Hi, Simon. I don't know any instrument than can produce a tune without harmonics.
It's not it's two separate sources and the IM is the tone of interest. The individual 13K and 14K sources have no mechanism for generation of a 1K tone only the mic does. Both systems as we are using them are weakly non-linear and obey closely simple polynomial relationships.
Well, what about the IM of the harmonics of those two sources ?
This said, I don't care too much about distortion of microphones. I use them in a totally scandalous way: I put one in front of an musical instrument and decide if I like the way it sound, or not ;-)
But, please, be kind to not repeat-it to anybody in this forum or I will be obliged to kill you (If i have some time left before to be killed ;-)
Last edited:
Ed you need to go through the math and physics of the reciprosity process before making claims. What do you actually get with two coupled perfectly reciprocal motor generator sytems when the first harmonics are in opposite phase?
My immediate answer would be no harmonic in either case. The same for when only one has a harmonic you would see it in both cases A to B vs B to A that is. I might be wrong, but if I cared I would work it out.
I said before I have not seen the reciprocity used that way. With microphones you can deduce acoustic sensitivity without an absolute measure of pressure. You can not determine R without an absolute volt or amp.
Last edited:
Each of the two sources produces its own set of harmonics but as long as sound pressure levels aren´t high enough to trigger severe nonlinearity of the air the only device processing both signals with a potentially nonlinear transfer function is the microphone/preamplifier combination.
So, the first order effect will be the impact of the microphone/preamplifier combination which will result in IMDs.
None they each are reproducing one tone. I think this can be set up to eliminate any confounders.
Of course. My remark was just to underline the vanity of these measurements
for those who look at numbers as God's word.
When I see the complexity of the whirling phenomena in the airflows, I wonder how this media transmits itself these awful pure sinusoïdes of dream. No IM produced in the transport ? Sure ? ;-)
Of course, I'm aware of the interest of diferencial measurements in this heroic struggle to reduce our own contributions to harmony (harmonics).
My goal is also to bring a pinch of good mood: this new year has started badly, over there and on twitter.
Last edited:
Thanks Ed
Each unit (X) will have some 1kHz THD as a transmitter (Xt) and some other 1kHz THD as a receiver (Xr).
When we measure at the receiving end, the 1kHz THD total =1kHz THD of transmitter+1kHz THD of receiver
Therefore your example will come out as follows
At+Br=0.01
Bt+Ar=0.011
At+Cr=0.02
Ct+Ar=0.021
Bt+Cr=0.021
Ct+Br=0.022
This is a system of 6 equations with 6 unknowns.
It’s solvable but the numbers you give unfortunately don’t match.
George
Handbook for Sound Engineers contains an excellent monography on microphones. I would look it up for you but unfortunately can't at the moment.
The fact that electrical wave form from moving coil mike runs 90 degrees behind the acoustical wave should be obvious from working principles. At zero crossings of the acoustical wave, velocity of the diaphragm is highest and thus output.
Condensor mike has highest output at crest of sound wave so is in phase.
No. Just the numbers.
As I have no experience with such a testing, do you think that excluding the influence of phase difference renders the approach fatally unrealistic (even at 1kHz)?
George
Yes I think. (Not so sure that I can).
But i never worked in a mic manufacturer's laboratory ;-)
I only use (used) acoustic measurements for response curves of speakers in anechoic chambers, taking them with a huge pinch of salt*, or for max SPL in concerts with sound level meters (In French: Sonomètres).
*The goal was to figure out the good phase alignements and best relative levels of multi-ways speakers for my and my colleagues own use, and to can bring to the marketing department the most beautiful curve you can provide moving the mike millimeter by millimeter ;-)
One of the puzzles is to can send to the marketing guys a flat response curve, while you speaker has to be slowly and regularly descending as the frequency increases for a nice listening experience.
By example, i hardly can tell the frequency breakdown of a membrane, looking at its acoustic response curve. While it is obvious looking at the impedance curve. After a while, you don't even look at nothing, because they all break more or less around the same frequencies, depending of(on ?) the diameter.
(Smileys everywhere).
The harmonic amplitude and phase as a generator probably will not match that as a receiver. The reciprocity process is good for absolute sensitivity but may not be useful for harmonic testing. Maybe with a tuned receiver you could learn something. I suspect it will be limited by the highest distortion device in the chain. Especially since the phase relationships become critical when the levels are within a few dB of each other. For the fundamental this is not an issue but it is for the harmonics since their relationship to the fundamental is not defined or even fixed as frequency changes.
Its worth a try. I'm pulling my GR reciprocity test set from storage today and if its reasonably quick to set up I may be able to try the concept. All this stuff takes a lot more time to do than to conceptualize. Even the dual voice coil testing chews through several hours of fussing before collection of a few seconds of info. I'll need to dust off the wave analyzer to have any hope on this project, if its working. . .
The claims are that microphone distortion goes down linearly with level so it should be predictable at lower levels until its lost in the noise. Is there a reason this may not be true? And is there a way the mechanism that causes IM would not cause HD?
The B&K and HP preamps I have all have a dominant 2nd harmonic (as John Curl predicted). its low enough (at least -85 dBC at the equivalent of 124 dB SPL on a B&K 4145.
The plot is a B&K 4145 on a B&K 2639 through a B&K 2609. The pistonphone has a rather amazing 1% 2nd harmonic given the mechanicals and the necessary shape of the cam (four lobes) to make this possible. You can also see the .5 dB difference between the amplitude at the fundamental and the composite amplitude from the addition of the harmonics.
Its worth a try. I'm pulling my GR reciprocity test set from storage today and if its reasonably quick to set up I may be able to try the concept. All this stuff takes a lot more time to do than to conceptualize. Even the dual voice coil testing chews through several hours of fussing before collection of a few seconds of info. I'll need to dust off the wave analyzer to have any hope on this project, if its working. . .
The claims are that microphone distortion goes down linearly with level so it should be predictable at lower levels until its lost in the noise. Is there a reason this may not be true? And is there a way the mechanism that causes IM would not cause HD?
The B&K and HP preamps I have all have a dominant 2nd harmonic (as John Curl predicted). its low enough (at least -85 dBC at the equivalent of 124 dB SPL on a B&K 4145.
The plot is a B&K 4145 on a B&K 2639 through a B&K 2609. The pistonphone has a rather amazing 1% 2nd harmonic given the mechanicals and the necessary shape of the cam (four lobes) to make this possible. You can also see the .5 dB difference between the amplitude at the fundamental and the composite amplitude from the addition of the harmonics.
Attachments
Working often with statistics means to be never sure, so no ....
But i tried to address that point in my post when mentioning the nonlinearity of air (aka transport medium),further it´s possible to examine its impact by subsequent experiments like varying the microphone distance while holding the excitation levels of the sources constant.
Second that too, but despite the "vanity" that might exist, it´s always of interest to know about the uncertainties our measurement equipment is associated with and additionally it is quite a good example for a way to circumenvent a typical "hen and egg" problem in metrology.
Scott,
Using "identical" transducers limits the phase shift also. There may be some but both amplitude and phase are data you can recover in each measurement. This has been done. How do you think they rate microphones for distortion?
Demian,
Reciprocity does assume the process of source or receiver is symmetric. An easy test is to look at the first harmonic phase as you change level.
Last edited:
And a ribbon mic is velocity sensitive so its phase is also different. So what. Unless your using these 2 different types of mics as a stereo ( multi channel) pair it doesnt matter. The different distances to the mikes mess up the phases anyway.
View attachment 726562
Pavel, please take a chill pill, seriously...also please consider that this post is not to lambast you.
I won't bother to 'splain this again' (copyright Jocko
) about Foobar and Foobar ABX being problematic except to reinforce that others have come to the same finding and in this case I simply couldn't be bothered to attempt 'formal' Foobar ABX.You proposed "Anyone able to tell the difference, in ABX of course, between the electrical signal and recording through condenser microphone from W18NX001 woofer? (the last 200Hz sine test)"
Facts are I used VLC as per the screen shot above.....no VU meters.
I merely mouse click flicked back and forth until I was certain that one track sounded 'louder' than the other.
IOW I did hear a slight difference between the recordings and assuming that the recordings were precisely fundamentals level matched asked if the difference in perceived level (SPL) was due to distortion components added by the (dirtier ?) speaker/air spacing/microphone/cable/preamp/cable/ADC signal chain.
A signal comprising clean tone with additional distortion products by definition contains more energy/information than a single pure tone and ought to sound louder as I understand it....please correct me if I am wrong on this.
This understanding/reasoning led me to ask if the louder recording was the result of your mic chain, see ?.
That said I did notice that something wasn't quite right with what I was hearing+reasoning.....the B recording did 'sort of' sound cleaner but went against my reasoning and caused me to do a few dozen VLC track swaps to confirm certainty that B sounded louder and then posed my bracketed questioning, double questioning actually.
It has been very forcefully promoted here in the past (perhaps including you) that valid ABX testing requires perfect level matching....by this measure your 200Hz_test ABX is by definition invalid (according to Reaper metering).
To be fair (as I said in my previous post) the difference I heard is fine, very fine and maybe testimony to the quality of your recording chain.
Noteworthy is that I did manage to pick a difference but not the difference you thought (and I thought) you were testing.
As to the audibility or inaudibility of your test chain components (speaker and/or mic), that is not proven or disproven at this time.
Dan.
Some further thoughts...with matched fundamentals levels the only difference to be heard is the actual difference (measurable THD+N, subjectively tonal/timbre change), with the mismatched fundamentals levels of this case the perceived difference is the result of two differences and the actual difference is masked/overcome by the fundamentals level differences.
This leads to further questioning of what to measure when matching.....fundamentals only, or rms values, or peak values....or what ?.
How did you attempt level match ?.
Last edited:
Haven´t seen anything contradictionary so far (my experience is limited to only a few of these condenser mics from B&K though), provided the capacitance loading case is avoided (which was mentioned already by Scott Wurcer and explained by Erlingsen in his various publications) as it will lead to departures from the predicted "linear" level/unlinearity relationship as this depends (not solely of course) on the ratio between the internal capacitance and external ones.
Adapters between microphone cartrigde and preamplifier should always be examined for this effect (or better avoided).
- Status
- Not open for further replies.