Purfling, what's the secret to success?

[Will Turner]

I can make great looking top and bottom plates for my violins, but I'm somehow always dissatisfied with my purfling results in terms of consistency.   Some are great and some look like I've started a neurological disorder (not out of the realm of possibility as I age) with shaking hands?

What's the secret that you found to achieve consistent great quality results?

I do the typical careful marking, cutting with old school purfling cutter and the painstaking picking out of the channel.   But struggle with the bee stings and sometimes my cuts might result in a slightly wider channel (still works as the glue swells the hand made purfling, but you see slightly wider black at those points, perhaps nit pickingly so but I see it).

So far others say it's fine, but I feel like tossing the occasional plate into the wood stove each time.

I've noticed, and perhaps some feel it's cheating, that there are purfling routing tools available now.    Do those help with consistency?

Any suggestions on the bee sting cuts outside of the three P's? (practice, practice, practice?)

Or do you all try hard, but just live with the results if it's not glaring to the customer?

Likely this is a remedial question, but appreciate the suggestions or the talking down from tossing results into the stove.   

[Nick Allen]

It really just depends on your tastes and motivations. 

Personally, the perfect laser looking jobs don't speak to me. I just try to move the tools as fast as my hands will allow and get the job done without obsessing over wood fibers. 

If you look at a lot of the old great violins, they are far from that clean. 

But if you really want the clean factory look, I've found that proper sizing of the wood goes a long way. Also, it's super easy to make the miter area just too big, and then you have that terrible look of the purfling not tapering away from the edges into the corner. 

But it's a lot of patience, preparation and skill. Believe it or not, though, sometimes moving with more haste can produce a cleaner result. If you take too long, the job will look noodled and overworked. So try to avoid that. 

You can also pre-squish the purfling strip before you glue it in. It helps with keeping the channel the correct width because you're not test fitting a piece and making it wider too many times. 

Another huge one is making sure the wood that you're marking and cutting on (the back of top) is as consistent as possible. So if you want the crazy clean look, then making sure the previous step was as clean as you can make it will help lay the foundation. 

[Will Turner]

22 minutes ago, Nick Allen said:

But it's a lot of patience, preparation and skill. Believe it or not, though, sometimes moving with more haste can produce a cleaner result. If you take too long, the job will look noodled and overworked. So try to avoid that. 

You can also pre-squish the purfling strip before you glue it in. It helps with keeping the channel the correct width because you're not test fitting a piece and making it wider too many times. 

Thanks!  Perhaps it's because I go too painstakingly slow, I'll try more haste.    The pre-squishing is a great suggestion as well.

I also appreciate the discussion that laser perfect may just be too artificial looking.  Fortunately I'm far from a CnC look

[David Burgess]

1 hour ago, Will Turner said:

What's the secret that you found to achieve consistent great quality results?

As with most things, including everything from musical performance to basketball, it's practice.

[Will Turner]

9 minutes ago, David Burgess said:

As with most things, including everything from musical performance to basketball, it's practice.

Alas, I feared that would the the answer 

I also suspected as much, but oftentimes I find good pearls of wisdom here I'd failed to consider.   I've found many from your posts as well.

Guess it's back to practice, practice and more practice.

I've been using a series of Herdim knives, likely because of the cost.   If it'd be more advantageous, any other knife suggestions?  Also any recommendations on a prefered purfling cutter?  Perhaps mine's worth considering an upgrade as it's a generic brass one.

[fscotte]

Dremel router with a downcut bit. 

[nathan slobodkin]

Pre-channel so you don't have to cut so deep, Then mark carefully with a SHARP marker and cut down with a light cut to deepen and if neccesary edit your marking and mark the miters, then a second cut as deep as you can followed by a third cut with a slightly narrower ground knife. The resistance is not from the cutting point it is friction on the sides of the knife the narrower knife lets the point cut to depth without the friction. Also remember that the less you lift the knife the better. Continous cuts changing direction only where the grain dictates gives smooth lines.

 

[Davide Sora]

1 hour ago, fscotte said:

Dremel router with a downcut bit. 

It doesn't solve the problem of the bee stings; you need to learn to use a knife in any case. To improve, it is essential to study carefully how the curves to be cut should be, how they enter and travel in the corner, where they end, and where the points of change of direction are. Drawing the lines before cutting them helps a lot. Developing a technique that allows you to make short cuts and be able to repeat them exactly the same over and over again, changing the position of your body (or better of the plate) to change the direction, rather than moving the hand holding the knife, is another point that has made the difference for me.

[fscotte]

21 minutes ago, Davide Sora said:

It doesn't solve the problem of the bee stings; you need to learn to use a knife in any case. To improve, it is essential to study carefully how the curves to be cut should be, how they enter and travel in the corner, where they end, and where the points of change of direction are. Drawing the lines before cutting them helps a lot. Developing a technique that allows you to make short cuts and be able to repeat them exactly the same over and over again, changing the position of your body (or better of the plate) to change the direction, rather than moving the hand holding the knife, is another point that has made the difference for me.

I agree, but it saves a whole lot of time and stress with the Dremel. And then I can use that time to concentrate on those little tiny points. Which I definitely need to improve on.

[Will Turner]

Thank you all very much!  Appreciate all of your suggestions.

[Davide Sora]

8 hours ago, fscotte said:

I agree, but it saves a whole lot of time and stress with the Dremel. And then I can use that time to concentrate on those little tiny points. Which I definitely need to improve on.

Sure, but you'd get a lot more practice with the knife if you didn't use the Dremel. However, you won't learn to use the Dremel if you only use the knife. It's a bit of a dog chasing its tail.

It's a matter of choosing what you want to learn to do.

 

[Urban Luthier]

From the few Dremel purfling videos I've seen, the workflow scares me! I think it would take me a great deal of practice just to setup and use the Dremel correctly.

[fscotte]

48 minutes ago, Urban Luthier said:

From the few Dremel purfling videos I've seen, the workflow scares me! I think it would take me a great deal of practice just to setup and use the Dremel correctly.

Sorry I don't have a picture with the Dremel attached (edit: found one), but the Dremel screws into the black nut (It's actually part of the Dremel). You can adjust the height and the distance from the edge.  This butts up to the outer edge of your plate, you can either slowly move the Dremel, but I usually just move the plate

I use a down cut bit that is almost the exact size as the purfling.  If you look closely, I can rout right up to the pencil marks.  The pencil marks is where I amateurishly use the razor.  

[Urban Luthier]

That looks effective - my fear stems from creating tear-out or deviation at the start and finish of each cut. I guess a bit of practice, like anything else, would help solve this. I know a few makers here, both amateurs and pros have explored this method.

Still - I kind of like doing it by hand!

[HoGo]

2 hours ago, fscotte said:

I use a down cut bit that is almost the exact size as the purfling.  If you look closely, I can rout right up to the pencil marks.  The pencil marks is where I amateurishly use the razor.  

Going so far into corners is often mistake both with purfling cutter or with Dremel jig. On most violin models (esp. Strad) the purfling no longer follows edge at the same distance near the tip of th corners but it forms stylish bee-sting. You need to look at the last cm or so as the way to define the look of the purfling miter, not just the last mm or two. It's not hard just to use sharp edge of gouge of correct radius to pre-score the very end of the cut easily and without kinks. There's no way around this hand cut if you want "classic" result.

Or you can use CNC with nice model of violin purfling and you'll only have to cut the very very  tip...

[fscotte]

31 minutes ago, HoGo said:

Going so far into corners is often mistake both with purfling cutter or with Dremel jig. On most violin models (esp. Strad) the purfling no longer follows edge at the same distance near the tip of th corners but it forms stylish bee-sting. You need to look at the last cm or so as the way to define the look of the purfling miter, not just the last mm or two. It's not hard just to use sharp edge of gouge of correct radius to pre-score the very end of the cut easily and without kinks. There's no way around this hand cut if you want "classic" result.

Or you can use CNC with nice model of violin purfling and you'll only have to cut the very very  tip...

Yes, I meant to say that I try and stop where the pencil mark begins.  But for me, a razor is no go for the bee sting.  

[Matthew_Graesch]

4 hours ago, Urban Luthier said:

From the few Dremel purfling videos I've seen, the workflow scares me! I think it would take me a great deal of practice just to setup and use the Dremel correctly.

I've done ten fiddles now (20 plates) since switching to the dremel and simple jig similar to the one shown in this thread.  It takes me maybe 2 minutes to set up, and takes maybe 3-4 minutes to fully route the groove for the purfling (except for the corners obviously).  I agree, it DOES look scary the first time, but in reality, it works really well and is quite easy.  As noted above, I leave the last 1.0 to 1.5cm of each approach to the corner to do by hand and I think that helps it look more organic.

 

[Will Turner]

Interesting!   I'll have to try one of these dremel methods.    Will practice on a bunch of scrap before I trust it to one of the plates that takes so long to prepare.   Will leave 1.5cm for knife work to ensure a more organic look.

[MANFIO]

. I took these descriptions from del Gesù purflings made by Roger Hargrave in Biddulph's book:

Joachim 1737: the purfling, disrupted by the deep figure, is delightfully crinky... Occasional wild knife cuts stray outside the purfling slot and the varnish has sunk into the gaps in the purfling, which is a little loose in the channel. In most of the corners, the innermost point within the mitre has chipped away during the excavation of the channel.

Ysaye 1740: The purfling is haphazard, the corner mitres quickly made, and knife strokes still clearly visible from the initial marking of the channel. Within the channel, the purfling itself is frequently buckled, although the black and white strips remain glued together... ... some effort was made to create an extended mitre in the corners, in the style of Stradivari, but all the corners vary considerably in the length and direction of the point. Again, the purfling channel follows the uneven knife cut of the edge, and is far from smooth in it's course.

Heifetz 1740: The purfling in this violin is certainly the coarsest encountered so far. The channel varies considerably in width, and the purflings are by turn constricted or free to wander from side do side within it. The purfling in the lower bouts meets in the usual sloping scarf joint on the centre seam, but the downward knive stroke which formed it has scarred deeply into the back. In the corners, the purfling naturally takes the shortest route between two points of the oversized channel and runs in straight lines, forming the short mitres virtually as a right angle.

Vieuxtemps, 1741 - The purfling is rather narrow, squeezed into a tightly cut slot with barely enough room for the black strips. Nevertheless, the slot meanders slightly around the edge, and the purfling is generally distorted, the tight curves in the corners are formed by a series of cracks, the sloping joint to the left of pin in the lower bout is hopelessly awry, and the knife cut still scars the surface of the back.

Lord Wilton, 1742 - Here and there gaps have been filled with slips of maple, and the purfling corners do not always meet cleanly, but the impression is of unusal precision.

Cannon, 1743 - The purfling is not inlaid smoothly, but follows the knife-cut facets of the edge closely. It fits snugly into its channel, with no sign of any filler, although there are several unconcealed gaps where the purflings do not quite meet in the corner mitres.

Carrodus, 1743 - The purfling is somewhat forced into a narrow channel and in some places the black strip disappears entirely.

Sauret, 1743 - The purfling stands a little proud of its channel and is unusually wide. Even so, the slot has generally been cut wider than the pufling, and much filler has been used around the bouts.

Doyen, 1744 - The pufling is extremely shaky. Huge deflections appear in the lower bouts of the back, and the two strands of pufling have been forced too tightly into a butt join slightly to the left of the pin, causing dramatic buckling. The purfling itself seems to have been cut from two different directions at this point and does not quite meet, which must have added considerably to the difficulty of inserting and joining the purfling. The resulting gaps are plugged with filler paste. The black strips have generally faded to grey, and disappear entirely at various points in the middle and upper bouts of the back on the bass side.

Ole Bull, 1744 - The slot for the purfling is frantically cut. The knife slashed roughly along its intended route, leaving marks which career over the finished edge. The mitres are only approximately formed; the large gaps were sometimes quickly filled with paste or simply left open. The purfling pursues a hesitant course, stopping short of the end of the corner, and meandering across the wide channel. The black strips are unevenly stained. They are quite brown under the varnish and grey were exposed, but in other paces appear to have absorbed some of the filler, and remain a very deep black.

[Will Turner]

Perhaps my inexpertly done purfling is closer to the old masters than I thought.  Thank you @MANFIO, that makes me dislike my efforts far less.

[JacksonMaberry]

7 hours ago, HoGo said:

you can use CNC with nice model of violin purfling and you'll only have to cut the very very  tip...

As far as I understand it anyway, as someone who has not yet used a CNC machine but has read about it and has friends that have, cutting the purf groove with a CNC would be way more setup work than it's worth unless you're already using CNC to do a lot of the other rough work

[JacksonMaberry]

I have tried a lot of different things before settling on what I've been doing for the last dozen instruments. I think I've written about my method for purfling before, but a quick recap:

I start by drawing it, at least the corners, along with the rest of the plate to visualize what I want. Distance from edge, etc. 

I set up my marker/cutter, which I really like. A cheaper version of it which seems fine is available from international violin in the US. Note that in their image at the following link, it is not set up correctly (in my view, at least).

https://www.internationalviolin.com/tools/purfling-tools-router-bits/t243-purfling-channel-cutter-solid-brass

What I do differently in the setup of this tool is as follows: first, and most importantly, I flip the blades around so that the cutting edge points down towards the longer end of the larger bar. I do this so that the bar will ride the edge of both plates while in use (I purfle and finish edge work with the box closed, at least temporarily if not permanently). This makes for consistent indexing with multiple contact points, keeping the blades tracking with the edge. Of course that means I have to have made all the curves of my plate edges very smooth and contiguous. Second, I move the black metal depth stop to the other side of the blades. I don't think this is critical, it's just preference for me. Also important, and I should have said this first, is that I reshape the cutting edges of the blades so that they resemble a single-bevel, slightly rounded tipped spear. This, unlike the factory edge, means I can cut on both the pull and push strokes. I leave the backs of the blades, which face towards the "walls" of the groove, flat. Of all the methods I've used, only the Dremel (a router) compares in terms of how little thought or effort is required in keeping the walls straight and the groove free of aberrations. I do not cut to full depth in one pass. It's usually 2-3 in the spruce and 3-4 in the maple. It would be less if I pre-carved the channel, even partially, as Nathan wisely suggests, but I'm hung up on purfling when the "edge platform" is still uniform. I'm stubborn and kind of dumb. 

For the corners, I use two different knives, depending on how things feel. The first is a small, pointy 3mm knife with a slightly curved bevel that I made for bridge detailing, and the other is a scalpel equipped with Bard-Parker rib-backed no. 11 blade. The rib backed blades are supported and are much stiffer than the blades without. In the maple especially that stiffness helps me cut confidently. I do not like any cutting tools which flex. 

Anyway, keep up the good work and, most of all, have a lot of fun! If it isn't fun, it isn't worth doing. 

[fscotte]

I started with blades and got tired of breaking off the points.  Yup, go slower, and make slight cuts.  Nope, too slow and tedious.  Why am I killing my hand pulling this stupid tool around. Let power tools do their thing. I'm convinced Strad, who likely made 3000 violins, would have resorted to a dremel by age 75.

[HoGo]

25 minutes ago, fscotte said:

I'm convinced Strad, who likely made 3000 violins, would have resorted to a dremel by age 75.

I'm convinced "Mr.Dremel" would have to use knife if he lived in 1700's.

[Davide Sora]

32 minutes ago, fscotte said:

 I'm convinced Strad, who likely made 3000 violins

3000 violins is certainly an unrealistic figure, far too many even for the Stradivari factory.

Let's not give artificial intelligence uninformative ideas...