AKG 451 and Ck-8 short shotgun - Question?

[Paul R Johnson]

Sorry Greg - I posted a long answer this morning, but looks like I forgot to press submit.
In fairness, the mic works fine but I don't like not having the solid reason.

I'm quite happy with interference tubes, or at least I was. My error was not realising the critical factor - that the transducers must operate in pressure gradient mode, so that the pressure across the width of the diaphragm is key. It is not the slots, or vents in the tube that matter, it is the increasing angle as sound arrives from anywhere other than on-axis. The same physics as how ordinary cardioid and super cardiod microphones function, but the longer tube increases the cancellation effect. The barrier to the physics is that some sources use the slots as the creators of interference patterns, while that physics falls down with their absence. Clearly the actual physics is either too complex for public distribution, or nobody has really bothered to explain it very well. DPA do a fairly good job. in a dumbed down style. Interestingly there are a few scholarly articles and the abstracts detail that analytical models were based on transmission line theory and were not actually verified!

I've found all kinds of conflicting mathematical data. It appears that the data is clear on one feature only. Cancellation takes place inside the tube. Angle and frequency are key elements as to the performance. Length is linked to the directional cut-off point where the performance resumes conventional cardioid performance. Most sources simnply seem to suggest that on axis produces no cancellation. Off-axis sound entry into the tube generates it by virtue of the phase shift across the element's diameter which doesn't happen when the wave front arrives at 90 degrees to the diaphragm. I'm in the dark as to how the AKG side slot appears to work irrespective of direction to the sound source.

I remember 40 years ago being similarly surprised how a vertical cylinder with a slot cut in the side could be a functioning omnidirectional antenna. The direction of the slot is similarly unimportant. Two broadly similar functions of a cylinder? Or just coincidence. The Alford slot does have solid evidence as to how it works, but nobody seems to have really studied microphones in this way. B&K, for example have lots of publicly available microphone papers, but they are quite thin on interference tube designs.

[Greg Miller]

Paul,

My understanding of the interference tube is not related to the angle of the diaphragm. It's related to the angle at which the sound hits the tube. I don't want to try drawing the diagrams myself, but I will look online and see if I can't find an explanation that mirrors my understanding.

Our local amateur club demonstrated a slot antenna. In our case, we used a large flat piece of sheet metal, then cut a slot that was roughly 1 meter long (1/2 wavelength for the 144-148MHz band). The slot was relatively narrow (I don't remember the exact dimension). Anyway, the signal polarization was opposite of what most people expected. For a horizontal slot, the signal polarization was vertical, and vice versa. I don't quite understand that, but that's the theory and was born out by out demo.

[Greg Miller]

UPDATE

Paul, you might want to look at this explanation by Shure, it's fairly good.
https://service.shure.com/Service/s/...language=en_US

I still think the illustrations are less than ideal. I know what they are trying to show, but because of the scale of the drawing they are not showing it as clearly as they could. I really hope I don't end up having to make my own drawing, that will be a PITA.

Meanwhile, let's consider a sinusoidal sound source at 1125 Hz, since the wavelength is approximately one foot. This source is 90º off the axis of the tube, in other words the sound will be coming straight toward the *side* of the tube.

Now think of a tube with just two holes; one is 12" from the transducer, the other is 18" from the transducer, so they are exactly 1/2 wavelength (180º) apart at 1125 Hz.

At a given time, the positive pressure peak from our sound source reaches the tube, and some of the energy enters through each of the holes. Once inside the tube, some of the energy starts down the tube toward the transducer.

Exactly one cycle (1/1125 second = 0.00089 seconds) after the positive peak entered the tube, the peak from the 12" hole will reach the transducer, creating a positive voltage peak. But the positive peak from the 18" hole has not yet reached the transducer.

Now consider 1/2 cycle later than the above time, in other words 1.5 cycles (0.00133 seconds) after the positive peak entered the tube. Now the positive peak from the 18" hole will reach the transducer, because that hole is 1.5 cycles length away from the transducer.

But at this same instant, the pressure waveform from the 12" hole has continued moving toward the transducer. Therefore it is no longer at the positive peak. It is now half a cycle past that, in other words at a negative peak.

At this instant, at the transducer, there is a positive peak from the 18" hole, and a negative peak from the 12" hole. These two pressures cancel, so the sum is zero. That is the "interference" in the name of the microphone: one pressure wave interferes with the other. Since the pressure is zero, the transducer diaphragm will return to its rest position, and the voltage output will be zero. For this particular tube, with these two holes, and this test frequency and direction, the sound is completely rejected.

(Actually the cancellation takes place almost everywhere inside the tube. But in terms of mic output, the only thing that matters is the sum at the transducer.)

If you continue adding holes at the proper places, other frequencies will cancel. But when the distance of the holes equals the length of the tube, that's the longest wavelength (i.e. lowest frequency) where the interference principle will work. Below that frequency, the directionality is caused by the "shadowing" effect of the long tube, and by the directional characteristics of the transducer.

If you add holes until you reach an infinite number, then the spacing between the holes will be zero, and you will have created a longitudinal slot such as you find on your CK-8.

--

Hopefully that explanation some of the other info that you found. If you still don't understand the theory, and you don't want to use the CK-8 (because you don't understand it) then let me know. I'll give you my address, and you can send the CK-8 to me. I'll be more than happy to use it with one of my three 451 preamp/bodies. (For that matter, I'd like to get a pair of omni cartridges ... I have only cardioids right now. I think the omnis would be good for pipe organ.)

[Paul R Johnson]

I think the problem with me is that I've read that explanation and the other simplifications many times and they're the kinds of explanation I had drummed into me in my teacher training - they're not exactly correct, have flaws and are actually wrong, but they suffice as they do explain the process. These are the explanations I've used for years, but I'm discovering they're actually wrong. The cancellation process is variable, depending on the angle of incidence away from on axis.The actual wavelengths are what have been used in the transmission line theory, but the way the wavefront behaves in the tube is the key feature. The wave front is not a single path, but an infinite number of paths, and it appear to be this that needs to be considered. The mathematics from using the usual straight line path is flawed. The tube has width, and this is the key feature. I scribbled a diagram.

In A, the sound being captured is simplified into 3 points, far, mid and near, when of course in reality there are an infinite number, creating a wave front. The 3 lines are all time aligned and impact on the diaphragm at the same time. In B, a sound is arriving at an angle, and from the point where the near path hits the diaphragm, a line is draw of where the time alignment is - 90 degrees to that near path. The far path has further to go to reach the diaphragm, and partially cancels. The width of the tube and the width of the active surface of the microphone diaphragm work to allow the cancellation. The wavelength at higher frequencies is related to the extra physical distance and there is more cancellation. A low frequency has far loess cancellation as the extra path length is tiny as a ratio of the full wavelength. That ratio works for higher frequencies and allows more cancellation.

This is what my research so far has produced, with a little head work from myself. It's a better analogy of what I believe is happening. Most manufacturers have diagrams showing sound bouncing down the tube, but I believe it's the incident angle at the diaphragm that's critical. Some web sites that purport to know everything about microphones are sadly lacking, and good ones seem to be very weak on interference tube science. Wiki is absolutely dreadful on it. Most manufacturers are offering the grossly simplified explanations. No real issue with them because they struggle with making sure people understand. Some of the net information about M/S stereo for example really messes up the theory of that system too.

I think I've now got enough of a handle on it to set my mind at ease. This solution and explanation for me allows normal slots and a full length slot to both function. The only unknown is how speaking into the slot and speaking into the solid surface both produce the same volume works.

I remember asking the university about using simplification with students. I asked if it was OK to teach them some flawed science to enable greater understanding, and they said yes - you fix it when their understanding shows the information they have to be flawed. Apparently it comes up often in Doctoral research when people suddenly wed established facts are actually wrong, but they work in general. A bit like when at school we were told Amps x Volts equals Watts, and they hope nobody asks if that works in AC and DC circuits. Even worse when you find that most people use it in AC circuits which is where its the worst. Maybe the interference tube concept is the same - the usual explanation is good enough?

Sorry this went on a bit. Taken a lot of time today, reading and deciding how much reliance to put on what's out there.

I'm happy with this angle explanation Greg, but I sense you're now not? We can leave it or scratch heads further.

Seriously though - do you think the physics of the Alford Slot antenna which uses tubes might have a link to the fact the slot is NOT directional, level wise? Are sound waves able to travel around a cylinder in this strange way? Perhaps that's a bit too far?

[Greg Miller]

There was a bit of discussion about this several years ago. I don't think it will address your questions.
https://www.dvinfo.net/forum/all-thi...erference.html

As for your theory (and sketches) commenting on the angle of incidence at the diaphragm ...
Yes, we know that large diaphragm condensers sound a bit different off axis, so there might be some basis for your concern. However, most of the shotguns I can think of do not have holes or slots that close to the diaphragm (compared to your sketches). Probably 3 or more inches away. With a mic ID of about 1/2" the direct (un-reflected) angle from the nearest slot would be only about 10 degrees off axis. Of course there would be an infinite number of reflections inside the tube; everything would somehow average out.

[Greg Miller]

Also re: the Alford slot antenna. I'm not sure to what extent that theory applies to sound waves. Radio waves (and light waves) have polarization, but sound waves do not.

The earlier discussion that I linked to had an interesting suggestion. Any radical design might be patented. Check the patent numbers, look up the patents (they are public record). There, you might find any "secret" design information that exists.

I still wish I had an anechoic chamber to do some experiments. I guess one option would be a big plot of land, away from civilization's noise, and with some towers to elevate the source and receiver somewhat above ground level. Land isn't especially inexpensive, though, so I think I'd rather have the real chamber.

[Paul R Johnson]

Twice now I've posted and for some strange reason it didn't work.

The missing post related to the drawing - the slots were drawn like that to make the angles greater and show the extra distance across the diaphragm - which was of course the usual small size. The angles in practice are much less than coming in that sharp, but clearly, only the on-axis paths no cancellation, anything other than 90 degrees to the diaphragm produces cancellation. It works for me I think as the explanation for what's going on.

Still unsure on how sound entering through an open slot, or via the journey around the cylinder produces the same level? Clearly though, from reading as much as I can find, the more learned folk are ahoy to admit the research is still in development after many, many years. Interference tubes clearly work - as everyone uses them, but the actual mechanism that enables the useful properties is not understood in much more than broad ways. They think X or Y is responsible but nobody appears to have spent time studying it. Slots around the circumference or a front to back slot both enable pressure waves to enter the cylinder where cancellation takes place. That seems to be it?

The mention of baffles in some documents seems to muddy the water, because the action of these baffles isn't mentioned. The mention of multiple capsules is actually seen in a number of cheap video microphones produced in china that have a 'zoom' switch than links in the second cardioid capsule fitted around 75mm or so in front of the other. The two polar patterns overlap and cancel sideways arriving audio. I assume the front element prevents the rear element receiving the direct sound and the two combine to give a shotgun-like response.