Idiot proof approach to guitar building? For this idiot

[Matthew Lau]

Dear MIMF,

Looking back at the years of postings I've done, I have to seriously cringe.
Over the past few decades, I've grown to appreciate my limitations.
I've created a system for my dentistry to be "idiot proof."

For building acoustic guitars, do you have best practices for an idiot proof system?
I can see stuff like a fox bender, deflection testing, and chladny diagrams being some....but am not sure how extreme to go?

-Matt, who overcomplicates things.

ps. I'm hoping to get back into guitar building because I met a lady I'd like to marry.
I'd like to build her a special guitar as a wedding/anniversary present...but don't trust myself.

[Matthew Lau]

Oh, and I don't know if this is the wrong section of posting.

Should this be in the tools/jigs section instead?

[Alan Carruth]

No matter how much information you keep on your instruments, you will probably find you'd like more at some point.

I start with measurements of the properties of the wood; density, Young's modulus along and across the grain, and the associated damping factors. I use this to choose appropriate wood (lighter tops for Classicals, and so on) and to determine the top thickness. Once I've chosen a top and trimmed it to shape I look at some Chladni patterns to be sure there is not some problem that will make it hard to 'tune' the way I want.

Once the top and back are thicknessed and braced, and trimmed to the final shape, I use Chladni tuning methods to try to get the top and bracing to work together as well as possible. This involves looking at the patters and trimming the braces to get 'good' shapes and frequencies that are in the right range. I use a form that has rows of outlines on it to draw in the initial patterns and note the frequencies. I keep track of how the frequencies change as I trim the braces, and make out a new set of forms when I have the final dimensions.

After the guitar is assembled and strung up I do an 'impulse' test. I hang the guitar up and hit the saddle on one side with a small hard plastic ball, and record the 'thunk' on my computer. I use an FFT program to break it down and find the resonant peaks, and print out the spectrum chart up to 1000 Hz. I record the low end 'signature' mode shapes, found with Chladni patterns, on the top and back, and also the 'main air' and lowest 'corpus' (bar mode) frequency on the same form.

All of that, and anything else I can think of, goes into a file folder for that guitar. The file cabinet is one of the most important tools in the shop. I can't always remember everything, but those files don't forget. It's easy when I have a question to pull out relevant files and look things up.

It's important that you do any such measurements in the same way every time. Even fairly minor changes in the room, or in microphone position, when you do a sound measurement, can alter the spectrum. When measuring the 'air' resonance be sure you don't have anything too close to the hole (no closer than one hole diameter, and preferably more).

It is certainly true that you can go down a rabbit hole with this stuff, and end up spending all of your time measuring and not building. OTOH, it can give you really sinking feeling when you've just made your best guitar ever, and you can't remember something like the weight of the bridge or exactly how thick the top is.

Hands and ears can be very fine measurement tools, and I tap and flex as much as anybody. But they can be unreliable, and it's been shown that most people are not nearly as good at that sort of thing as they think they are. A friend of mine says he wants to be able to have a fight with his wife and still go into the shop and do his best work. Objective measurements are what they are, and don't change with your emotional state.

Speaking of fights with the wife; I gotta go...

[Alan Carruth]

I wanted to add that I've been at this for a long time. I set up my file system before the computer era, and have become pretty set in my ways. For those of you who are more technologically progressive, I could suggest a nice Android app called 'Luthier Lab' that was put together by a friend of mine, Mike Mahar, and his wife. He originally started out on it some years back as a Java app, but his wife, who is an Android developer, suggested the switch. I gave him some input, especially regarding the acoustic stuff, like the signal generator and spectrum analyzer, but there's 'way more in there than I'll ever use.

It's based around a 'project' orientation: you can save all results for a given project in one folder. He has a drawing module based on arcs and lines that is set up with a number of more or less standard outlines, which you can modify, or you can come up with your own. You can also design brace patterns, and the drawing program will generate arch contours based on 'curtate cycloid' cross arches, which include the recurve. You can use the signal generator to produce Chladni patterns by plugging your tablet or whatever into a suitable amp, and use the camera to record the patterns, label them with frequencies, and store them in your project folder. Later you can use the spectrum analyzer to record taps, and label the frequencies of the peaks and dips to compare with Chladni patterns from the assembled instrument. There's probably more in there that I haven't even looked at yet, and Mike keeps upgrading it. Best of all, it's free. Look it up.

[Dick Hutchings]

Allen, your method will not work for me. I've never been a note taker and lean towards repeating mistakes and getting it done... 2 or 3 times. At 67, it's too late to change. In the end I think I'm building a fine first guitar and I like the tone of my $20 Marketplace Stagg so the bar isn't very high. I hope to build a Martin sounding guitar but I'll be happy with what ever I end up with and on to the next one. If I live long enough to build 10, I may start thinking your way, but I doubt it. I think my method is custom made for idiots like me.

[Peter Wilcox]

I've never been a note taker and lean towards repeating mistakes and getting it done... 2 or 3 times. ...I think my method is custom made for idiots like me.

I build the same way. At 77, after building 30 or 40 guitars (I don't even count them) in the last 12 years, I've never built one without at least one mistake, and usually more and sometimes the same one. But that's what makes it interesting for me - each guitar is a new experience.

But I've gotten pretty good at fixing and covering up mistakes.

[Alan Carruth]

I don't think I'll ever make one without any mistakes. As far as I'm concerned the idea is to avoid making the same ones too many times, and to make smaller and less consequential ones over time.

[Freeman Keller]

I have thought about Matthew's original question a lot as I go thru my lutherie learning experience. There are the obvious things that we all know (but maybe don't always do) - choose a known plan, use good materials. If we deviate from the plan know why. Keep notes. Strive for good worksmanship. Those things just make sense.

The one that baffles me is the whole concept of "voicing". I've sat thru GAL seminars by people I respect, bought the expensive books and read every article in American Lutherie. When it comes to actually doing it I still don't have a clue.

The one thing that has helped me tremendously was getting a copy of Alain Lambert's incredible spectrum analyzing software. I had fooled around trying to do this with Audacity and it was just too much of a hassle. Alain's software makes it easy.

I have used this in two completely different fashions. For my last guitar I just started tapping the top and recording the waveforms each time I did something to it - raw hunk of wood, planed to thickness, braced, braces shaved, box closed.... I was in no way using this to guide what I was doing, simply documenting the changes. However each spectrum lets me think about what I am hearing. Is the bonk tone getting "more musical" (I keep hearing people say that, what does it mean?). When I close the box I immediately see a new frequency peak - must be the air resonance. Is it what I expect? Can I hear it?

The second way I have used the software is when I make a change in a guitar. Someone wants a new nut, take a couple of sound samples with the old one and the new one. Are they different? Do I hear the difference? Does what I hear make sense? Change the pickup on an electric guitar - same thing. Ditto strings - are they "brighter"? Do I see it in the spectrum? Do I like it?

I've developed some techniques for doing both that are easy to do and more or less standardize my method, and its become almost a regular part of most guitar projects that cross my bench. I'm still an idiot and my guitars aren't any better but for once in my lutherie journey I have a bit of a map.

[Alan Carruth]

I never 'got' Dana Bourgeois' tap tone method on the finished box until he actually walked through it at a GALL convention, and allowed me to track what he'd done. We planned on making an article out of it, but the top of his guitar cracked on the trip home, so we did not get a chance of actually evaluating the sound.

Tap tuning of the 'free' top and back and Chladni tuning get much the same information at different levels of detail. Tapping and flexing is quicker, and works pretty well once you learn how. Chladni tuning gives you more information, which may or may not be helpful.

Spectrum analysis of the completed guitar is certainly useful, and it's even more so when combined with Chladni pattern information. It's possible to get similar tap spectra from guitars with somewhat different patterns, and they do tend to sound different. Chladni patterns are harder to generate at higher frequencies, and guitars with virtually identical modes and spectra in the range below 1000 Hz can sound noticeably different due to differences at higher frequencies, particularly in the 204 kHz range. Matching the low range seems to produce instruments with a similar 'character', but the higher range seems to be where 'quality' lives.

More information is generally better, but you can't spend all of your time making measurements.

[Matthew Lau]

Thanks, Al and Peter!

Al is actually one of the biggest reasons why I got into MIMF, with his explanations of the "why" of voicing and tuning a top. Before MIMF, there was a lot of misinformation and folksy truisms (like the state of Kung Fu).

I think that I'm working on first bending sides that are even, and not wonky looking....figuring a way to have a neck be consistently aligned straight with minimal fuss, and with a certain level of uniformity...making sure my joints don't implode/explode.

If I build more, density...deflection...chladni diagrams, etc...a feel for flexing the top. etc.

My late mentor (Randy Angella) was horrible at explaining stuff....but maybe it was because he was very Italian.
He'd explain that a good G-string was the key to a great classical, and that the profile of the sound should be sensual and well balanced like the figure of a beautiful lady (hard for this nerd to relate). He'd also rhapse lyrically into how the universe, the sine wave, and God and man are united in guitars. As for building guitars, I didn't learn too much aside from selecting wood.

[Alan Carruth]

Yipes! that should have been "2-4", not "204" kHz range.

[Chuck Morrison]

Wow, that's an interesting description of the desirable g string tone. Not helpful, but interesting.

An idiot proof system for building an acoustic guitar ? Hmm... This is the approach I have developed over the past 40+ years. I didn't start out with it, but it developed over time. Unfortunately, it doesn't keep idiots out, at least not at first.

The first step is to ask what is it you are trying to accomplish ? If you don't know the goal, how can you evaluate whether you succeeded ? I set a question for each instrument I make, It may not get answered, but it should help me narrow down the question for later. Each instrument should add to your knowledge base.

Next, build a few instruments and identify your problem areas, so you can concentrate on solving those problems. This can be workmanship, building techiques/jigs etc, finish work, whatever.

Take as many measurements as you can and record them in the same notebook. You will probably change what you measure as time goes on. That's OK, you learn what is relevant as you go. It's unlikely you will remember months later what you did for that guitar. Thickness/stress test results, Bonk test spectral analysis, basic Chladni tests are a good place to start.

Make sure you know how to set up the instrument for optimal playing ease. Neck profiles, action setups, etc. are important to player satisfaction and first impressions.

You need to develop a concept of the sound qualities you want in your instrument(s). Listen to other guitars you are impressed with (as you play them), and identify what it is about them you like. As you build instruments play the new instruments long enough to get a sense of what they really sound like (a minimum of a month, an hour a day if you can). If possible play them alongside the "target" instrument to see how they measure up. If you luck out and build a really outstanding instrument, KEEP IT! Use it as the one to match or beat. It will be a valuable reference. You know what you did and you know the result. As you refine testing techniques, you can use it as the base reference.

Don't be afraid to try new, bizarre approaches. It's all about learning "what happens if I do this ?" Much of what I think I really know came from building unsaleable instruments.

Sorry to not be specific on tools and methods, but those are variable and replaceable. Materials are not all the same and our approach to using them must be based on knowledge, the most valuable of which comes from experience.

[Alan Carruth]

When initially bending sides I go for over-bent. Getting a smooth curve is more important than getting the right curve. Once the whole thing is over bent you can go back the next day and un-bend it as necessary. If you make a corner, bend the flatter areas on either side to match that, and then come back tomorrow and unbend the whole thing. If you try to unbend the corner and then bend the flat spots on either side the corner just bends again, since the wood stays plastic for a while.