Member
Joined 2009
Paid Member
Well I think it's because the sound source is not moving - which I tried to explain in my own way of thinking in earlier posts. What I think is that the reference point for the movement of the cone is stationary. Unfortunately I don't have a big enough woofer to try your experiment with. Besides, I think this issue has run it's course, each of us has our way of thinking about it and we will probably stick to it.
Last edited:
I am going to experiment with 16 per side.
I disassembled half of them breaking off capsules. The first question is, how much power can they tolerate? Both electrically and mechanically.
I already use 64 capsules from cellphones as a tweeter array, now going to try microphone capsules as full range array. The question number 2 is, how full will be that range.
Last edited:
There was a frequent poster on rec.audio.pro about ten years ago who didn't believe this doppler distortion thing. The argument went over weeks, I and dozens of others tried to explain it to him. There were literally hundreds, perhaps even several THOSAND posts made on this very topic. It went on so long I got tired of reading, but then saw him post several months later that he had finally saw and agreed that doppler distortion was generated, but I didn't go back through all the posts to see what convinced him.
But in either case, put the speaker in a fluid with finite speed of sound, and detect the sound with a microphone, and both cases will detect frequency modulation of the high frequency by the low frequency. The high frequency IS reproduced at the cone, is impressed onto the air from the cone, and the low frequency changes the position of the cone over time, thus the signal as detected in the air will "see" the cone closer or farther away depending on the position as set by the lower frequency. It doesn't matter whether the cone is moved by the lower frequency being electrically impressed on the higher frequency, or by the whole speaker (cone, suspension, frame, magnet) being moved by the lower frequency (by some larger "super speaker), it's the cone movement that is the superposition of the two signals AS WELL AS the detection of the generated sound IN AIR that causes the frequency modulation.
The speaker diaphragm's larger movement when reproducing a low frequency signal is a larger percentage of the wavelength (in air) of the higher frequency being reproduced, and so it's a more significant distortion than in a microphone diaphragm.
What is the difference between the movement of the full-range cone reproducing the algebraic sum of both frequencies, and the 10kHz membrane riding on the bass-reproducing cone? I maintain there's no difference. They both move with the sum of the instantaneous amplitudes of the two signals. This is the superposition theorem. At the membrane the output is exactly the same as the electrical input.
But in either case, put the speaker in a fluid with finite speed of sound, and detect the sound with a microphone, and both cases will detect frequency modulation of the high frequency by the low frequency. The high frequency IS reproduced at the cone, is impressed onto the air from the cone, and the low frequency changes the position of the cone over time, thus the signal as detected in the air will "see" the cone closer or farther away depending on the position as set by the lower frequency. It doesn't matter whether the cone is moved by the lower frequency being electrically impressed on the higher frequency, or by the whole speaker (cone, suspension, frame, magnet) being moved by the lower frequency (by some larger "super speaker), it's the cone movement that is the superposition of the two signals AS WELL AS the detection of the generated sound IN AIR that causes the frequency modulation.
All very true so far, but this doesn't describe the interaction of the moving cone with a medium that has a finite speed of sound.
It's true that any real-world speaker is imperfect and will make various modulation products when fed with two frequencies, but that's in addition to doppler distortion.
All other things being equal, the effect (at the same amplitudes) would be lower in water due to sound traveling faster in water than in air. If you use something like a laser to determine the cone position, or (as I recall in some issue of Audio Amateur so long ago) microwaves, you won't see this distortion, as electromagnetic waves travel about a million times faster than sound in air, and so the effect is reduced by a factor of a million.
I think I agree, and my earlier 'thought experiment' (a method which people around here seem reluctant to use, instead only ever thinking about 'real' speakers and audio waveforms) would seem to prove it.
However, is there not a difference between a tweeter glued to a solid woofer cone, and a tweeter glued to a similar woofer cone made of wire mesh? Would you expect to measure exactly the same Doppler-type shift on the tweeter signal in both cases? Does the low frequency modulation of the air in which the 10 kHz signal propagates have no effect? If not, I agree with you. But if there is an effect, it would be very neat (in the British sense) if it turned out to compensate for the Doppler shift.
Thought experiments are very useful (I use them a lot myself to think through how something works) but have to be used carefully. They're the mental equivalent of navigation by dead reckoning.
Like dead reckoning you have to start from a known, provable point, or the result will be wrong. Likewise you need to understand the subject matter fairly thoroughly and not make any mistakes along the way in the deductive process, one mistake or logical error and you will end up at completely the wrong conclusion.
So it's important to experimentally verify things part way through the thought process, if possible, to see if you haven't run off the rails.
In this particular debate a key point of contention part way through the thought experiment from electrical input to measured result at the microphone is where is the acoustic centre of the driver, and is the acoustic centre modulated with low frequency cone excursion.
Without having the correct answer to this, a correct conclusion about the existence of doppler modulation cannot be reached. There are those in the thread that still believe the acoustic centre of a driver does not move with cone movement - however the onus of proof is on them, since it flies in the face of decades of measurements and theoretical calculations which show it does in fact move and that doppler effect on speakers is real. (Albeit small)
It's not up to "the establishment" to prove once again that something that has been well understood and known for decades does in fact exist. However those that doubt the existence of the phenomena and even doubt that acoustic centre doesn't move with the cone have taken no steps to experimentally prove or disprove this. (Where are the measurements ?) Thought experiments alone are not sufficient in this case, if understanding of the preceding steps in the deduction process is faulty.
The debate in this thread really does feel like "no, really, the earth is a sphere not flat"
We have sent people out into space to see the earth as it really is, yet a significant percentage of the population of the world still believe it is flat. Such is life.Not in doppler shift no.
Maybe not exactly the same shift to every last decimal place, but very close.
The key point is that the distance from the driver to the listener is varying with cone movement, and the average speed of sound propagation across the full air path length is not going vary in any significant way due to the small localised pressure variations that the woofer produces.
The high particle velocity just in front of the cone is very localised - it very quickly dissipates from near field to far field - the air molecules immediately in front of a large woofer might be moving back and forth +/- 5mm but by the time it spreads out and gets to the listener it will be an infinitesimal movement. Most of the path length will be through air which is hardly moving at all.
Let me pose a thought experiment to you. A large truck with a blunt nose is driving along and just behind the front grill is a horn. The front of this truck will be "carrying" a pocket of air along with it much like a woofer cone.
Will this horn still have doppler shift despite this ? Of course. Would a horn attached on a stick at the top of the truck also have doppler shift ? Yes of course. Would the doppler shift be exactly the same ? Maybe not but it will certainly not be absent in either case.
Last edited:
Now we're talking!
What if we were standing in an open tunnel, and the truck made a good seal against it like a piston in a cylinder. i.e as it advanced towards us it pushed all the air ahead of it in the tunnel past us. Would we still have Doppler on the sound of the horn it was carrying?
What if, instead, the tunnel was sealed behind us, and the advancing truck was compressing the air in our listening chamber. Would we measure the same Doppler shift on the sound of the horn?
OK, well I'm going to suggest that there is no Doppler in the case of the medium itself being displaced at the same speed as the acoustic source towards the listener. (as in the truck advancing down a tunnel and pushing a sausage of air past the stationary listener).
From there, I suggest that if the medium is also being moved by the displacement, it is not simply an indisputable truth that displacement of emitter relative to listener = Doppler. In real speakers with real audio signals, my intuition is that there is not much in the way of Doppler shift and that in principle, the 'perfect' (which probably means infinitely sized) full range driver doesn't suffer from Doppler.
Thereby reversing my earlier assertions!
From there, I suggest that if the medium is also being moved by the displacement, it is not simply an indisputable truth that displacement of emitter relative to listener = Doppler. In real speakers with real audio signals, my intuition is that there is not much in the way of Doppler shift and that in principle, the 'perfect' (which probably means infinitely sized) full range driver doesn't suffer from Doppler.
Thereby reversing my earlier assertions!
Gosh, my own personal thread.
I'm now thinking that as long as all frequencies propagate out in the same way, it doesn't have to be an infinitely big cone for there to be no Doppler. Probably similar to what people were saying a few pages back, but in terms that were too complex for my feeble brain. The crux of the matter is comparing the effect of a tweeter glued to a solid woofer compared to a tweeter glued to a wire mesh woofer. In the former case there is large displacements of air literally 'flapping your trousers' at the same moment the sound of the tweeter reaches you. In the second case, the tweeter has moved in the same way, but there's no air movement corresponding to that displacement. The former case is, of course, the full range driver, and the second is what I naively imagined was going on.
Quite nice to have got there in the end (for now!). It never sounded right that there was this strange effect that got worse the bigger your speaker. I accept that real world speakers may still suffer from it a bit, but it's not fundamental to the whole concept.
But I could be wrong, of course.
I'm now thinking that as long as all frequencies propagate out in the same way, it doesn't have to be an infinitely big cone for there to be no Doppler. Probably similar to what people were saying a few pages back, but in terms that were too complex for my feeble brain. The crux of the matter is comparing the effect of a tweeter glued to a solid woofer compared to a tweeter glued to a wire mesh woofer. In the former case there is large displacements of air literally 'flapping your trousers' at the same moment the sound of the tweeter reaches you. In the second case, the tweeter has moved in the same way, but there's no air movement corresponding to that displacement. The former case is, of course, the full range driver, and the second is what I naively imagined was going on.
Quite nice to have got there in the end (for now!). It never sounded right that there was this strange effect that got worse the bigger your speaker. I accept that real world speakers may still suffer from it a bit, but it's not fundamental to the whole concept.
But I could be wrong, of course.
Last edited:
Thought experiment is foe, but let's experiment with arrays of tiny speakers.
I've found yesterday this lot on ePay, but I can't afford it alone. Anybody to share?
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=300654649644
I've found yesterday this lot on ePay, but I can't afford it alone. Anybody to share?
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=300654649644
Thought experiment is foe, but let's experiment with arrays of tiny speakers.
I've found yesterday this lot on ePay, but I can't afford it alone. Anybody to share?
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=300654649644
Nice.
Is the idea to drive them all in parallel, or to do a Peter Walker delay line thing?
Dont confuse air vibration (sound) for air movement (wind) there not the same.
No, it gets worse with excursion. Since you need less excursion with a larger driver you will have less doppler.
Heres my though experiment: An 8" full range speaker with a 10ft excursion and very low freq. is playing .5hz (2 sec for 1 cycle ) at full excursion and 1khz at the same time. I have difficulty seeing any way around the doppler distortion (the 1khz sound comes from the cone (its acoustic centre) not the magnet or anything else).
This would imply that the low freq sound (air pressure) effects the highs freq sound and adds distortion to a tweeter in a 2way system (actually implies that any noise (like a truck idling outside) will cause distortion in your audio!
Last edited:
Member
Joined 2009
Paid Member
cdbd, you have inspired me with another gedanken experiment:
Take the 8" full range speaker with the 10ft excursion and a very low freq. playing at your 0.5Hz. Now place a 2nd driver behind that cone and drive it with the 1kHz at the same time. The main speaker magnet and the tweeter magnets are fixed relative to each other.
We are going to assume that the physical set up is such that the sound from the 2nd driver can not reach the listener other than by traveling through the cone of the speaker in front of it.
The 1kHz signal from the 2nd driver propogates through the air to the large cone and excites a 1kHz movement in that cone. The larger cone then excites the air in front of it generating a 1kHz signal to the listener.
The listener can not see the 2nd driver, he/she hears only the 1kHz tone and it appears to radiate directly from the main driver. Since the 2nd driver is stationary there is no Doppler.
How is this different from a single full range driver excited with both the low frequency and high frequency signals. Perhaps it isn't, and no Doppler shift exists in that case either.
Take the 8" full range speaker with the 10ft excursion and a very low freq. playing at your 0.5Hz. Now place a 2nd driver behind that cone and drive it with the 1kHz at the same time. The main speaker magnet and the tweeter magnets are fixed relative to each other.
We are going to assume that the physical set up is such that the sound from the 2nd driver can not reach the listener other than by traveling through the cone of the speaker in front of it.
The 1kHz signal from the 2nd driver propogates through the air to the large cone and excites a 1kHz movement in that cone. The larger cone then excites the air in front of it generating a 1kHz signal to the listener.
The listener can not see the 2nd driver, he/she hears only the 1kHz tone and it appears to radiate directly from the main driver. Since the 2nd driver is stationary there is no Doppler.
How is this different from a single full range driver excited with both the low frequency and high frequency signals. Perhaps it isn't, and no Doppler shift exists in that case either.
No, without any delay, just a narrow row from floor to ceiling. They have to keep up with higher power than microphone capsules, but I suspect they have harder and heavier cones.
Can you really use 5k pcs? On Taobao looks like a single quantity price is just 2.5rmb (roughly $0.40):
Small horn speaker 20MM
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.