Long Arch Drawing

[Dennis J]

Here is a drawing of the long arches as I see them.

The components of the front arch, upper and lower, are drawn from the centre and extend to the ends of the plate producing a full arch with small recurves at each end. Rib taper is 32 to 30 mm. Arching thickness is about 3.3 mm along its entire length and I don't see any reason to vary that anywhere across the arching. Upper and lower bout edge thickness is 4.3 mm.

The back arching is more complex having long recurves at each end, especially the upper one. As a result the lower bout part of the arching is slightly higher than the upper.

Arch thickness varies from 4.5 mm at the middle bout to about 3.5 mm near the blocks, widening to 4.5 mm under the back edge of each block. The upper block, being a little thicker than the lower, extends further into the body.

The front arch is 15 mm and the back is 14.5 mm.

I haven't tried to flatten the front arch. The difference between a flat section, near to one-third of the length of the top, and the one I have drawn is only about 1 mm.

 

 

[Andreas Preuss]

Question:

Apparently you define all parameters before building the instrument. But wouldn’t it be reasonsble to have somewhere room for adjustments either in the thickness or more important in the arch height and curvature? I think this is not only necessary to adjust to slightly different material proportions, but can be crucial to match overall thickness and arching for the overtone response.

 

[Don Noon]

If you haven't already, get the Strad3D CT scans of the long arches (it's free now).  Compared to them, your long arch goes too abruptly from convex to recurve... on the scanned instruments, there is a straightening out of the convexity before the recurve.  And the back recurve is only slightly more than the top, but the convex part is less curved and maybe the straightening portion is longer.

[JacksonMaberry]

2 hours ago, Don Noon said:

If you haven't already, get the Strad3D CT scans of the long arches (it's free now).  Compared to them, your long arch goes too abruptly from convex to recurve... on the scanned instruments, there is a straightening out of the convexity before the recurve.  And the back recurve is only slightly more than the top, but the convex part is less curved and maybe the straightening portion is longer.

God bless, look at how sunken that top is! Thanks for sharing, Don. I agree with your comments on Dennis' drawings.

That said, nice work Dennis. I should follow your example and make time for some drawings.

[Davide Sora]

52 minutes ago, JacksonMaberry said:

God bless, look at how sunken that top is!

I see in these deformations a good reason not to start with an already flat sixth (long arch), but it is only my very personal interpretation.

[JacksonMaberry]

Just now, Davide Sora said:

I see in these deformations a good reason not to start with an already flat sixth (long arch), but it is only my very personal interpretation.

Agreed. After using a belly plateau for years, I have moved away from it. I'm encouraged that the above instruments are still working and at a high level, but I'd like to avoid extremes of distortion if possible.

[Dennis J]

11 hours ago, Andreas Preuss said:

Question:

Apparently you define all parameters before building the instrument. But wouldn’t it be reasonsble to have somewhere room for adjustments either in the thickness or more important in the arch height and curvature? I think this is not only necessary to adjust to slightly different material proportions, but can be crucial to match overall thickness and arching for the overtone response.

 

I have posted this drawing as a follow-up to what I have done with cross arches. It has clarified some aspects of the long arches I wasn't sure about and I was particularly interested in the inside profile of the front and back. I'm thinking about another variation of the front arch but it won't be much different. So whatever designs I come up with will be used to make a template and there is no point in making a template and not following it. As I say I'm keen on the idea of an even, pre-determined top thickness for the front.

I'm sure wood density and arching thickness would have a big impact on what an instrument sounds like, but I don't see why that can't be taken into account when designing arching templates.

All I have done here is to explain a simple way to analyse arching geometry.

 

 

 

 

 

 

[Marty Kasprzyk]

3 hours ago, Davide Sora said:

I see in these deformations a good reason not to start with an already flat sixth (long arch), but it is only my very personal interpretation.

Do you want an instrument to sound good today or 300 years from now?

[Dennis J]

I don't think you would need to be an experienced luthier to hazard a guess as to what instruments with sunken tops might sound like.

My guess is that they would be unresponsive and uninspiring.

It's hard to imagine that some of these scans represent their original condition. Wood oxidises, shrinks and deforms over time. If it wasn't for the varnish they probably would not be viable.

 

[JacksonMaberry]

7 minutes ago, Dennis J said:

I don't think you would need to be an experienced luthier to hazard a guess as to what instruments with sunken tops might sound like.

My guess is that they would be unresponsive and uninspiring.

It's hard to imagine that some of these scans represent their original condition. Wood oxidises, shrinks and deforms over time. If it wasn't for the varnish they probably would not be viable.

 

The massively deformed instrument in the image Don posted is actually known to work quite well, but I'm sure none of us believe that's it's original condition.

[Dennis J]

Deformation and possible shrinkage aside those three instruments' archings look very thin. To me it seems to be something which occurs in a lot of instruments, not only the very old.

I don't know if thin tops are the  primary or only reason for it but it doesn't look good.

 

[Davide Sora]

7 hours ago, Marty Kasprzyk said:

Do you want an instrument to sound good today or 300 years from now?

Obviously I want them to sound good immediately, I don't think that the customers would be willing to accept such a wait to get a good sound...

But I like that the violin offers some resistance to the bow so that you can dig in without choking the sound, and it's my belief that a more curved and not flat longitudinal arch helps in this direction, even if this cannot be considered only in isolation, as always.

[David Burgess]

12 hours ago, JacksonMaberry said:

God bless, look at how sunken that top is!

General consensus among restorers seems to be when a top has become distorted that much, re-arching it (again?) will greatly improve the sound and playing properties.

[Andreas Preuss]

10 hours ago, Dennis J said:

I'm keen on the idea of an even, pre-determined top thickness for the front.

So if I understand your idea correctly it looks like this:

The correct calculated arch makes it possible to build a top plate with a uniform thickness. 
 

Do you include then the rib taper from the upper corner blocks to the top block as an ‘arch factor’ ?

And does this also mean that you have very narrow selection criteria for your wood concerning density and speed of sound?

Your approach reminds me of the method of Josef Kantuscher, though to my knowledge he left a few parameters for adjustments. One was that he apparently changed the width of his mould to adjust to to the wood.  He worked thicknesses with a special caliper to the precision of 1/100mm. He used this to draw the thicknesses with an attached pencil in the plates and was very meticulous to have smooth transitions from thicker to thinner areas. (I have watched him doing this when I visited his workshop very long ago) He would make the thicker areas larger or smaller depending on the wood properties. 

[Dennis J]

My main focus lately is structural design concerning arching. I think I've just about covered all I need to know.

I've read a lot of your posts and I'm not involved in the sort of detail that you are.

As far as rib taper is concerned I prefer a taper from the back block to the upper one, no bending.

I see a front arch thickness of 3.3 mm fitting well with an edge height of about 4.3 mm, as simple as that. I just see a uniformly thick arch as something fairly easily achieved and worth trying. I've got a front and back glued up and I'll put into practice my latest ideas.

 

[HoGo]

9 hours ago, Andreas Preuss said:

The correct calculated arch makes it possible to build a top plate with a uniform thickness.

You can always build a plate with uniform thickness but if you mean "structurally sound" top plate I think the traditional arches are not going to satisfy that. "Flat" arches are a clear no go and even less than flat arches of uniform thickness will eventually develop a hint of camel back arch over few hundred years. You only can control how long it will take before it needs arch correction.

I htink 3.3mm target thickness is more than modern makers tend to do. I just repaired my old "usual" fiddle and the top is just about 3.2-3.3 mm thick with slightly thinner bouts (but never below 3 mm) and the top weight was 78g (complete with bass bar and dozen or so small cleats).

[Don Noon]

33 minutes ago, HoGo said:

the top weight was 88g 

Solid body electric?

My last 3 viola tops (40cm) were right around 80g complete.  Yeah, I know I build on the light side, but 88g for a violin top is in student brick zone, IMO.

[Andreas Preuss]

6 minutes ago, HoGo said:

You can always build a plate with uniform thickness but if you mean "structurally sound" top plate I think the traditional arches are not going to satisfy that. "Flat" arches are a clear no go and even less than flat arches of uniform thickness will eventually develop a hint of camel back arch over few hundred years. You only can control how long it will take before it needs arch correction.

I htink 3.3mm target thickness is more than modern makers tend to do. I just repaired my old "usual" fiddle and the top is just about 3.2-3.3 mm thick with slightly thinner bouts (but never below 3 mm) and the top weight was 88g (complete with bass bar and dozen or so small cleats).

I was just trying to find an interpretation of Dennis’ approach, this is not what I am currently thinking or doing.

My questions to Dennis are also motivated to understand something where I thought long ago that it goes into a dead end. It is well possible that back then I didn’t figure out the right approach. 

——————-

For the figures you are giving, hmmm, I’d say heavy and thick. But all this is in the end a personal choice.
 

The motto of my choice is 

‘make the top as light as possible and adjust the rest to it.’

 

[David Burgess]

6 hours ago, Andreas Preuss said:

The motto of my choice is 

‘make the top as light as possible and adjust the rest to it.’

How can one make a top "as light as possible", without pre-knowing the environmental conditions it will be exposed to over the next 10 to 300 years?

At 20% humidity, there won't be dramatic distortion to a normally constructed fiddle from string tension and "wood creep".

100% humidity is a totally different ball game. How much arching distortion do you consider to be acceptable in either 10 or 300 years, and how much of a burden would you like to pass on to your customers and fiddle-trade descendants?

 

[HoGo]

7 hours ago, Don Noon said:

Solid body electric?

My last 3 viola tops (40cm) were right around 80g complete.  Yeah, I know I build on the light side, but 88g for a violin top is in student brick zone, IMO.

Sorry, it was 78g just before gluing it back full with bassbar and varnish, (not 88 g), some 10 grams more than what makers prefer these days. It's not a Strad, but medium grade old violin. The top is just about 3.2-3.3 mm thick just like in OP's drawing. I thought for a while about thinning it a bit but decided I will leave it as is so the cracks will hold better and I didn't want to risk reopening of cracks. Some Gesu's are quite thick as well and play OK.

[Dennis J]

I tend to think top distortion may be caused by the body bending under string tension. So it's not so much a product of downward force through the bridge but the upward movement of the upper and lower bouts.

Whatever the case maintaining a thickness of at least 3.2 mm or more between the sound holes seems wise to me. And after reading some of the comments here I might settle on a more convex top arch.

 

[JacksonMaberry]

5 minutes ago, Dennis J said:

tend to think top distortion may be caused by the body bending under string tension. So it's not so much a product of downward force through the bridge but the upward movement of the upper and lower bouts.

It's demonstrably both, but you're right that the longitudinal compression is underrated in considering this phenomenon.

I, for one, agree about maintaining more thickness in the central areas of the top plate. Depending on wood properties my bellies are as thick as 4.5 mm at their greatest point.

[GoPractice]

May I ask if anyone believes that the direct effect of saddling is increased when the tension is static, or active?

If someone owns a nice instrument that is not played, better to store it at pitch or looser strings? Sorry for the detour.

[donbarzino]

String tension is the primary cause of most deformations so instruments are much better stored

with reduced string tension.

[Marty Kasprzyk]

Maybe we should try to duplicate great violins a little faster.   If the famous violins we like so much have saddle shape creep deformed arches like Don's CT scans show then maybe we should make an undeformed arch to start with and then quickly creep it into the deformed shape if you don't have the patience to wait 300 years.

So I suggest putting your new violin with tensioned strings into your kitchen oven with a pan of water and then leave it until you get a Strad Titian, or DG Plowden shape or your favorite.  Sort of like sticking a fork in it to see if it's done.