Kinky Stardust (Acoustic Guitar Pickup)
This page is about the design of my very own acoustic guitar pickup design.
Lately I have been recording songs with my guitar and vocals. I used (above average quality) dynamic microphones and the results with that setup weren't very satisfying. There are several reason for that: There is very significant cross-talk between the two microphones when recording simultaneously, it is very susceptible to background noise and the handling is cumbersome (both mics need to be well aligned to both my guitar and my mouth). Usually those problems are mitigated (except for the background noise) by recording the instrumental first and then the vocals. But since i also would like to be able to do amplified live performances there is only one solution (i can see): using an acoustic guitar pickup. Well, and since they are rather expensive, I'm a hacker and lately have started to build my own audio equipment (e.g. a dual channel mic preamp), i decided to hack my own acoustic guitar pickup.
The Name: "Kinky Stardust"
Well, the name of the pickup is due to the resemblance of the mounted pickup to an eye patch. And especially on my white guitar it reminded me of "Ziggy Stardust" the fictional character by David Bowie which had his face painted white and wore an eye patch. The prefix "Kinky" is kind of an reiteration since Ziggy Stardust was most an embodiment of kinkiness but I still felt I wanted to emphasize it, because the elastic strap used for mounting the pickup is a re-purposed sex toy made from re-purposed bike parts (i think it was made by the people behind http://bikesexual.blogsport.eu though i couldn't find that specific toy). I think it is worth noting that all the parts used are re-purposed parts i found in the whateverlab except for the audio jacks and the piezo.
My specifications for a good pickup:
- Good sound quality: Very low noise, a rather analogous and easy to correct frequency response, no artifacts (like the infamous piezo-quack), etc.
- Easy to mount, without any modification to the guitar (even easily removable glue or tapes)
- Doesn't inhibit, influence or hinder any guitar playing techniques (like e.g. soundhole based pickups prevent tapping the strings)
- Doesn't (significantly) influence the acoustic characteristics of the guitar (e.g. doesn't try to pickup sound from the strings directly or is installed in the soundhole)
- Very low susceptibility to background noise and feedback
- Cheap and easy to build
Research on existing pickups
So, i did some research on usual pickup technologies and guess what i found? None of them really met my specifications. Also existing DIY solutions tended to fail my specifications utterly and were mostly very crude (gluing a piezo directly to guitar - seriously?). Well, after ruling out picking the sound up from the strings directly i saw two options: Picking up vibrations of the corpus directly (e.g contact mic) or picking up "airborne" sonics (e.g. classic microphone). Both have their pros and cons: e.g a contact mic has a very low susceptibility to background noise whereas a classic microphone is more true to the sound of the guitar. That ment i needed a third option - a mix of both techniques. I investigated different designs for acoustic emissions sensors (AES) and decided to modify one of the approaches i found:
The Approach
After some experiments i build a pickup to the above AES design (which despite the similarity works differently) with three major modifications: 1. There is no couplant and the wear plate doesn't directly touch the object. Instead i left an air cushion of about a 1.5 centimeters between them. Note that the connection between object and case is airtight. 2. I also introduced a gap (about 3mm) between the backside of the piezo element and the dampening material and drilled a small hole (~ 1mm) in the dampening material, so that the backside pocket is not airtight. 3. I used hard rubber as case material.
Theroes™
The theories i have (IANASE™ = I am not a sound engineer) behind those modifications are: An AES directly picks up the sound from a solid object (the guitar corpus) and therefor isn't susceptible to background noise. On the other hand i experimented with a classic AES setup and the sound was really bad (far from preserving the acoustic characteristics of the guitar) which in my opinion is due to the fact that sound behaves differently in different media (materials) and that the wear plate and the object are not perfectly connected (e.g. fused or glued). So i decided that while loosely adhering to the design i still wanted to measure changes in air pressure like a classic mic - hence the (air tight) air cushion. I Introduced the (not air tight) backside gap so vibration of the membrane wouldn't be inhibited - covering the hole actually results in less gain and a dampening of the higher frequencies. Well at that point i already was pretty satisfied with the results except for two issues: 1. Even the most careful handling (touching) of the guitar was picked up very loudly. 2. The FFT (of a Rigol oscilloscope) showed significant beat noise. After some theorizing I figured the beat noise was the result of the different speeds of sound conducted in the (solid) cover and the air cushion. So i replaced the metal cover with a hard rubber cover which actually fixed both issues. The beat noise was reduced to a very minimum and the handling noise became very acceptable. I'm aware that thought I started with an AES the final pickup works very differently :)
The Result
- Has very decent sound quality, with a slight dampening on the lower frequencies and a bit stronger on the higher frequencies - resulting in an easy to correct curve.
- Is rather easy to mount.
- Doesn't hinder guitar playing (at least as far as I'm aware of)
- Hasn't been tested with different guitars but it seems to do a good job on preserving the characteristics of the guitar.
- Has a very, very low susceptibility to background noise and feedback
- is cheap and easy to build
- Profit!
Video, Pics and Sample
Usage video (Please note that i forged the audio recording it a second time directly to my usb sound interface because i didn't have a decent microphone with me - gonna re-record tomorrow)