Thinking of making a new mould.

[Wood Butcher]

I fancy making a Guarneri model violin, but I was hoping the experienced makers, may be able to help suggest which model might be best to use, and also from which period.

I have read that it is believed Guarneri used a single mould, to make all of his violins. The differences being only the edge overhang, corner shapes, and how well bent, or not, the ribs were. I’m not quite buying this, the variation seems far too great.
Therefore, rather than making a generic Guarneri mould, I will want to use a single instrument, to create the patterns.

I know some are much more famous than others, but for those who have access to the real instruments, which ones have you found to sound the best?

Thank you in advance.

[Dr. Mark]

This could provoke an interesting and inconclusive discussion - are you the real Wood Butcher?

 

[Don Noon]

1 hour ago, Dr. Mark said:

This could provoke an interesting and inconclusive discussion - 

IMO the result depends so much on the details that you won't find much agreement about the model.  I'm using "model" here mostly regarding outline, as arching has not been mentioned and can be tweaked to make a large difference in result.

I have made one Kreisler(ish) model that performed very good, and a  generic Guarneri(ish) model that seems to work better.  Now I have another family of generic Guarneri(ish) models, but the jury is not going to reach a decision until VSA next week.  But again, it won't be the model (outline), but the other things.

[Davide Sora]

1 hour ago, Dr. Mark said:

This could provoke an interesting and inconclusive discussion - are you the real Wood Butcher?

 

[Michael Darnton]

The mold may have been the same but the results weren't. Unless you're thinking of removing the ribs and pushing them around to get the size you want each time, I'd pick one that works [tonally] better than the rest and looks like something you'd like to make and follow that. My nominee is the "Cannone", but there are other good ones, of course. Personally I'd stay away from 1734-37, the smaller ones, and the very early ones. If I wanted what the early Strad "copy" ones (1730-33 or so) have, I'd just cut to the chase make a Strad instead. I guess what's left is something 1738 or after, mostly, but not one of the weird ones if you want a generic del Gesu. This is all general, and there are outliers within the range I stated I would avoid and ones I wouldn't make in the date range I think is OK.

In addition to the "Cannone", the ex-Kochanski has a nice neutral look and is available as a STRAD poster, for instance.

[Wood Butcher]

19 hours ago, Dr. Mark said:

This could provoke an interesting and inconclusive discussion - are you the real Wood Butcher?

 

I think many discussions end up this way, but there is always something to learn, especially from those who have had experience with the real thing.

[Wood Butcher]

18 hours ago, Don Noon said:

IMO the result depends so much on the details that you won't find much agreement about the model.  I'm using "model" here mostly regarding outline, as arching has not been mentioned and can be tweaked to make a large difference in result.

I have made one Kreisler(ish) model that performed very good, and a  generic Guarneri(ish) model that seems to work better.  Now I have another family of generic Guarneri(ish) models, but the jury is not going to reach a decision until VSA next week.  But again, it won't be the model (outline), but the other things.

Thank you, Don.
I should have been more specific, but in this context, the outline, scroll F-holes. If my resource material has arching information, I would then try to incorporate this, allowing for distortions.

[Wood Butcher]

17 hours ago, Michael Darnton said:

The mold may have been the same but the results weren't. Unless you're thinking of removing the ribs and pushing them around to get the size you want each time, I'd pick one that works [tonally] better than the rest and looks like something you'd like to make and follow that. My nominee is the "Cannone", but there are other good ones, of course. Personally I'd stay away from 1734-37, the smaller ones, and the very early ones. If I wanted what the early Strad "copy" ones (1730-33 or so) have, I'd just cut to the chase make a Strad instead. I guess what's left is something 1738 or after, mostly, but not one of the weird ones if you want a generic del Gesu. This is all general, and there are outliers within the range I stated I would avoid and ones I wouldn't make in the date range I think is OK.

In addition to the "Cannone", the ex-Kochanski has a nice neutral look and is available as a STRAD poster, for instance.

Thank you, Michael.

There is a lot of food for thought here, and your experience with Guarneri violins is greatly appreciated.
Generally, I have used either the back outline, or purfling channel to generate my mould pattern. Looking at pictures of some Guarneris, the back and belly outlines look quite different to each other.

Part of this is clearly down to wear, but in others, it must be due to the ribs being angled significantly. Therefore, do you think it is a better idea to use the belly to make a mould from? I feel if I were to use the back, it's not going to look like the instrument it is copied from, and there will be issues placing the sound holes correctly.

[Michael Darnton]

@Wood Butcher That's why I suggested finding the most neutral-looking one--to find the violin most likely to correspond with the original form, with the least distortion. I don't think there's any reason the front should be better than the back: if the ribs were distorted then both back and front are affected. One other approach might be to look for the most symmetrical example, though that's not a foolproof idea, either. But there are examples which are obvious outliers, and those are the ones I'd avoid.

[HoGo]

I'm in similar situation as OP. I've got several dG violins in photoshop and comparing the backs and CT scans of backs or ribs (all resized to full size) and considering symmetries Plowden comes as one of the candidates. The symmetry is stunning. I remember R. Hargrave suggested looking at Du Diable years ago but that one doesn't have the corners as symmetric as the Plowden.

[Davide Sora]

I think the Plowden is a good starting point to derive the form, on the poster there is the CT scan of the ribs at the form level, definitely a big help, so you won't get distracted by the outline of the plates. In my opinion, the most important thing is the position and size of the blocks, without a rib scan you will not be able to have this information. You can superimpose this CT scan on as many other Del Gesù violins as you wish, to understand how to manage any differences in the outline of the plates.

[HoGo]

18 minutes ago, Davide Sora said:

 You can superimpose this CT scan on as many other Del Gesù violins as you wish, to understand how to manage any differences in the outline of the plates.

That's what I'm doing in Photoshop and they are all surprisingly close even the larger ones like Kreisler match Plowden form very well, just extend the upper andor lower blocks tiny bit and reshape the point blocks. Here is quick comparison of Plowden and Kreisler ribs I posted some time ago (Kreisler is the darker longer one)

[Davide Sora]

Yes, that's what I did too, and it's very informative. Another thing to understand the outline of the plates is to observe the protrusion of the corners from the tips of the ribs, which Del Gesù often let protrude a lot, and sometimes less.

[HoGo]

Here is another comparison of Plowden and Cannone. The cannone drawing was from internet I'm not sure how precise it is and whether it was corrected for symmetry, but it certainly adheres to the same basic mould shape everywhere except near corners.

[HoGo]

I just found out there are some nw interesting things published on https://www.strad3d.org/

Lots of good resolution photos and CT scans of whole plates rim and also crossesctions of two Strads and the Plowden del Gesu. Nice set of precise measurements as well. I just put the back photo into Photoshop and after quick resizing the basic four measurements of back are within 1 or 2 10ths of a mm from the published numbers. Looks like there is very minimal distortion in the pic.

I'm working on some simple drawings of the Plowden and will post them (may take few months to finsh it).

BTW, queation for experienced makers.... The ribs of the Plowden appear to be quite bulged in some areas. In such case would you prefer using back as main source for outline of instument "inspired by Plowden" or the ribs CT crossection (taken through the middle of the ribs) which would result in slightly larger outline when overhang is added.

[David Beard]

There is a theory that he could perhaps have worked from one mould.

On the other hand, there is no evidence to support that, only the conjecture that it be a possible thing.

We don't know. He might have used twenty moulds.  The thing is, we don't have them.  We're only guessing.

[HoGo]

We know about the "one mould theory" but in this case I'm interested in approach at copying a certain violin (not exact bench copy). I personally lean towards using back purfling as the main source of outlines not the ribs CT.  I'm not sure if the bulging of ribs was there (at least partially) when the instrument was built (so the perspective mould should copy that) or majority of the bulging appeared later as result of the years under tension and back is closer copy of the mould.

[Bodacious Cowboy]

3 hours ago, David Beard said:

There is a theory that he could perhaps have worked from one mould.

On the other hand, there is no evidence to support that, only the conjecture that it be a possible thing.

We don't know. He might have used twenty moulds.  The thing is, we don't have them.  We're only guessing.

While I understand the interest in this matter from a historical/scholarship point of view, to me it's kind of irrelevant. The important (practical) thing for me is that I can use a single middle period mold to build any model I've wanted to from the Biddulph book, just as accurately as I could with a dedicated one piece mold for each one. Just by packing out the neck/tail blocks as required.

[JacksonMaberry]

If I were trying for a close copy, I'd start with the belly - it's the "face" after all.

[Michael Szyper]

4 hours ago, Bodacious Cowboy said:

While I understand the interest in this matter from a historical/scholarship point of view, to me it's kind of irrelevant. The important (practical) thing for me is that I can use a single middle period mold to build any model I've wanted to from the Biddulph book, just as accurately as I could with a dedicated one piece mold for each one. Just by packing out the neck/tail blocks as required.

Love that approach.

[Davide Sora]

8 hours ago, HoGo said:

We know about the "one mould theory" but in this case I'm interested in approach at copying a certain violin (not exact bench copy). I personally lean towards using back purfling as the main source of outlines not the ribs CT.  I'm not sure if the bulging of ribs was there (at least partially) when the instrument was built (so the perspective mould should copy that) or majority of the bulging appeared later as result of the years under tension and back is closer copy of the mould.

1 hour ago, JacksonMaberry said:

If I were trying for a close copy, I'd start with the belly - it's the "face" after all.

The top plate must be taken into serious consideration for the position of the Fs, the stop length at the notches, the cornerblocks position, and the overall dimension, as it is the most decisive for the sound. However, the back generally has a better outline and being the first to be glued to the ribs, it is probably the most faithful to the contour of the form (pretending there is no off square and rib bulge...). So what I would do is start drawing the outline of the top, overlay the outline of the back drawn on transparent paper and choose the curves that look best to me making a patchwork, not necessarily using one half or the other but also assembling single areas, respecting the widths of the violin but not necessarily the correspondence with the centerline of the various individual parts. You will thus have 12 parts to assemble (6+6) i.e. the two halves of the upper bout, the two halves of the lower bout, and the two C's of each plate. This is how I proceed if I don't have an original form or a CT scan of the ribs available, but if I had them I would have no doubts, I would use that to make the form and I would leave the interpretative work of the outline to my eye and inspiration. For me, talking about a copy only makes sense if it is an exact reproduction (distortions, asymmetries, errors, etc.), otherwise the result will always be a new model, so take care of the measures and elements that are important for playability and sound and don't worry too much about everything else.

[JacksonMaberry]

57 minutes ago, Davide Sora said:

the first to be glued to the ribs

Not in my method, but possibly in the Cremona method, I don't know. 

I like the outline of your approach to developing a model based on limited information. My approach is similar in some ways, I think, but different in some details. In a nutshell, I like to measure and draw, back and forth, until I have what could plausibly have been the original drawing.

[Don Noon]

1 hour ago, Davide Sora said:

However, the back generally has a better outline and being the first to be glued to the ribs, it is probably the most faithful to the contour of the form 

The back is also far stiffer than the top, and more likely to retain the original outline... unless the maple shrunk a lot.

Life is so much easier NOT trying to make close copies.  Or cellos.

[David Beard]

From an historical perspective, it seems to me that the back outlines are more 'chosen' clean designs.   The top outlines have a greater aspect of 'following' whatever vagries are introduced by the unsquareness of the sides, which in classical work can be very significant.

[Davide Sora]

14 hours ago, JacksonMaberry said:

Not in my method, but possibly in the Cremona method, I don't know. 

I like the outline of your approach to developing a model based on limited information. My approach is similar in some ways, I think, but different in some details. In a nutshell, I like to measure and draw, back and forth, until I have what could plausibly have been the original drawing.

We will never be absolutely sure what was glued first, if the back or the top. One of the reasons that make me think it was the back is the better correspondence with the outline of the form (Stradivari), even if as Don Noon says it could simply depend on the fact that the maple deformed less than the spruce, who knows.

Regarding the method of drawing a new model, As for the method of designing a model, I think any approach is valid, if it leads to results that satisfy you and to an instrument that works. Whichever method you use, the key stage for me is to leave the final drawing hanging on the wall for some time, to see if I will still like it even with a fresh mind from the drawing stage, making corrections if I feel the need.