CNC mill to make a good violin

[liu01china]

The tech data for Guarneri del Gesu violin is available as shown in the picture.
Now I want to use a CNC mill or a CNC router to make a good violin.
3-D data will be generated for CNC data input.

Does anybody know how much should I pay for a CNC for this purpose?
I only need to make the front and back boards; others can be taken from a 'cheap' violin.

Really need to look for two pieces of quality wood for the violin.

[turmite]

Hello liu,

Cost is exactly propotional to the quality of the machine......or for the most part it is. There are many small machine out there that would do this job, but I think you need to ask yourself how accurate do you want to do the job? I have a machine that is huge and vibrates if I run it too fast, but it would do the job. On the other hand, a vertical machining center designed for metal work would be capable of making that piece so smooth you would barely have to sand it.

Now.......what does the second pic represent? Is that energy from a recording or a computer simulation?

Mike

[rlrhett]

I'm not sure what those pictures represent, but I can tell you that you can do a microscopic copy of a Stradivarius and not get the same tone. It is not just the raw dimensions of the instrument that give it its tone. If it were, people would have long ago made perfect duplicates of the best violins. After all, CNC wasn't discovered just yesterday.

Each piece of wood has its own characteristics. The masters were able to "tap" a piece of wood and know where it needed more carving and where it needed less. They worked the individual piece listening to whatever magic they listened to, slowly modifying the individual plate until it sounded right to them. No two Stradivarius are alike in dimensions, although they have similar tone.

Oh, and just so you know, the wood also changes with time. Resins in the wood crystallize, moisture leaves the wood, and some people swear that the linked cell structure of the wood begins to modify in response to the constant vibrations it receives. There are some who believe that the famous 17th century violins that are now impossible to match were not that impressive when new.

If you are planning on mass producing a violin, you don't need a CNC machine. You can build your self a duplicarver and simply copy a generic top shape by buying a good quality violin. One man with a duplicarver can probably copy a top in about 30 min, inside and out. Maybe if he was doing it all day, even 15 min. It would take a small non-industrial CNC probably a couple of hours.

On the other hand, if you are intending to hand make violins and want a little help with the initial carving, you can use a CNC machine to rough carve your plate. You can then focus your energy entirely on tuning each plate by hand once the machine is done.

Just don't expect that because you copied a famous violin you will have the same instrument. "Cheap" violins are themselves copies of the violins you are looking at. What makes them cheap is that they are mass produced and no master's hand ever touches the thing. They are perfect copies, but no one has taken the time to tune the plates to the individual wood and instrument they are intended for.

One can spend a life time learning how to modify each plate to make a glorious instrument. It is a very rewarding practice, like anything that takes mastery, and there are a lot of groups out there willing to help. Google the Catgut Acoustical Society or the Musical Instrument Makers Forum. There are just no shortcuts.

[liu01china]

Hello liu,

Now.......what does the second pic represent? Is that energy from a recording or a computer simulation?

Mike

Thanks Mike!

BTW, the second pic is the CT scan of the back plate while the fisrt one is the top plate; The color represents the actual thicknesses of the plates.
The 'Cannon' is the famous violin that Paganini had played.

[liu01china]

I'm not sure what those pictures represent, but I can tell you that you can do a microscopic copy of a Stradivarius and not get the same tone. It is not just the raw dimensions of the instrument that give it its tone. If it were, people would have long ago made perfect duplicates of the best violins. After all, CNC wasn't discovered just yesterday.

These are certainly useful advice from an expert. Thanks a lot, rlrhett!

While it's not possible to duplicate a million-dollar violin, is it still possible to make a medium grade violin in a much faster way using a CNC combining the existing data?

Now it's possible to construct a 3-D CAD model out of the scanned data (actually I'm doing one with iges). And I'll also do a modal analysis using FE. Combining math material models, it's possible to decide an optimised thickness for a violin. The data in above pictures can be referenced and used as scale factors.

If we can work out the optimised thickness for each particular wood, then a mass production is possible. Let's talk about 200 violins a year selling $2000~$3000 each. The violin head, neck and finger board, can be ordered from outside places like from China at very low prices. (They are selling a whole piece including bow and case plus strings for $40). This looks like a pretty good small family business model, isn't it?

It would make more sense to make a violin with carbon fibre in which the material property would not change. The carbon fibre violin current on the market is $4900!

[rlrhett]

You actually don't need to do a scan of great violins to copy a master violin. The old and new masters use mathematical formulas to create the basic shape and then hand tuned the tops to create the specific tone of each instrument. The most common formula was to trace a curate cycloid from the head to tail and again across both bouts. Lots of books on the subject, and even a guy who has a plug-in for Rhino CAD specifically for creating optimal violin plates.

As to whether there is a faster way to do it using CNC, the answer is both "yes" and "no." "Yes" in that you can create reliable blanks from which only a little hand tuning is necessary. "No" in that, again, a duplicarver can do the same for less money and much more speed in a production environment. Of course, to use a duplicarver you need a high quality master. I think some factories in Asia have even gone so far as getting an aluminum CNC'd master that they copy.

Put bluntly, a CNC carved violin that has not been hand tuned will not be a $3,000 violin, it will be a $600 violin. You can see from the scans that the thickness varies over the plate and between violins. These aren't random variations, the master luthier carved the violin where it needed it and left it thicker where it did not in response to each specific piece of wood (even though they are the same species).

I don't want to discourage you, but the market for hand made instruments is traditionally difficult to get into. You can't swing a dead cat in San Francisco and not hit an aspiring luthier. The problem is that most people are perfectly happy to play a mass produced violin, guitar, cello, etc., as long as they can go to their biggest music mart and pay very little money. I have seen acceptable quality instruments for less than it costs me to buy the wood. Economies of scale, I suppose, and greatly subsidized production in some Asian countries. Oh, one last discouraging thought from Mr. Sunshine... if you saw a $3000 violin at a music store you liked, they probably paid $1,500 or less to the luthier for it.

The market for people who can tell the difference between a hand-crafted violin and a mass-produced one is relatively small. To do it right is a real art, and developing an ear for a properly tuned instrument is a life long endeavor. Those two elements combined tend to limit the potential market and make it hard for a new comer to break in. Like any luxury item, up market violin buyers tend to look for famous name luthiers to justify the price.

Whew... all that bummer stuff!!! Well, I just want to give you an accurate picture. If you already have access to a CNC, I would say "Go for it!" It sure would be fun!!

If you are curious about what results you would get, I would suggest buying a violin kit from Stewart MacDonald. For $150, its a steal, and the plates have been carved to "optimal" dimensions just waiting for your hand to tune.

I don't know if they use CNC or a duplicarver, but either way the kit would be pretty much exactly what you would get out of your own machine and Chinese parts. If you are really thinking about this as a family business, I would buy 10 or 20 of these kits and practice making the best instrument you can with them. If you really are building instruments people want to buy for $3,000, then you are ready to bring it all in house.

Good luck, and have fun with it!

[dbassva]

the one thing ive learned over the years is there is always a market for anything as long as you position yourself correctly in that market...musicians are always looking for more gear, especially if its different from the massed produced options....

if you are planning to mass produce anything, you have to go with cnc period, unless your place in the market is hand made low production models, which is fine as long as thats your target....what you need to do is decide what you want to do, put it on paper and stick to it...after that you can decide what machines you will need...i can tell you from experience, that the least amount of hands required to make your product the better, people make mistakes that are costly....

now, im assuming you know how to build a proper violin and you want to go with cnc since thats what you asked about....and im assuming you want to do high production runs, cuz if you dont you may be better off having someone else machines these for you....there are a few companies that specialize in whats called ghost building for other intrument makers...the cnc can be used to machine just about everything you will need including jigs, fixtures and tools....and if you are going to make money with this project and want to buy a machine, you better get a good quality one...all of those rinky dink do it your self cnc routers that are cheap just wont cut it...they are great for a hobbyist, but they just arent fast enough, accurate enough or durable enough for making intruments....you should get a vmc like a fadal or a haas..you dont need a big one for violins so it wont be that expensive, and i would go with a new one or factory remanufactured or demo one so you can get some training on it..im not sure on the price of the little ones, probably in the $20-30k range.....haas has a interactive website that you can price out your machine...you wont need many extra options except maybe 10k spindle and way covers...fadal is a better deal right now and dont require any options to their base model