Hello,
It's a busy life!I would like to do more wavelets than I have time for..
For a looong time I've been thinking to start a thread on 'How-To' with wavelets, and more specifically how to do it free with Octave
The thing is I have a full audio measurement system coded with Octave and it's about 3000 lines of code written over the years. The wavelet is only a small part of it, the code measures and calculates millions of other things as well. The code is not very modular though, so separating the wavelet from it and making it work universally with any set of data by any user would require rewriting that part.
Meanwhile here's some reading about different time-frequency methods and their possible applications:
http://www.listeninc.com/PDFs/paper_123AES_Time-Frequency_Analysis.pdf
- Elias
its that 3000 lines of code available?
Hello,
It's a busy life!
For a looong time I've been thinking to start a thread on 'How-To' with wavelets, and more specifically how to do it free with Octave
May I ask what user interface you use for Octave and how you would comment on XOctave for the purpose of easy going - or what else you'd possibly suggest ?
Michael
Getting back to the Honk.
So what makes a horn honk? Why does it sound like that?
If we are talking about musical and car/truck/train horns, the reason they honk is resonance. To play a note, they have to resonate. That "amplifies" the sound and gives you a loud noise. At frequencies where they don't resonate, no loud noise, no honk.
Take a bugle for example. It naturally plays just 5 notes - C,G,C,E,G. Those are its natural resonant points. Other notes can be forced, but it is hard work. The bugle has a fixed length and a limited number of notes. If you want more notes, you have to change the length of the horn. Enter the trumpet and its valves, or the trombone with its slide. By changing the length of the tube the resonant points can be changed, thus giving us more of them - and more notes.
Have a look and listen to a truck or train horn. Just one note is used (plus harmonics). The horn and diaphragm are tuned to resonate together, reinforcing the sound. If you lengthen or shorten the horn, you get a very weak sound - it won't honk! I know this well having tried to install a flexible tube between the horn and the diaphragm on one of those horns with the can of compressed air. It didn't work at all. The system has to be tuned.
So what does this have to do with the horns and drivers we use? I contend that if they "honk" then they must be resonating. Just like their musical and automotive cousins. That should show up in FR and impedance plots. And it may show up as resonance that lingers in time (CSD.)
I don't think it is just because they are called horns that we say they honk. It can sound like the honk of a tuned horn.
So what makes a horn honk? Why does it sound like that?
If we are talking about musical and car/truck/train horns, the reason they honk is resonance. To play a note, they have to resonate. That "amplifies" the sound and gives you a loud noise. At frequencies where they don't resonate, no loud noise, no honk.
Take a bugle for example. It naturally plays just 5 notes - C,G,C,E,G. Those are its natural resonant points. Other notes can be forced, but it is hard work. The bugle has a fixed length and a limited number of notes. If you want more notes, you have to change the length of the horn. Enter the trumpet and its valves, or the trombone with its slide. By changing the length of the tube the resonant points can be changed, thus giving us more of them - and more notes.
Have a look and listen to a truck or train horn. Just one note is used (plus harmonics). The horn and diaphragm are tuned to resonate together, reinforcing the sound. If you lengthen or shorten the horn, you get a very weak sound - it won't honk! I know this well having tried to install a flexible tube between the horn and the diaphragm on one of those horns with the can of compressed air. It didn't work at all. The system has to be tuned.
So what does this have to do with the horns and drivers we use? I contend that if they "honk" then they must be resonating. Just like their musical and automotive cousins. That should show up in FR and impedance plots. And it may show up as resonance that lingers in time (CSD.)
I don't think it is just because they are called horns that we say they honk. It can sound like the honk of a tuned horn.
I believe you read my post 33, 40, 51, and 52 and yes it is resonant so you are correct. Did I leave out something? These are the most common reasons for honk.
Ok but if that's the case we should be able to damp the daylights out of them and make it go away. Most of the horns/waveguides I have used are not metal like the older Altec 511/811 horns. You should be able to damp them with clay or another suitable material. That should help.
Have you tried??
Rob
Thanks for the link...which version are you using presently...
I think I remember using a Scalogram back when, no wavlets involved though...
Try damping a trumpet with clay. Will it still play? Yes. It will sound different, but it will play. The most important factors are shape and length, not the material.
I am currently upgrading to 8.5.Thanks for the link...which version are you using presently...
[If wavelets were in 7.x, I never stumbled upon them in 5 years of using it....
]
Last edited:
I did read those. Good stuff. I will have to try the Dayton H110 1" horn.
My post was to relate this directly to the honk heard in other horns, which had not yet been discussed directly here.
Question: If we were to extend the diameter of the bugle bell (mouth) to about 2 feet, what would happen to the honk? That would be ~1/2 wavelength of middle C. Would the bugle then not resonate as much?
Were are not talking trumpets. That's a long tuned tube designed to resonate. With horns you have just the bell without the tube. Sure you have a driver there and a short throat but if you look at the lengths of the "tubes" they seem kinda short to be an issue at cut off or the lower ranges where I think most of the "honk" is. If it was the air resonating inside the horn causing the honk with the walls acting like the walls of the pipe all you would have to do is run sine sweeps to find the resonance. Not only that but they all would honk including Earls Waveguides if it was that simple.
Rob
I"m not sure about that Rob.
The hand held alarm horns as well as truck, train and boat air horns are not very long. Some horns we use are as long or longer. The W.E.-15A is 15 feet long!
But what the musical horns do have is a small mouth and bore diameter compared to the tube length. Perhaps that is important?
The hand held alarm horns as well as truck, train and boat air horns are not very long. Some horns we use are as long or longer. The W.E.-15A is 15 feet long!
But what the musical horns do have is a small mouth and bore diameter compared to the tube length. Perhaps that is important?
are we in lack of resonance seen in this plots ?
The problem IMO is more to keep it "on display" when fading out by measures of round over or foam dampening – as obviousely our ear brain system is way more sensible in picking up that pattern than we can easily visualize for now with more real world horns
One reason might be that FR until the first echo (mouth reflection) is seriously different to what comes after - I do not know
We still are in a need for better visualization in that short time span yet.
Michael
The problem IMO is more to keep it "on display" when fading out by measures of round over or foam dampening – as obviousely our ear brain system is way more sensible in picking up that pattern than we can easily visualize for now with more real world horns
One reason might be that FR until the first echo (mouth reflection) is seriously different to what comes after - I do not know
We still are in a need for better visualization in that short time span yet.
Michael
Last edited:
I've seen it for at least a year, I believe.
Somebody there is very much up on this art....
Hello Panomaniac,
The methods to design horns for brass instruments and horns for musical reproducing differ a lot.
For musical instruments, we are looking for controlled resonances at multiple frequencies. A good reading on that topic is Noreland's thesis.
"Numerical Techniques for Acoustic Modelling and Design of Brass Wind Instruments."
by Daniel NORELAND
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.1.3417&rep=rep1&type=pdf
Then for musical reproducing there is no need for resonance to obtain an increase in sound intensity with a horn. What leads to the increase of sound intensity is the acoustical loading of the diaphragm. The horn acts as a transformer and allows an optimal adaptation of the impedance. The efficiency of a driver mounted on an excellent horn can reach 50% (e.g WE555 on a WE15A horn between 500 and 1000Hz).
If you try to use a Le Cléac'h horn as a trumpet you'll be deceived - and try to produce different notes will be difficult as there is very few reflection from mouth to throat- this means no interference between the direct wave and the reflected wave (reflection only happens near the acoustical cut off and below) - and no resonance.
Then, horns having a truncated mouth used in musical reproducing induce the reflection of an important part of the energy from mouth (or whatever rapid profile change) to the throat and the resulting interference with the direct wave (or another reflected wave of order 2, 3...) leads to resonance and honkness. This phenomenon is more important when the frequency decreases toward the acosutical cut-off as there is more reflected energy in that case.
But near the cut-off (defined as the frequency above which the resistive part of the throat impedance becomes larger than the reactive part of the throat impedance) there is also a rise in group delay and if we have a note the fundamental of which has a ferquency near the cut-off then it is delayed versus the harmonics of that note and the timber is altered (and the image of the sound source too). This is IMHO another face of the honkness (a Le Cléac'h horn used with a high-pass crossover the -3dB frequency of which is too low can present that timber alteration, while there will be no resonance).
Best regards from Paris, France
Jean-Michel Le Cléac'h
The methods to design horns for brass instruments and horns for musical reproducing differ a lot.
For musical instruments, we are looking for controlled resonances at multiple frequencies. A good reading on that topic is Noreland's thesis.
"Numerical Techniques for Acoustic Modelling and Design of Brass Wind Instruments."
by Daniel NORELAND
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.1.3417&rep=rep1&type=pdf
Then for musical reproducing there is no need for resonance to obtain an increase in sound intensity with a horn. What leads to the increase of sound intensity is the acoustical loading of the diaphragm. The horn acts as a transformer and allows an optimal adaptation of the impedance. The efficiency of a driver mounted on an excellent horn can reach 50% (e.g WE555 on a WE15A horn between 500 and 1000Hz).
If you try to use a Le Cléac'h horn as a trumpet you'll be deceived - and try to produce different notes will be difficult as there is very few reflection from mouth to throat- this means no interference between the direct wave and the reflected wave (reflection only happens near the acoustical cut off and below) - and no resonance.
Then, horns having a truncated mouth used in musical reproducing induce the reflection of an important part of the energy from mouth (or whatever rapid profile change) to the throat and the resulting interference with the direct wave (or another reflected wave of order 2, 3...) leads to resonance and honkness. This phenomenon is more important when the frequency decreases toward the acosutical cut-off as there is more reflected energy in that case.
But near the cut-off (defined as the frequency above which the resistive part of the throat impedance becomes larger than the reactive part of the throat impedance) there is also a rise in group delay and if we have a note the fundamental of which has a ferquency near the cut-off then it is delayed versus the harmonics of that note and the timber is altered (and the image of the sound source too). This is IMHO another face of the honkness (a Le Cléac'h horn used with a high-pass crossover the -3dB frequency of which is too low can present that timber alteration, while there will be no resonance).
Best regards from Paris, France
Jean-Michel Le Cléac'h
Last edited:
From what the driver or the "horn"?? Where are the impeadence and phase plots for the driver unmounted and on the "horn"?? Are they gated plots or do they have the room reflections contaminating them?? Are they a monopole or dipole set-up?? Hard to "see" what's there if you don't know what you are looking at.
Rob
Interesting ETC comparison between 90deg OS (no lip rounding), other configuration, and direct radiating driver.
It seems OS is closest to direct radiating driver in the ETC. However, the acoustic loading would be less than ideal. I assume the foam that Earl uses will improve the loading. Now I am more interested in listening to the Summa's.
It seems OS is closest to direct radiating driver in the ETC. However, the acoustic loading would be less than ideal. I assume the foam that Earl uses will improve the loading. Now I am more interested in listening to the Summa's.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.