It won't be easy to simulate the open baffle in front of the woofer. I'd want to take polar measurements on your prototypes, and to keep an open mind.
The things you ask are quite easy to do. There is a good place for hornresp questions here - https://www.diyaudio.com/community/threads/hornresp.119854/unread
As tmuikku noted, WRT WL dimensions, he's referring to the circumference (fundamental) of my 'bubble', i.e. we measure its circumference as linear Vs my choosing to view it as the diameter because it gives (me) a more accurate 'snapshot' of the wave as it's moving over/around/through the room/whatever, so generally speaking I don't need a complex computer sim of it for basic design work.
Note that 'round' presumes a circular driver, horn, whatever whereas the radiator + any baffle, near room boundaries shapes the 'bubble' over 'x' distance, then as the signal weakens it collapses closer to round, but AFAIK this is so far away to be moot for even most prosound apps and even then only in the HF and why long throw horns (trumpets) are added to the normal system to limit it, i.e. super tweeters in the truest sense.
Last edited:
Need to learn TL theory since horns are 1/2 WL resonators closed cones, but must be at least a 1/4 WL long, closed cylinder fundamental.
Comparison
"""The drive unit is effectively coupled to a long column of air which has mass. This lowers the resonant frequency of the drive unit, negating the need for a highly compliant device. Furthermore, the column of air provides greater force on the driver itself than a driver opening onto a large volume of air (in simple terms it provides more resistance to the driver's attempt to move it), so to control the movement of air requires an extremely rigid cone, to avoid deformation and consequent distortion. ""
Does this applies to horns/guides similarly but to varying degrees ( obviously the driven air space is more open in a horn than in a straight cylinder/box ?? )
@GM Not sure i understand why the cone has a standing wave similar to the open cylinder , is it because the reducing volume at the apex forces the node to be at this specific point ?
So this kinda make sense, but how does it affect horns/waveguides ?
The sound wave produced by a driver at the tip of the cone is moving through it with time,
unsure how standing waves affect this other than creating undesirable resonances ?
Does this applies to horns/guides similarly but to varying degrees ( obviously the driven air space is more open in a horn than in a straight cylinder/box ?? )
@GM Not sure i understand why the cone has a standing wave similar to the open cylinder , is it because the reducing volume at the apex forces the node to be at this specific point ?
So this kinda make sense, but how does it affect horns/waveguides ?
The sound wave produced by a driver at the tip of the cone is moving through it with time,
unsure how standing waves affect this other than creating undesirable resonances ?
Spacing is related to interference. If you have two point sources and space them apart an interference pattern emerges at any observation angle (except one, the main axis) and when ever path length difference from the two sources is 1/2 wavelength there is a dip or null, destructive interference as the signals are out of phase at that wavelength and cancel out at that frequency. Similarly, when the path length difference is multiple of full wavelength the signals are in phase and the interfere constructively. Key realization here is that wavelength changes all the time as frequency goes from low to high thus resulting interferemce pattern looks like a comb.
Now, 1/4 wavelength distance is just roughly the turning point when the combfilter starts. Below this frequency there is only constructive interference and soon above there is first null of combfilter which continues all the way to high frequencies, nulls and peaks alternate.
On a multiway speaker the ways bandwidths overlap some, both reproduce sound at crossover frequemcy and around it and here the above scenario happens. In addition to path length difference there might be delay from crossover for example. Also, there might be multiple drivers, like a two way speaker with 2 woofers and a tweeter, woofers would both reproduce same frequency bandwidth and tweeter joins in at some point.
1/4wl is some distance in cm or inch and also in time, depending on what frequency you look at, calculated from speed of sound.
How sound interacts with objects like baffle or horn depends also on frequency / wavelength. When wavelength is short in comparison to the size of the object the sound tends to reflect and when its long it diffracts (goes around). Simplified. I have no deep understanding of wavefronts vs. horn profile other than whats observable from messurements and simulations. Any abrupt change in the profile, like sudden change in profile or mouth without rollback there is big change in impedance and part of sound transmits and the rest reflects back.
Anyhow, with imagination one can get far for understanding various phenomena, seeing things happen although sound is invisible. Wavelength is key gerring into it.
Good stuff @tmuikku , much appreciated.
So When the angle/flare of horns gets open to fast, the impedance ( because of the lower restriction ) changes which then acts similarly or to some varying degree, as the open end cone/tubes and part of the waves gets sent back as a resonance while another part starts enlarging around the opening/walls ?
would that be a good assumption ?
As fort the overlap on multiway ..totally makes sense .thanks for the clarification.
I knew about interference but didn't realize how it was linked here.
Would i be correct in assuming that if there was no overlapping of frequencies at all ( theoretically ) the 2 drivers wouldn't then interact and cause interference ? If this is correct, then crossover with very steep filter will lessen the interferences, while those with very " slow " pass will make it worse ?
The distance between the drivers ( delay ? ) would also cause similar effect ( additionally ? ) then ?
I think you are right, similarly to GM's statement, that visualizing the waves and understanding the wavelength stuff is kind of mandatory .
So When the angle/flare of horns gets open to fast, the impedance ( because of the lower restriction ) changes which then acts similarly or to some varying degree, as the open end cone/tubes and part of the waves gets sent back as a resonance while another part starts enlarging around the opening/walls ?
would that be a good assumption ?
As fort the overlap on multiway ..totally makes sense .thanks for the clarification.
I knew about interference but didn't realize how it was linked here.
Would i be correct in assuming that if there was no overlapping of frequencies at all ( theoretically ) the 2 drivers wouldn't then interact and cause interference ? If this is correct, then crossover with very steep filter will lessen the interferences, while those with very " slow " pass will make it worse ?
The distance between the drivers ( delay ? ) would also cause similar effect ( additionally ? ) then ?
I think you are right, similarly to GM's statement, that visualizing the waves and understanding the wavelength stuff is kind of mandatory .
I think so this would be the rough generalization. In the end any resonance on a loudspeaker system would color sound for better or for worse and in this light this is enough information to try judge if a horn is better or worse in this respect, although in reality the audible difference could be small. My current waveguide has diffraction slot and doesnt have mouth rollback and it sounds just fine although next one will perform better and perhaps sound better so I preserve rights to change my mind on this 😀
Yes thats correct and steep filters are doable with DSP with cost of delay (FIR), which might be fine for music playback but bad for movie lipsync or doing work with sound like recording. Steep IIR filters make lots of group delay which can have effect on perceived sound quality. And one can do just fine with low order filters like most speakers typically have, because
yeah this, if the drivers on both sides of the crossover can physically fit close enough not to cause too bad interference all is fine. Or slide the crossover or driver spacing to adjust the interference to various directions.
Loudspeakers radiate sound to all directions and if listening inside a room most if not all of it reaches the ear at some point and hearing system processes some kind of perceived sound from it. Some directions are not as important as others so if there is going to be interference one could try and take advantage of the interference (line array for example) or to point bad stuff to directions that are less important for perceived sound quality. Tilt and toe in, positioning relative to walls and objects both the speakers and listening position are part of this. Radiation pattern both horizontal and certical, the directivity.
only if you want to go bottom of it, to the deep end 😀 its lots of fun, trying to figure stuff out and then to reason over the knowledge and make improvement. Surely a key thing to tie sound to physical objects, to loudspeakers and to rooms. Dont forget to study the hearing system as well! and have fun! 🙂
ps. simulations will help a lot, sometimes question at hand is just too complicated for imagination to tackle but even then ability to imagine plausible scenario makes simulations easy to do and test if one can kind of imagine the basics, or options. Also analyzing simulation results gets more educated with ability to try and imagine what features in the simulated (or messured) object makes the features on the simulation result graphs (or in real measurements). Educated guess for improvements speeds up the process and help get better results on the next sim/prototype. All of this feeds back to the ability to imagine the stuff.
Next level would be an ability to imagine how all of it actually sounds in the application, how to translate graphs to sound, but this would need lots of experience of different systems. Perhaps some one can do this as well. Experiment like crazy, ask questions, seek answers, dont be afraid to make mistakes and be wrong on things, that equals learning, progress and I see you on this path already with all the questions 🙂 Im writing confidently because of this, if someone knows better I hope he will eventually come and correct and we will all learn. As disclaimer, take my messages as is 🙂
Last edited:
Again much appreciated !! I am always on the learning path... not really quick in this stuff and i keep forgetting everything cause i get into other things for a few years then need to come back and relearn lol .
I am slowly trying to finish Dr. Toole book, for the 2nd time lol
but at least this time i understand most of what he is talking about, in contrast with the first reading a few years back, which was quite fuzzy.
Interesting FIR crossover info i just found :
https://www.prosoundtraining.com/2013/09/27/fir-ward-thinking-part-2/
have a good one !
I am slowly trying to finish Dr. Toole book, for the 2nd time lol
but at least this time i understand most of what he is talking about, in contrast with the first reading a few years back, which was quite fuzzy.
Interesting FIR crossover info i just found :
https://www.prosoundtraining.com/2013/09/27/fir-ward-thinking-part-2/
have a good one !
Zingali Twenty Evo
the bottom driver is a 12" supposedly, so this is not bookshelf size thing.
So hmmm .. why the spacing between the horn and the woofer is so large ?
It has to be aesthetics
and the front cabinet edges could've use a little more of a filet edge perhaps ?
Just so i can start to notice the shape of horns/waveguides , what type of angle do you guys think this is ?
Looks like is it has a first section with a lower angle and then flares open up to the mouth ?
I think i've seen quite a few horns that seem to use different profile types in the mouth to the flare section
VS the rest of the horn , is this something that has a specific use ? limit diffraction and resonances ?
Lastly for tonight ( i'm tired , covid for almost 2 weeks now and i'm 3x vaxxed ...headaches, weakness and now starting to cough again )
I'm trying to think about the relation about the expansion of the wavefront and how it is restricted by the horn walls VS the volume at different time
Does the wavefront behave similarly in horns of proper sizes, for HF and for mid/low ?
If the frequency is high enough, doesn't the waves start reflecting off the walls ?
Or is the following wavefront pushing it out and making it behave or something ...
I guess theoretically would be possible to produce a single wave with a sound test and then there would be nothing to follow it
so it must be doing its own thing then ... hmm
Trying to use Hornresp wave simulator but it behaves weirdly i think ....
Specs says 1 kHz/2nd XO, horn 120 deg; a 12" is ~120 deg/1 kHz, so in theory if the spacing is set based on measurements to this, which it apparently is, then no comb filtering in the entire 1 kHz - up pass-band.
Not enough info beyond it appears to be a basic conical horn like the Danley Synergy except round, so just another WG.
Bummer! 🙁 Hope you can 'shake' it PDQ. A shame the rush to get vaccines has rendered them a 'hit or miss' affair. 🙁
Correct, you can scale based on BW WL size.
As asked, no, so need to define the design, etc., though in general as long as there's an included angle > ~12 deg, no 'slap' echo (reflections) can form.
Dunno, never used it, just looked at other's.
You may be right, but it's difficult to be sure. While it must be properly considered and accounted for in various ways, c-c spacing is an often misunderstood thing.
Just guessing, but it's not unlike a truncated tractrix. This family of horns are often well terminated.
Thanks again both of you. Appreciated.
So it seems i don't understand the CC spacing ...
i'll play with this :
https://www.falstad.com/interference/
and xdir to try and understand better ...
having tools to learn makes it much easier 🙂
thanks again
So it seems i don't understand the CC spacing ...
i'll play with this :
https://www.falstad.com/interference/
and xdir to try and understand better ...
having tools to learn makes it much easier 🙂
thanks again
18" is about 1.5wl at 1kHz, this should make nice smooth power response, vertical sidelobes yes. Then additionally the horn is probably at ear height while the woofer is midheight of the box, which would reduce/eliminate the lowest mode inside the box, well below 1kHz so on the woofers pass band and the hardest one to kill with damping material inside.
Of course I have no idea what the designer thought, how the speaker measures and sounds but it seems to be very reasonable layout, something that one would end up if looking at cta-2034 "ideals" and tweaking things in a simulators.
Of course I have no idea what the designer thought, how the speaker measures and sounds but it seems to be very reasonable layout, something that one would end up if looking at cta-2034 "ideals" and tweaking things in a simulators.
You're welcome!
Based on 18" puts spacing at ~752 Hz, so maybe based on effective baffle width and/or horn's -F6.
Also much easier to learn the wrong thing, so try to confirm what you are seeing.
OK so played with XDIR quite a bit ..was super fun 🙂
So i used this Linkwitz Riley crossover graph ( hope it is somewhat accurate ) as i've seen this type of filter around quite often...probably popular
and tried the reference points i noted in XDIR .
this is the base interference @ 1khz with different wavelength spacing for the drivers
and second line was just playing to see where can we get back to near 0db which ended up matching the LR2 fitler i guess
When i look at those first graph, from my very inexperienced POV, i'd conclude that anything over 3/4WL is quite risky as it will boom on the floor and ceiling more than the forward firing..also only +- 15degree
I'd guess also that something near to the 1/2WL could be per design as this would give some good directivity ? is this something that designers actually use in their loudspeakers ?
So i used this Linkwitz Riley crossover graph ( hope it is somewhat accurate ) as i've seen this type of filter around quite often...probably popular
and tried the reference points i noted in XDIR .
this is the base interference @ 1khz with different wavelength spacing for the drivers
and second line was just playing to see where can we get back to near 0db which ended up matching the LR2 fitler i guess
When i look at those first graph, from my very inexperienced POV, i'd conclude that anything over 3/4WL is quite risky as it will boom on the floor and ceiling more than the forward firing..also only +- 15degree
I'd guess also that something near to the 1/2WL could be per design as this would give some good directivity ? is this something that designers actually use in their loudspeakers ?
Then i tried to play again in XDIR with the values from the LR filter graph
WOOFER SIDE :
and then HF side :
Lookin at the graphs i would've think the HF side would've been influenced differently because of the shorter spacing in the frequency on the filter
but it seems the difference is less than anticipated.
In this particular setup i think we can conclude that the mids are less affected as even the 1 wave length spacing doens't look awful
The HF side though has problems for around 100hz ( which is not that much but what do i know )
and a 1/2WL or shorter would be the better solution ?
May i ask for correction here ? i'm probably wrong on most of it lol
And also if you would give different conclusions or pull different info from this thing, i'd very much like to read it ! 🙂
This quick test was still very fun and informative i think.
Ah yeah, i played with phase somewhat, doens't seem to have as much effect as i would've anticipated,
but anyhow 2022 = digital so phase shouldn't be much of a problem right ?
( would phase simulate driver spacing axially ? )
WOOFER SIDE :
and then HF side :
Lookin at the graphs i would've think the HF side would've been influenced differently because of the shorter spacing in the frequency on the filter
but it seems the difference is less than anticipated.
In this particular setup i think we can conclude that the mids are less affected as even the 1 wave length spacing doens't look awful
The HF side though has problems for around 100hz ( which is not that much but what do i know )
and a 1/2WL or shorter would be the better solution ?
May i ask for correction here ? i'm probably wrong on most of it lol
And also if you would give different conclusions or pull different info from this thing, i'd very much like to read it ! 🙂
This quick test was still very fun and informative i think.
Ah yeah, i played with phase somewhat, doens't seem to have as much effect as i would've anticipated,
but anyhow 2022 = digital so phase shouldn't be much of a problem right ?
( would phase simulate driver spacing axially ? )
High ceilings? In any case, it depends what you are doing. Would you rather a hole?
Was phase ever a problem?
If you mean can you create those lobing patterns by manipulating phase, no you can't but you can cause variations. It therefore depends on what it is you're trying to do.