Playing sound through a piezo pickup

  • Thread starter Thread starter McMajik
  • Start date Start date
McMajik

McMajik

New member
Right, I have absolutely no idea where to post this but this seems like the place I'm most likely to get useful results.

I want to make an impulse response of my acoustic guitar's body. I have K&K pure mini's in it. No electronics, just piezo pickups connected straight to an output jack. I've been messing around with just knocking the body, with my knuckles, with a pick, hitting the muted strings, but it seems to me a good way to stimulate the body would be playing a pop or test tones out through the pickups. Would that work? How risky is it? And how should I go about it?
 
Hey man,
I don't know much about it but I guess you'd need something other than the piezo. I'm thinking maybe recording DI guitar then pipe it into the the guitar body using a speaker against, or inside, the body.

If you mic that up then you have the 'dry' source audio and the recorded audio, from which to generate your impulse.

As I say, just guessing since you've no other replies yet, but I can't see a piezo making enough noise to be useful.
 
IDK if it'll be loud enough either, but the worst that can happen is you pop the piezo and have to replace it. A quick first test would be to just plug a line output into the guitar and run something through. Next would be to plug it into a headphone jack somewhere. Then maybe try some small solid state amplifier.

I'm not sure it's ever going to pass much bass in this configuration, though. These things literally have a high pass filter built into them by their very nature and the low impedance of any amp you plug into it is going to make it cut off pretty high. You might be able to precondition the signal to compensate some...

I'm kind of interested in the results, so definitely let us know how it goes.
 
Not sure how efficient piezo pickups are going "the other way" I suspect they might be optimized for sound to lekky conversion?

Not going to be high efficiency anyway since if you take a guitar as producing say, 80dBSPL at a foot you only get a few tens of millivolts from a piezo pup? You could get a few tens of volts drive from a power amplifier but not a guitar amp unless you can get at the PI input, you don't want the preamp voicing messing with the response. Put the correct load (resistance AND power rating!) on the amp (even if it is transistorized) and feed the pup from a 100nF capacitor .

I am assuming you regard the pickup as "sacrificial"?

Dave.
 
What exactly are you attempting to achieve?

An impulse response for an acoustic guitar is going to be dependant on many things but most importantly the frequency and place at which you excite it. There is no single response. Thats what makes it such a fascinating and bewildering acoustical model.

If you give an idea of what you want to investigate I can point you in the direction of where to start reading.. I can not see what you stand to gain from the classic understanding of an impulse response which attempts to include all frequencies. Whilst it can be enlightening in the case of loudspeakers to auditorium acoustic modelling its pretty worthless on an acoustic instrument which is essentially a Helmholtz resonator made of precise anisotropic materials..

Aside from that piezo would not be the way to go as they are a one way system which translates energy from physical movement to an electrical signal via compression of the piezo element.



What are you looking to do?
 
Tweeters are often piezo elements. It's pretty common to rip piezo buzzers apart and use them as contact mics. It's a two way street very much the same way that a paper cone speaker can be used as a microphone (see "sub-kick") and a dynamic microphone can act as a speaker (learned that one the hard way). Whether it'll make enough noise, or cover a wide enough spectrum is a couple other questions. That one about "why" is also a good question, but I can see a couple of decent reasons.
 
ashcat_lt said:
Tweeters are often piezo elements. It's pretty common to rip piezo buzzers apart and use them as contact mics. It's a two way street very much the same way that a paper cone speaker can be used as a microphone (see "sub-kick") and a dynamic microphone can act as a speaker (learned that one the hard way). Whether it'll make enough noise, or cover a wide enough spectrum is a couple other questions. That one about "why" is also a good question, but I can see a couple of decent reasons.
Click to expand...

A piezon pickup is not a piezo buzzer the latter is more robust and normally attached to a membrane slighty more forgiving than a guitar top. You can try it is you like but you will more than likely ruin the element and with very little reward.
 
muttley600 said:
An impulse response for an acoustic guitar is going to be dependant on many things but most importantly the frequency and place at which you excite it. There is no single response. Thats what makes it such a fascinating and bewildering acoustical model.
Click to expand...

That's what I was thinking. The pickups are on the inside on the guitar, under the saddle, my thinking was exciting the guitar where the strings meet the body would give a...an impulse response more true to how the guitar responds to being played. I've been messing around with knocking the guitar body in various places and strumming all the strings muted, with mixed results (Clicky). This was with a different guitar that has no pickup in it, to be applied to a different guitar again. I'm mixing some live stuff with DI'd guitars for someone, thats what started this whole thing, but it got me thinking about my own live acoustic sound. If I miced my guitar up properly in the studio and got a good impulse response of it (With any EQ e.t.c. recorded tracks need also applied to the impulse response, the one in that link needed lots of high boost/lower mid cut), then loaded that into a pedal that uses impulse responses, I could have the best live guitar sound.

I don't want to lose the pickup though. It's a good pickup. The output doesn't have to be loud (Or does the body respond to different volumes differently?), I'd be recording in a vocal booth using the lowest self noise stuff if I have, but if I can't drive it safely I'm not going to risk it, I don't know enough about piezo elements beyond that they turn electric fluctuations into compression/expansion and vice versa.
 
OK before we move on to a discussion on what your objectives are a little background may help.

An acoustic guitar sounds the way it does for a number of reason. First and foremost are the materials it is made of and specifically the top and how it is braced. On most decent but not all acoustic guitars the top is made of some variety of spruce. Spruce is chosen because it has a high stiffness to mass ratio. That allows it to be relatively light but remain stable as well as vibrate freely compared to other woods.

The top needs to respond to the vibrations of the string and it does it by passing some of the energy through the top and some travels back down the string. This is what causes our sustain or more correctly decay. The energy is limited and it is lost through the top via the bridge. It is lost in other places but the bridge is our prime concern. The speed of sound in spruce is around ten times greater along the grain than radially or tangentially so the braces not only strengthen the top but pass the strings energy around the top. The sound hole is not there to push sound out it's there to aid the way the top responds to vibrations. In fact less sound waves leave the soundhole than are pushed out by the top vibrating.

This moves us along to the second most important aspect that shapes the timbre of an acoustic instrument, the enclosure or body cavity. Like all enclosures there is a lot going on inside the guitar sound wise. Air is constantly being compressed and expand as the top moves. The size and location of the sound hole aids (tune) what are known as the AO or Helmholtz resonances. They play a large part in the efficiency of the whole system helping the top to vibrate freely and across as many frequencies as possible without getting dead or wolf tones. There is a lot more to it than that but those are the two principle factors that shape the resonance of the acoustic guitar.

As you can imagine the resulting frequency response for each note or group of notes will be highly complex and very different. In short you would need to measure all the fundamental frequencies of each note and it's component harmonic series frequencies to be able to accurately model an acoustic guitar. How the top vibrates and the enclosure responds physically to say a high e will be way different say a low A'. How they respond in tandem should be a sum of those parts but that is not the case as there is a deal of cancellation that occurs as well as the influence of impedance in the materials that take effect.

The guitar is a beautifully simple acoustical model but also a highly complex one. One of the reasons it is so hard to get a really accurate sound live is because of this. There are some sound (no pun intended) reasons for this. Whatever method you use to capture the sound is only giving you a small part of the envelope. A pickup only samples the sound envelope at that point. Even micing up live is a compromise. In the studio you would likely have over the shoulder mics and room placement would also be a major factor as would room mics and the preferred type. That is rarely possible on stage. The second factor is that when we listen to an acoustic what we hear is not what is going on inside the enclosure or right close to the bridge or soundboard. What we hear is either several feet away and above if we are playing it or several yards away at least if we are listening to it.

If the what you suggest were easily possible it would have been done to death. One of the holy grails of acoustic guitar players and pickup manufacturers is a true and accurate amplified sound even at quite a large cost.

It remains true today that even though live sound for acoustic guitars has come on a mile in recent years what we get is still a facsimile of what is actually going on and just an acceptable usable tone. Most players I have built for and worked with have spent years developing their live sound and it is often way off their unplugged or studio sound.

If you are interested in looking into more detail at the top behaviour and Helmholtz resonances I can point at some accessible entry point stuff. I used to teach this stuff to student luthiers and student music engineers.
 
Start with a line level signal. Can't possibly hurt the pickup. If that's not quite enough, maybe try a headphone signal.

muttley's post leads to the idea that a sine sweep might be more appropriate than an step function here. Pick attack will be more like a step/impulse, but the sustain/decay of the note will be more like a sine wave. In most cases, the two methods are equivalent, but this might be a special case.
 
OK, I have given you an over view of what you are attempting and have taken the time to outline what is involved. It seems that others appear to be offering advice without any real idea of what is involved.

Here is the short version....

It wont work... if you want to understand why refer to my previous post. The key reasons and the rationale are there. If you really want to do a thorough investigation into what is going on with an acoustic instrument when subjected to various frequency stimuli good old FFT analysis is the way to go. Yes it has been done. Yes I have done it. No there is no golden egg...;)

A final word of caution. There are lots here that have a depth of knowledge and understanding about sound engineering and processing but zero understanding of what acoustic modelling and analysis involves. Be wary, be very wary before you commit large amounts of time that could have been better spent understanding the principles and differences between various disciplines and situations.

Good luck..
 
You must log in or register to reply here.
Back
Top