The Talking Machine Forum — For All Antique Phonographs & Recordings

The large drawing appears to be a version of a folded orthophonic horn. That is a far more complex method of reflexive sound control which is more aimed at making a 17 ft long horn fit into a more compact arrangement, which is suitable for a phonograph that will fit into a living room.

In the smaller drawing, it is not clear whether that reflector is a cone, or a fold. The following will work for either style, but a cone shape would be more efficient. That is, much like the reflective cone at the center of an electrically amplified bullhorn, that little reflector could be doing just what its name implies. Reflecting back SOME of the sound. Depending on the angle of the horn and of the reflector, it could be lengthening the distance to the horn exit, at least for a fraction of the sound energy. According to Helmholz laws, the longer a tubular device is, the lower its resonant frequency. Therefore, it is conceivable that the little reflector has a small influence on the lower frequencies.

Since you describe it as reducing the tinny-ness, it could be performing a second function. That is, it could be changing the phase of certain frequencies, by reflecting those phase changed sounds back into the throat of the horn, thereby reducing or even cancelling, some sound(s). To be most noticeable, this sort of trick requires some tuning, or adjusting to optimize the affect.

Startgroove, I'm sorry to say but your post contains many misconceptions about acoustical physics.

In the "large" drawing it is pictured a folded and split orthophonic horn. There are no soundwave reflections, and actually sound reflections are avoided at all costs in orthophonic horns.

Megaphones ("electric bullhorns") are in turn a three-times folded exponential horn. Again, there are no soundwave "reflections", just bends. Perhaps you can figure it out better by the sectioned drawing attached to this message.

Finally, horns are not resonators: they are impendance adapters. Hence, Helmholtz principles don't apply. If a conical/exponential cone, without modifying its original contour, is tapped at the mouth with an object of whatever size and shape, as first thing the lower frequencies will be destroyed by acoustical impedance mismatch.

Wow! It seems to me you do not get it. My suggestions are purely speculative, yet valid. I'm not going to cite formula's or other too technical data here, rather I'll keep this discussion uncomplex. Besides, it does not take a scientific mind to understand that any hard surface is a sound reflector, and it is not hard to understand that sound reflects (bounces, if you prefer) off of a hard surface.

See: https://www.nde-ed.org/EducationResourc ... ection.htm

If it didn't reflect, we would never hear an echo, and there would never be a need to foam line a sound studio, and Sonar could not detect objects in water. Since a sound wave normally travels in a relatively straight direction, it must reflect off of the interior surfaces of a bent horn, such as in a Victor back mount phonograph, several times before it exits the bell. (Note: We are not considering the pressure wave here).

I think your second supposition is just as flawed. Possibly you meant to say, "impedance adapter". Which more accurately should be called a loading column.

Again, it is well known that a tubular device exhibits a resonant frequency, and such frequency is dependent on length and diameter, and the viscosity of the fluid through which sound waves propagate (in this case air). Without this principle, neither a trombone, a tuba nor any of the other horn musical instruments would work. They all vary the length or diameter of the loaded column to change the frequency of the note, or the resonance. The longer, or larger the diameter, of a column of air in those instruments, the lower their resonant frequency. The shorter the length, and smaller the diameter, of the column of air, the higher the frequency.

Further proof of this is with the phonograph horn itself. A small "witches hat" horn has good high frequency capabilities, but very poor low frequency reproduction ability, while a longer and larger diameter orthophonic type horn "passes" the lower frequencies quite well. Although, with phonograph horns, resonance is an unwanted effect.

Feel free to pass to fomulas: I am M.Sc. in Physics.

By "impedance adapter" I meant exactly what I have written: impedance adapter.

A gramophone horn is NOT, by any means, used as a "resonating column". Your trombone example is completely off and shows a severe misunderstanding of how gramophone horns act as impedance adapters. It is just like comparing an acoustic guitar body with a loudspeaker cabinet: both are made with wood and both deliver sound, but the first shall resonate, while in the second resonances are to be avoided at all costs.

The reason why witches' hat horns are unable to deliver a good bass response is because they have a contour far from the ideal one, and hence their impedance adaptment properties are only modest. They are also, in general, too short.

How impressive, you have a degree and so do I. However, and again, we don't need to go beyond the basics of acoustic dynamics to have an understanding of this subject.

Yes, you meant "impedance adapters", but you wrote "impendance adapters", therefore I corrected you.

With your education, I would think that you would understand exactly what I am saying. However, since you seem to misunderstand this, I will patiently describe it in more detail. While it is true phonograph horns are not intended as resonators, they are resonators inadvertently, though they are not tuned to a frequency that is imposed on them. You cannot take away the resonant properties of any object, it is always there.

All horns, as you correctly pointed out, have a quality called "Impedance". That can be defined mathematically as: "Acoustic impedance, which has the symbol Z, is the ratio of acoustic pressure p to acoustic volume flow U. So we define Z = p/U." See (1) below. To reduce that to simplicity, it means that impedance usually varies strongly when you change the frequency. In other words, the acoustic impedance at a particular frequency indicates how much sound pressure is generated by a given air vibration at that frequency.

From Wikipedia: "We can specify the z of air or of water. The acoustic impedance Z is the property of a particular geometry and medium: we can discuss for example the Z of a particular duct filled with air. Usually, Z varies strongly when the frequency changes. The acoustic impedance at a particular frequency indicates how much sound pressure is generated by a given acoustic flow at that frequency."

In other words, frequency and impedance are interactive, and inversely related. "When a sound source transfers its energy to a medium, the medium opposes the movement of the source with some kind of average impedance that is dependent not only on the medium, but also on the size of the air mass pushed by the sound source." Therefore, larger horns, having more air mass, will have more impedance to a given frequency. That directly links frequency response to the size of horns. (2)

"Resonating column" is not a term that I used at all, you did. However, we'll go with that for now. When a resonant object is not tuned to the frequency that is applied to it, that object will resonate inefficiently. When that object resonates at its natural frequency, that object does not attenuate the amplitude of the frequency. The longer and larger a phonograph horn is, the closer it gets to the resonant frequency of the lower notes. However, to be resonant at 20 Hz, the horn size would have to be so enormous that it would be too large to fit into a living room. Therefore, a phonograph horn that efficiently produces the lower bass range, could never be a practical item. Yet, the smaller horns, such as a witches hat horn, are definitely able to optimize the volume of a range of notes which are around, or near, that horns resonant frequency. By the way, the resonant frequency I am speaking of is not that of the material the horn is made of, rather it is the resonant frequency of the column of air within, which is activated upon by pressure waves emitted from the diaphragm.

In your comparison of a guitar body and a loudspeaker cabinet, did you forget that a bass reflex speaker cabinet is designed to accentuate lower frequencies through tuned resonance? In that case, the speaker is the diaphragm, and the port is the 2nd exit for bass notes. That is the "resonating column" you mention. Did you also forget that a guitar produces notes in the frequency range from 80 to 15KHz, therefore its resonance should not be tuned? If a hollow body guitar were to resonate in the tuned fashion that I am talking about, then one range of frequencies would be louder than all the rest. Therefore, a hollow guitar body is NOT tuned to any specific frequency, and besides, it does not contain a "resonating column", rather an open chamber. Significant difference there.

When you say, "The reason why witches' hat horns are unable to deliver a good bass response is because they have a contour far from the ideal one, and hence their impedance adaptment properties are only modest. They are also, in general, too short", you are confirming what I say. Although the use of the word "contour" is not quite accurate. Perhaps you meant equalization curve? That is yet another facet that could be incorporated into this discussion if you like.

References cited: (1) http://newt.phys.unsw.edu.au/jw/z.html (2) http://www4.uwsp.edu/physastr/kmenning/ ... edance.pdf

If you would go to the reference you provided, and then follow the link to the page "How do woodwind instruments work", and then to the paragraph titled "The air column determines the pitch" you will find a quite accurate description of how sounds are generated in a woodwind instrument. In two words, the player has to generate standing waves into the instrument's duct, and the length of the duct has to be changed in order to change the pitch. (forget about harmonics for a while).

This is mostly the opposite of what happens in gramophone horns: there are no standing waves to be generated, but almost contrarywise there is a pressure wave that propagates from the throat to the mouth. Also, the length of the horn has not to be changed in order to change the emitted pitch, because it is not a pitch generator, it is an impedance transformer.

If you fail to grasp these two fundamental differences, we will go on arguing forever but to no avail.

I agree that reflex cabinets can be mathematically treated as Helmholtz resonators, but again a reflex cabinet is not used to generate standing waves. All the contrary, standing waves have to be avoided at all costs in a reflex cabinet. If you want, you can think at a reflex cabinet as a device that, thanks to its physical properties, flattens out the resonance peak of the loudspeaker. The final effect is that the frequency response graph looks a bit extended in the lower spectrum, but then the cutoff is twice as fast than with pneumatic suspension cabinets, and also the distortion figure will be higher. Personally, I wouldn't say -straight and simple- that they deliver more bass. It's questionable, to say the least.

Finally, the witches hat. With "contour" I meant "profile", "envelope". The curve that, once rotated, generates the horn. Witches hat horns are basically conical, with a negligible flare at the mouth. To put it very short, P. Viappiani on his monumental work about horn systems, describes conical horns as "not apt to be practically used at low frequencies". Enough said.

Moreover, witches hat horns were also produced too short to be very efficient. (But all horns that are too short are inefficient. Many morning glory horns were also too short, and I've also read about some orthophonic exponential horns criticised as being too short).

It appears to me that we are using different verbage to describe the same thing. That's a merry-go-round situation which likely will not result in an agreeable conclusion. In addition, we have wandered way off the subject which started this discussion. I’d like to get back to the point.

To review, I speculated thus “…that little reflector could be doing just what the name implies. Reflecting back SOME of the sound."

Maybe another question is more relevant: What would that reflector really do?

A quick empirical test could yield some insight into what effect such a reflector could have on the sound. So I created one, and here is what I learned:
1. An angled device as shown in the drawing was fabricated, approximately 4X4 inches on each flap. As my Edison Triumph with large flower shaped horn played a tune, I inserted the reflector, open end in. My ears could not detect a significant change in the sound until the reflector was shoved so far into the throat of the horn so as to block about 80% of the opening. Then the volume began to diminish. I performed the same procedure a second time, this time listening for a change in tone. I thought the tone changed a tiny bit at a certain distance in, but it was so small that I could not quantify it.
2. A conical device with the small end closed and the open end about 4” in diameter was tried next, the angle of the cone being about the same as the angle of the phonograph horn. As it was inserted into the horn, open end toward the throat of the phonograph horn, the sound did not change until the opening of the phonograph horn was nearly blocked. Then the volume diminished significantly. My ears could not detect any significant effect on tone.

From this test, it would appear that the “reflector” as described in the first post does not cause a significant change in the sound. Perhaps more testing with different combinations of phonograph horn sizes and shapes, and different styles and sizes of reflectors, would yield more comprehensive and quantifiable results.

Well... I'm not really persuaded that we were just calling the same things in different ways, but I appreciate the fact that your practical test convinced you that what was predicted by the horn loading theory was correct.

If you would take some time to re-draw on a graph the diameters of the cross-sections of the horn modified with the cone, you would perhaps see better that there is a narrow hole tapping the horn at a point, and later a flare expanding too fast to be of any real use. The hole acts as a high-pass filter, and bass waves will not pass through the hole. I figure that the decrease of bass frequencies is almost undetectable with cylinders (which almost don't have bass) and horns that are less accurate than orthophonic ones.