to insert equalization there. It's just a little more software. For normal listening, I think it's a good solution. In the bass range in particular it does things that are hard to achieve without much more cost or complexity. Purists will probably always frown on EQ though.
Automatic EQ can be good or bad. If it's averaging multiple measurement locations and isn't doing anything crazy, it'll probably be OK. The ability to go in and manually adjust things after the automatic routine is useful, in case the automatic result is not to your liking.
Removing breakup issues is typically done with deep, narrow notches, and normally only on the leg of the crossover that has the problem. I don't think a filter like this is as common with automatic correction. Many manually adjusted multichannel DSPs would be able to handle it though.
The farther you push with DSP, the more likely you are to run into latency issues. Commercial implementations tend to have different settings for low latency or maximal correction (both phase and frequency response).
Some people seem inherently drawn to complexity. I've seen this repeatedly at work, so I think it's a thing. Some are also pursuing the flattest frequency response no matter the cost. It's another case where the designer gets to decide what is important to them and how they want to solve the problem.
EQ seems to be gaining popularity again. Some of this is because it's often done in the digital domain now, and if your signal is already digital at some point it's minimally invasive to insert equalization there. It's just a little more software. For normal listening, I think it's a good solution. In the bass range in particular it does things that are hard to achieve without much more cost or complexity. Purists will probably always frown on EQ though.
Automatic EQ can be good or bad. If it's averaging multiple measurement locations and isn't doing anything crazy, it'll probably be OK. The ability to go in and manually adjust things after the automatic routine is useful, in case the automatic result is not to your liking.
Removing breakup issues is typically done with deep, narrow notches, and normally only on the leg of the crossover that has the problem. I don't think a filter like this is as common with automatic correction. Many manually adjusted multichannel DSPs would be able to handle it though.
The farther you push with DSP, the more likely you are to run into latency issues. Commercial implementations tend to have different settings for low latency or maximal correction (both phase and frequency response).
Good morning thank you sincerely for your valuable advice
The combination driver plus horn by Celestion you kindly mentioned look perfect Its is just a matter to place a decent high pass filter before them
Still i am drawn to mod some used PA speaker A more challenging path
Problem is that i cannot do any wood work at all I should buy an already made cabinet
As i have nothing against the horn outside the cabinet on the top i could use at least the cabinet and add Celestion components
I think i will end doing this
The weird thing is that my interest for amplification is collapsed Only speakers from now on
I have already a good integrated to try
you are right And even worse people like me say that they love simple things but then they choose the complex way
The combination driver plus horn by Celestion you kindly mentioned look perfect Its is just a matter to place a decent high pass filter before them
Still i am drawn to mod some used PA speaker A more challenging path
Problem is that i cannot do any wood work at all I should buy an already made cabinet
As i have nothing against the horn outside the cabinet on the top i could use at least the cabinet and add Celestion components
I think i will end doing this
The weird thing is that my interest for amplification is collapsed Only speakers from now on
I have already a good integrated to try
Typically, a PA system, which is really what you are building, would use an electronic crossover or 'LMS' (Loudspeaker Management System) before a separate amplifier for each individual speaker within the system - one each for the sub, midrange, and horn, so six amplifier channels all fed by one crossover, which usually has EQ built-in. Analogue, high-level speaker crossovers generally only exist in low powered/low cost PA speakers.
Further, PA drive units are not designed to have a ruler-flat frequency response and will always need EQ.
Further, PA drive units are not designed to have a ruler-flat frequency response and will always need EQ.
Hi thank you very much for the precious advice I was looking to some Klipsch speakers very well received and quite expensive
Looking at their characteristics they seem more PA than hifi to me
The woofers for instance have fs of around 80Hz No hifi woofer of the same size has a fs so high
They have passive xovers and no eq of course And still people love their sound
so the idea is to follow their approach
Another consideration Can an audio system be good for movies and bad for music ? why ?
and most cinema speakers are of the PA type
Still the soundtracks are played back very nicely Like in some music movies like the Blues Brothers
In any music clubs the sound is through PA speakers and people enjoy it
Looking at their characteristics they seem more PA than hifi to me
The woofers for instance have fs of around 80Hz No hifi woofer of the same size has a fs so high
They have passive xovers and no eq of course And still people love their sound
so the idea is to follow their approach
Another consideration Can an audio system be good for movies and bad for music ? why ?
and most cinema speakers are of the PA type
Still the soundtracks are played back very nicely Like in some music movies like the Blues Brothers
In any music clubs the sound is through PA speakers and people enjoy it
Hi yes It was a provocation
Some years ago i read about a contest DIYers were invited to show their high efficiency speakers to be driven with a same source and amp
The amp was a Sonic Impact T-amp
Some of the best speakers sounded impressively
But you are right This is an extreme
But HE is a good thing to have imho
Some years ago i read about a contest DIYers were invited to show their high efficiency speakers to be driven with a same source and amp
The amp was a Sonic Impact T-amp
Some of the best speakers sounded impressively
But you are right This is an extreme
But HE is a good thing to have imho