So I've searched all around and couldnt find an answer to this question: Does directivity decrease or increase with amplitude?
In other words, say you're using a 1st order crossover at 1038hz with a 15" driver with an sd of 855cm. Thats a piston diameter of 13" roughly,
or a Q of 10 at 1038hz. Being that its amplitude down 9db at 2076hz, does the directivity decrease with increasing frequency or stay the same?
In other words, say you're using a 1st order crossover at 1038hz with a 15" driver with an sd of 855cm. Thats a piston diameter of 13" roughly,
or a Q of 10 at 1038hz. Being that its amplitude down 9db at 2076hz, does the directivity decrease with increasing frequency or stay the same?
Last edited:
Ah ok, thanks. That would explain I suppose why some people prefer using steep crossover slopes. Seems easier to match directivity between drivers if thats the case. I had wondered if it was possible to create a smooth directivity transition between two drivers by using the decreasing amplitude slope of the crossover to add to the directivity sum of the next driver.
I'm not sure what you mean by this, but have you also considered taking care of breakup?
Absolutely. I usually try to get out of a driver before the breakup of course.What I meant by my previous statement (and you answered for me) is that I was trying to figure out how to calculate the total directivity sum of two drivers before and after a crossover point in order to achieve the smoothest directivity transition.
Hi, amplitude can vary while directivity stay the same. If you turn up volume on the amplifier directivity of the speaker doesn't change.
Directivity changes with frequency and is the same for single driver on a given construct regardless of crossover slope, or amplitude. You can change directivity of single driver speaker by manipulating the construct around it, or changing the driver to different size, different shape, also the mentioned cone breakup would differ some.
On a multiway speaker where there is two (or more) different transducers, that usually are of different size and different physical location, the combined system directivity would change with the used crossover, around the crossover bandwidth, as sound from the two sound sources interfere. Interference is related to the physical construct: where the transducers are located and what their environment is. Interference depends on at which angle and distance you observe the system. Its all due to sound wavelength and path length difference from each sound source to the point of observation.
Directivity Index typically refers to sound power directivity index, which is calculated from sound towards some (on-axis) direction subtracted by power response. Power response is usually weighted average response to all directions. These are defined more accurately in CTA-2034-A measurement standard you can download here https://members.cta.tech/ctaPublicationDetails/?id=367c1410-7212-e811-90ce-000d3a004988&reload=timezone &reload=timezone
Easiest way to get the power response and DI is to use VituixCAD, read its manual how to measure your speaker and how to process and load the data into the program to get the graphs. https://kimmosaunisto.net/ Basically you measure each driver separately, in the enclosure they are in, and load the data into VituixCAD. There you can inspect each of the drivers alone if you wish, or the combined system response and what the effect of crossover is to system directivity.
Directivity changes with frequency and is the same for single driver on a given construct regardless of crossover slope, or amplitude. You can change directivity of single driver speaker by manipulating the construct around it, or changing the driver to different size, different shape, also the mentioned cone breakup would differ some.
On a multiway speaker where there is two (or more) different transducers, that usually are of different size and different physical location, the combined system directivity would change with the used crossover, around the crossover bandwidth, as sound from the two sound sources interfere. Interference is related to the physical construct: where the transducers are located and what their environment is. Interference depends on at which angle and distance you observe the system. Its all due to sound wavelength and path length difference from each sound source to the point of observation.
Directivity Index typically refers to sound power directivity index, which is calculated from sound towards some (on-axis) direction subtracted by power response. Power response is usually weighted average response to all directions. These are defined more accurately in CTA-2034-A measurement standard you can download here https://members.cta.tech/ctaPublicationDetails/?id=367c1410-7212-e811-90ce-000d3a004988&reload=timezone &reload=timezone
Easiest way to get the power response and DI is to use VituixCAD, read its manual how to measure your speaker and how to process and load the data into the program to get the graphs. https://kimmosaunisto.net/ Basically you measure each driver separately, in the enclosure they are in, and load the data into VituixCAD. There you can inspect each of the drivers alone if you wish, or the combined system response and what the effect of crossover is to system directivity.
Last edited:
Yeah think about sound wavelength. Physical objects, like loudspeaker enclosure, whose size is roughly size of the wavelength have huge impact on how the sound propagates, reflection and diffraction happens. Bafflestep might be a familiar thing. When wavelength is much longer than size of the speaker the speaker is practically invisible to it, has no impact on sound propagation.
Here is simplified thought experiment that should help you get started reasoning the stuff by yourself 🙂
100Hz is 3.4 meters long, 1000Hz is 34cm, 10000Hz is 3.4cm. From this we can reason that a loudspeaker whose size is rougly 30cm has less and less effect on how sound radiates below ~1000Hz and more effect on sound propagation above ~1000Hz. If you imagine it was a fine loudspeaker whose on-axis frequency response is flat, what the power response would be? Roughly below 1000Hz sound would propagate the same to all directions since the box / transducer has less and less effect on it due to being small in comparison to wavelength and response to any direction would eventually be the same, so power response (average response to any direction) would be roughly the same as on-axis response. Basically DI would approach 0 as wavelength gets longer. Roughly above 1000Hz the box affects how sound radiates and for example you could imagine there is "shadow" of sound behind the box, less high frequencies are propagating behind the box as most is reflected forward toward the on-axis. Also the transducers would "beam" sound forward with increasing frequency, when wavelength approaches diameter of the transducer (same phenomena, interference with wavelength and path length). Now sound power would be less than on-axis response, DI would be > 0. From this we get that loudspeaker DI rises from 0 at some frequency to some db number at some high frequency and its all related to physical size and shape of the system.
How any of it relates to perceived sound in room? you'd probably want to reason backwards from psychoacoustics to room acoustics to how a speaker should radiate at the position it happens to be in, in order to make maximally positive sensation in the brain. When you do this you can approximate what kind of physical construct with some xo would lead towards it in your application, in your room and positioning of the speakers and listener.
Its much more info than what you asked for 😀 Hopefully enough for you to think about the stuff and get toward good results relatively fast. Have fun!🙂
Here is simplified thought experiment that should help you get started reasoning the stuff by yourself 🙂
100Hz is 3.4 meters long, 1000Hz is 34cm, 10000Hz is 3.4cm. From this we can reason that a loudspeaker whose size is rougly 30cm has less and less effect on how sound radiates below ~1000Hz and more effect on sound propagation above ~1000Hz. If you imagine it was a fine loudspeaker whose on-axis frequency response is flat, what the power response would be? Roughly below 1000Hz sound would propagate the same to all directions since the box / transducer has less and less effect on it due to being small in comparison to wavelength and response to any direction would eventually be the same, so power response (average response to any direction) would be roughly the same as on-axis response. Basically DI would approach 0 as wavelength gets longer. Roughly above 1000Hz the box affects how sound radiates and for example you could imagine there is "shadow" of sound behind the box, less high frequencies are propagating behind the box as most is reflected forward toward the on-axis. Also the transducers would "beam" sound forward with increasing frequency, when wavelength approaches diameter of the transducer (same phenomena, interference with wavelength and path length). Now sound power would be less than on-axis response, DI would be > 0. From this we get that loudspeaker DI rises from 0 at some frequency to some db number at some high frequency and its all related to physical size and shape of the system.
How any of it relates to perceived sound in room? you'd probably want to reason backwards from psychoacoustics to room acoustics to how a speaker should radiate at the position it happens to be in, in order to make maximally positive sensation in the brain. When you do this you can approximate what kind of physical construct with some xo would lead towards it in your application, in your room and positioning of the speakers and listener.
Its much more info than what you asked for 😀 Hopefully enough for you to think about the stuff and get toward good results relatively fast. Have fun!🙂
Last edited: