Back to top tuning?

[John Bertotti]

I see prints of top and back thicknesses and see people shooting for these target numbers, generally, they just work but my new question concerns the relationship of the top and back.

Do you more experienced builders take the time to try and fine-tune the top to the back in some way other than approaching plate stiffness and mode reactions and your gut instinct? In some way tweak how they couple to improve the tone just that little bit more? In my mind, it seems like something someone with experience would also work on, maybe that little bit extra that gives some instruments that little bit of magic beyond the normal fine instrument? If you do just what are you looking for or how do you approach it? String it up in the white and play a bit then tweak the archings and bracing?

[sunburst]

About the best I can do is use the numbers I've accumulated.
As an example, I had an A-style mandolin that sounded really good to me. Strong fundamentals, good chop, loud yet with superior tone and response. When a customer asked about getting a mandolin and described such a sound (I'm building it now) I went to my notes, selected wood from the same sources, carved to similar arches, thicknesses and weights, and I expect the mandolin to sound very much like the other. Basically, I had and example that worked for the sound I wanted so I did my best to repeat it.
I don't string mandolins up in the white and play them. I never have, so I'm not likely to start doing that now. At that point in the process, the arching cannot be adjusted beyond minuscule amounts. Only thicknesses can be slightly adjusted, and in order to do much of that safely the plates must be very slightly overbuilt. I basically use my notes and experience to guide me and try not to leave much excess material in the plates.

Others, obviously, do it differently. We all tend to develop our own methods with experience and we tend to stick with what works for us.

It's different with new builders who don't have records of their previous work and don't have experience to drew upon. Fortunately for them, there is more information on arching, graduation, bracing and so forth at their fingertips than might have existed anywhere back when some of us started making mandolins. If one simply does what has been proven to work for others, things will turn out fine and the luthier can start to develop his/her own data base and experience from that enviable starting point that we old-timers wish we could have had.

BTW, I forgot to mention that there are a few builders who have been able to track free plate modes throughout the build process and are able to gain some insights from that; notably Peter Coombe. He is able to do that because he has many years of detailed records for contrast and comparison. Free plate tuning is basically of no use to me (and most other builders) because I do not have the detailed records that I would need to use the info. Without something to compare to, the information is useless.
I gave up trying to gain any insight from free plate tuning early in my building experience because I was simply not finding any consistent relationship between plate "notes" and the sound of the completed instruments.

[John Bertotti]

Thanks! And O am constantly amazed at the information I can find buried on the net! Truly a lot of knowledge is out here!

[peter.coombe]

If you would like to know how I do it

http://petercoombe.com/publications/jaamim7.html
http://petercoombe.com/publications/jaamim8.htm

This was all done under the assumption that violins are similar to mandolins (now proven wrong), and is entirely empirical. Why this works makes perfect sense to me now after Dave Cohen's research on mandolins and the Gore/Gilet books were published. Match the free plate ring modes and you consistently end up with a difference of 4 semitones between the main top mode and the main back mode (in arch top oval hole mandolins). Flat top mandolins and F hole mandolins are different, but the principles are the same. The more mandolins I make just adds to the data presented in those 2 articles, so I have an overwhelming amount of data now and it all supports what I said then.

[Rob Roy]

Since the seasoned veterans are weighing in here, what about the idea that the two halves of the top should be tuned to slightly different pitches? I think I read that in Seminoff's book a long time ago. On the few I've built, they seem to come out about 30-40 cents different on their own. Is it worth worrying about?

[John Bertotti]

As much as I like to read I am not giving Stew-Mac 350$ for the books, that just doesn't work into my budget currently. I'll watch for used online somewhere. Enjoyed those article Peter thanks for the links!

Was I reading the chart with the mode frequencies and differences between top and backs correctly in that it seems the instruments that were preferred had at least two modes with frequencies almost identical?

[peter.coombe]

Was I reading the chart with the mode frequencies and differences between top and backs correctly in that it seems the instruments that were preferred had at least two modes with frequencies almost identical?

Correct, mode 4 (the ring mode) are tuned to be as close as possible. Don't do that on a F hole mandolin or a flat top, the back will be tuned too low. In the finished instrument that gives a difference of 4 semitones between the top and back main modes, which are the 2 modes that vibrate like a speaker. 4 semitones is what Gore and Gilet recommend for guitars, and mandolins vibrate like guitars so it works.

I think I read that in Seminoff's book a long time ago.

Once again, take the tap tuning chapter of the Siminoff book rip it out and burn it. What he says bears no relation to the physics. The reliable info is from the scientific research of Cohen and Rossing, Graham Caldersmith, Trevor Gore and Gerard Gilet, and others from the guitar world.

[John Bertotti]

Another mode-related question. When those of you looking at modes want to quantify them do you just play a freq through a speaker and watch the pattern? How is this possible with a carved top instrument? I would think the shape itself would preclude using glitter or any type of light medium. I am also curious how it changes once assembled. I am also curious how deflection under string tension affects modes. Most builders build so I figure there probably isn't a lot of data on this. Thanks, everyone. I'm not sure why but to me this is fascinating.

[peter.coombe]

There are a number of ways to measure the modes of vibration. You can use Chladni patterns (glitter), you can tap and record and do a FFT, or you can use laster interferometry. Laser interferometry gives the most information, but it requires expensive equipment and that makes it impractical for most people. Chladni patterns are still possible with carved top instruments. Adding the strings and bridge does not affect the shape of the patterns, but will change the frequencies because the bridge adds mass to the top. Other fittings such as the tuners, tailpiece and pickguard will also change the frequencies, although not as much as the bridge. If the pickguard is attached to the fingerboard then I have found no differences, but if it has a bracket to the sides, then the main top mode will drop in frequency, as you would expect if you have read the Gore/Gilet books. If the frequencies change then that must affect sound in some way.

[John Bertotti]

I think I found the book. The Science of Stringed Instruments, is that the right one?

[John Bertotti]

The Science of Stringed Instruments is the correct book and it is ordered, should be a good read. In the meantime, I can read up on modes, what we should look for in carved mandolin tops and backs and how to do FFT's on my mac in Logic.

Thanks, everyone for the replies and messages, it is helping a lot!