Jump to content
Maestronet Forums

Andreas Preuss Super-Light-Violin Project

Recommended Posts

3 hours ago, Anders Buen said:

There is one article on japanese experiments with an x shaped bassbar in a violin. It is documented in the Benchmark papers in acoustics edited by Hutchins et al. The experiments were from the 60ties, I think.

I think there were many good experimental ides in the past. However, nobody tried to put the combination of different new ideas into a new concept. *)

One of the most important things I have learned from the new concept violin project is that trying to judge alterations on a structure which is too stiff and solid doesn’t really show what is needed to come to a well sounding instrument. The new concept violin started as a unfunctional violin, too light and structurally too weak resulting in a sound which was ‘woofy like hell’. (Or the sort of violin you would rather throw into the garbage with no hesitation.)

Well educated guesses in strengthening different parts (in particular the ribs) would result in clearly audible differences, without ever looking at a spectrum. 

I learned that ‘good sound’ requires a structure which needs to stretch in for a while. (I had a very similar experience on a Guadagnini copy I had made 8 years ago.) So my current hypothesis is that the ‘stretch in’ is absolutely necessary and must be governed by permanent sound calibration in different parts to bring out the ‘full sound’. In practical terms I abandoned the idea that a ‘functional’ sound with a sort of ‘dense overtone range’  can be achieved the moment the violin is strung up for the first time.

In this context I see the x shaped bass bar as a functional element to ‘distribute tensions’ in a favorable way. On instruments made in a violin school manner this doesn’t make much sense and is most likely the reason why the idea of research paper got forgotten over time.



*) My new concept violin in includes now:

  1. Exterior mould construction to build an instrument with a top larger than the back with
  2. laminated ribs
  3. guitar type of laminated lightweight neck, set as neck graft
  4. x bass bar
  5. laminated baroque type finger board
  6. balsawood for the corner blocks
  7. bent top
  8. assymetric f holes
  9. thinner sound post
  10. irregular lining structure with top side stiffer than on the back
  11. Asymmetric bridge
  12. New type of tailpiece (a recent addition)
  13. height adjustable lower nut (for sound calibration purpose)

and I dare to say that the combination of all those features is necessary to make it function to its full potential)

Link to comment
Share on other sites

  • Replies 319
  • Created
  • Last Reply

Top Posters In This Topic

  • 3 weeks later...
Posted (edited)

So here is the latest graph after a few sound adjustments though. I waited by purpose a month, knowing that the playing characteristics need to settle in. Presumably this is a stretch in effect.

So now its a flat walnut back. Graduations made from the outside with the goal of creating a thicker Island in the center. So I was removing wood only in the upper and lower bout. In the center I touched only the bass side C bout area. So very roughly (measured with a Hacklinger) it looks like 5mm between the C bouts except on the bass side where it goes down to 4.5. Upper bout thinnest spot around 2.8 Lower bout around 3.5. Rather quick increase to the thick area of the center.

The adjustment included a new tailpiece which I had made for experimentation 30 years ago! (See pictures below) The idea was to string it up on one string over the lower nut. Additionally the string spacing was set so that the strings would follow the straight line from bridge notch to the lower saddle. On normal tailpieces the outer fixture holes are outside this line. This is the reason why the tailpiece moves when the E string is loosend. I might later compare sound graphs with a traditional tailpiece and the one string tailpiece. It seems to increase slightly the lower resonances, but I need to check this maybe on several instruments.

Sound has a pretty good body, though it needs some adjustments to the bowing arm to draw the full sound out. 

Technical data 

  • 333g
  • string angle 160 with a neck overstand of 10mm
  • bridge height 32 (assymetric bridge model)
  • thin soundpost 5mm
  • new tailpiece without E string adjuster. 





Edited by Andreas Preuss
Add photos
Link to comment
Share on other sites

I was thinking yesterday the whole day what still needs to be improved. 

  1. low resonances are still a bit too strong and I should do something about this.
  2. High frequencies (I guess somewhere between 2khz and 4khz) are better than before but could be still a bit more. 

Since changing the back doesn't promise too much tonal changes it is probably time to think again about the top. For a quick experiment just putting the ground should be interesting. If Hargrave is right his POP method should do the job. 

Otherwise it would be time to make a new top with a higher arching and small corrections in the entire arching shape.


However, looking at the spectrum from where I started, there is definitely a huge improvement. But it is hardly the same fiddle anymore after replacing top back and massive alterations to the ribs. Slowly a 'picture' emerges how to adjust all parts to each other while maintaining a fairly low weight. 

Link to comment
Share on other sites

1 hour ago, Don Noon said:

Looks a lot more like a graph of a viola than a violin.  B1+ especially.

Yeah, it has a rather dark sound and as I said in my second comment that's one of the first things I need to work on. 

There is a chance that the viola=ish sound is a little bit amplified by the new tailpiece and I will check it out tomorrow.

Despite, I like the quality of the sound quite a bit, because it has a nice overtone crust. If possible I want to get more 'texture' though. ('texture' as mentioned by Ray Chen when testing 5 high class instruments at Tarisio)

Link to comment
Share on other sites

8 hours ago, baroquecello said:

Have you tried stringing it like a viola once? There is a market for good sounding small violas, isn't there?

Actually when it was first finished, the violin turned out to have many nice viola characteristics and yes I am thinking to make violas with an enlarged model. But before I need to figure out how to make everything on a violin to get a really good sound.

Stringing it up like a viola doesn't work because string tension will be too low.

Link to comment
Share on other sites

15 hours ago, Don Noon said:

Looks a lot more like a graph of a viola than a violin.  B1+ especially.

Here is the graph with a normal Tempel boxwood tailpiece. Looks not so different to me.

Playing the violin feels a little bit better in the midrange.



Link to comment
Share on other sites

On looking again at the last two response plots, there might be another interpretation that I think makes some sense: 

B1+ could be the peak around 700 Hz... extremely high

B1- would be the peak around 470 Hz... relatively high for  a violin

CBR would be the strong peak around 390 Hz... normal frequency, but abnormally high amplitude

Looking for signature modes in a bowed spectrum is often difficult due to the granualarity of the bowed note spacing at low frequency.  It is more clearly seen in an impact spectrum.  Try tapping the bridge a bunch of times with a pencil or stick of soundpost stock, and plot it out at 4096 resolution.

Link to comment
Share on other sites

Basically I posted this already in the thread about rib height reduction to increase the air frequency. I also made a new thread on the question which posed itself to me.

So with the latest change I am very happy. The violin is definitely on the darker side, but all what was making it too viola-ish has basically disappeared. Despite my limited abilities as a violin player, it is really fun to play now, this means I am getting very close to the goal.


It is an experimental violin, so I have still some changes in mind simply for testing. 

  • What happens if I reduce the ribs further? Right now they are already very low at 26mm, but if want to switch back a back plate with an arching this has to be done anyway.
  • Recently I had the idea to use one leg of the X bass bar to stretch it to the Bermuda triangle of the wolf vibration below the bass side f hole.
  • If I have time I want to try a top with a different bent arching, because right now it looks compared to normal standards a little bit weird.
  • And finally (YES that's important too!) the fiddle doesn't have any varnish. I think I am going to follow there the foootsteps of Roger Hargrave at least for the ground.

So here again the latest graphs and pics first audacity graph is bowed half note scale, second is done with impact hammer, and the photo of the back for those who are interested in Thickness distribution patterns.1925800679_(5).thumb.png.ee944a023bb88909ee169a97fea341df.png492128110_(6).thumb.png.107d8816b533cde8aca38b98509c027e.png


Link to comment
Share on other sites

5 hours ago, Marty Kasprzyk said:

How much does this violin weigh?  

At that stage with a traditional Tempel boxwood tailpiece around 335g.

The completely flat walnut back is now 135g and switching back to an arched maple back I guess I can reduce the weight at least 20g.

Yesterday I reduced the rib height once more. Just out of curiosity I  tried once more my single tail gut tailpiece and this time it hadn’t such a negative effect.

I let it stretch in over the weekend and post the newest data on Monday.

Link to comment
Share on other sites

I let it stretch in over the weekend and readjusted it on the sound post position. I noticed that sound post pressure is changing the playing characteristics more than the position itself.

Right now I am pretty satisfied with the sound and for any corrections it is more about details. There are a few spots where the response is weaker or lets say one must be more careful to get the entire sound out of the fiddle. Maybe soon there will be the time to show it to high class player for a good critique.

Anyway getting much closer to the sound I have in mind. To put it in one word I like the expression 'dense sound' . 

Looking at the graphs I noticed that there seems to be more resonances in the higher regions now. In general I feel always a bit lost in reading and analyzing graphs. In the end I want to figure out just simple methods to steer the sound in the direction I want to have it.

First graph from half note scale played 5 times in a row1770549357_2021-05-17233803.thumb.png.35acb589e6a78dac7cda91ef64b3c87a.png

Second graph from impact hammer. I don't know if it has any meaning that from 450 up to roughly 3500Hz there is only a slight decline on the peaks. I find it interesting too that there is again a prominent hill in the region from 4000 - 8000Hz. Taking the Dunnwald descriptions of frequency bands this corresponds to 'roughness'. I suppose this is the reason why the violin has what Ray Chen describes as 'texture'. (For my own taste it could have even more texture)1717925046_2021-05-17234554.thumb.png.586d3309d9a2b0e471cac8006f5029de.png

Link to comment
Share on other sites

  • 2 weeks later...

Just for fun.

I replaced the back with a piece of copy paper and even managed to insert a sound post. I played it before and after inserting the post and there was not too much difference.


  Interesting to see how the paper stretches.IMG_8766.thumb.JPG.99bcc6132ffd41813fc30f2d71b5c4e1.JPG

The red lines indicate  where there is most stress. By touching the paper with a finger it seems that there is more stress on the straight lines in the upper half than in the lower half.IMG_8774.thumb.JPG.9992e8cba0fd3a3e326cac5a5aabd4e9.JPG

I made a graph as well in this setup.


I think it is important to locate the point from where high resonances decline. On this setup this starts around 1400Hz. On the best setup with a flat walnut back the decline started around 2500HZ. It seems that on really good fiddles this point is at over 3000Hz.

Without the paper (meaning without a back) it is impossible to tune up a violin. the fingerboard goes flat on the top. A Paper back holds the neck surprisingly well.

Sound was like returning to the very beginnings. Woofy, tubby no real loudness and colorless. 

It seems that the back must hold with the help of the soundpost  the treble side bridge foot in a fixed position. Only then the rotation movement of the bridge can work efficiently. To test it I crazy glued a piece of plywood across the c bout area. This improved the overall sound, but didn't make it sound like a normal violin.



Link to comment
Share on other sites

2 hours ago, Andreas Preuss said:

Without the paper (meaning without a back) it is impossible to tune up a violin. the fingerboard goes flat on the top. A Paper back holds the neck surprisingly well.

Tie a piece of  bailing wire to the end pin, then over a piece of bent copper, ( a penny?) to protect the gluing edge, then on to the neck heel, where a screw can attach the wire to the heel of neck.  Run a piece of wood across the lower corners to support the sound post, then place a stick in between the sound post brace and the lower block to support it from twisting.

Have you ever tried filling a violin with sand to kill the back plate and see what it sounds like?,,,, for one, it is very heavy. The back becomes really solid. You can diddle with the post to accommodate the solidity of the back,,,,, maybe something to see,, I don't quite remember,,

Have fun,, you enthusiasm is refreshing,

Link to comment
Share on other sites


So here is the next stage. The graph shows impact hammer spectrum.


The previous arched back is back on the fiddle, though with some modifications. Since I thought the back became too thin I installed 2 reinforcement bars in the upper half and a cross bar at the center made of 2mm bamboo forming a triangle. Except the cross bar, the reinforcements follow approximately the force lines found on the paper. (see above)

The previous back got quite deformed in an interesting way. It curved inwards upwards from the upper block zone. To fit it back to the ribs I decided to adjust the rib height to it. Accordingly I made a taper like on the ribs of classical makers. Because of the pretty steep deformation I had to reduce the ribs on the top block area 3.5mm compared to the height of the upper corner blocks.

I made also one reinforcement bar on the outside of the top for another experiment. (More of that later)

I think the impact hammer graph looks pretty interesting.

There is a double peak around the air resonance. No idea if this good or bad. The amplitude of the air resonance seems to be good. When playing an instrument with a strong air resonance one has always the feeling of volume and depending how the instrument is bowed it can sound a bit tubby. (Right bowing technique is a pretty complex theme)

Maybe most interesting for @Don Noonis the region between 1kHz and 2kHz which looks pretty elevated. 

For me the decline starting at 2.3kHz looks pretty good. It extends well to 10kHz and seems to contribute to a certain richness in the overtones. My personal preference is however a slightly rough timbre and I guess for this the decline must start in a higher region.

From playing the violin the sound is still a bit 'bass heavy' and from my previous experiments I would say this can be corrected with reducing the air volume slightly. (Maybe about 0.5mm-1.0mm rib height reduction)

Bridge is the same as before, weight is down to 287g. (Makes me wonder if the light weight can be anywhere detected on the graph.) As said before, minimum weight is not the final goal anymore and I am happy with anything below 330g. With a small chinrest attached the violin weights now 323g.

For the moment one achievement in this project is that it is possible to build a light violin which sounds good (don't say fantastic right now, but maybe I am getting there). The other achievement is that slight assymetry does NOT cause major trouble for the sound. However I don't think that a lighter violin must be assymetric, there is nothing which hints at this.

For completeness the graph from a bowed half note scale.390780346_2021-05-29000006.thumb.png.e832af8a76a459559cb7a675c6e4e9e4.png


Link to comment
Share on other sites

Ok, here is a little brain twister.

After I put the arched back on the instrument it still sounded a bit too deep. (at least for my taste) Previously,when the flat back was still on the violin,  I could correct this with reducing the rib height. So I thought because the arching adds volume compared to a flat back, the volume became too big again. 

Therefore I reduced the rib height and to be on the safe side i reduced just 1mm and only down from the top blocks. (now you might guess where my story is going...)

The sound definitely got worse and the violin sounds now much more like a viola. At the moment I have absolutely no clue where this comes from.

I will probably first check out the linings which became a sort of narrow during repeated rib height reductions.

Eventually I might flip the flat back on it again to see if it behaves the same.

Link to comment
Share on other sites

  • 2 weeks later...
On 5/29/2021 at 8:29 AM, Andreas Preuss said:

After I put the arched back on the instrument it still sounded a bit too deep. ...So I thought because the arching adds volume compared to a flat back, the volume became too big again. 

Geometric volume really only has an effect on A0 frequency, but not on other mode frequencies.  And amplitudes are something else entirely.

My take is that the high amplitudes between 300 and 400 Hz are what's making it sound "too deep", and that has very little to do with air volume.  Somehow, it looks like you are getting a lot of dB's out of the CBR mode and the B1- mode, in addition to perhaps the B1- frequency being lower than normal.  I'd say forget about volume, and concentrate on structural mass, stiffness, and the distribution thereof.

Link to comment
Share on other sites

10 hours ago, Don Noon said:

I'd say forget about volume, and concentrate on structural mass, stiffness, and the distribution thereof.

While I think this is in some way in general correct, there are from my perspective a few thoughts on this. Might be all nonsense in terms in physics but nevertheless.

Before I think about mass and stiffness distribution, I think it is very important that the forces moving the body must be balanced. Force comes from the strings and to balance them means to find the best sting angle. I work with bridge height, string angle which is governed by the overstand and lower nut height. In a more detailed setup the balance between upper and lower strings has some effect too and the distance between the two tailguts running over the lower nut. (Each time I replace a normal two string tailgut with my one-tailgut-tailpiece, the lower register becomes too strong.) 

I see in 'bad sound' first of all two important causes

  1. wrong air volume
  2. bad setup in the above mentioned terms of string angle etc.

So far I found that I could do much more with just those two components than altering structural mass and stiffness. And I can say now that I regard the air volume as 'the base' for everything, meaning that if the air resonance is too far off, other changes are more or less without effect. This became one of my basic rules: First set the A0 at approximately the right frequency range. (270 - 280)

When it comes to look at resonance peaks of the signature modes the amplitude of each might have some effect, but now my take on that is, that it is possible to 'overpaint' it with overtones. And this comes all from the setting of string angle. This became my second rule: After the air volume is set, balance the string angle. For this purpose I invented a new type of lower nut which makes it very easy to change its height for testing. (more or less plus minus 3mm)

What I try to achieve is the best clarity under the given conditions and if possible some sort of ear piercing sound. Both are in my view more dependent on just the string angle than anything else. 

It seems that in this setup environment the bridge height should always stay in the established limits between maybe 29mm and 31mm.

However I found that (at least for the new concept violin) an completely exaggerated overstand of 10mm worked  better.



So here we go. 

First I reduced the air volume once more (to get it from 267 to 275) and at first it didn't show the desired effect BECAUSE I didn't pay attention to the pitch. When I tried to make a higher bridge to ajust for it (the bridge would have been 35mm) I realized immedeately that this just makes things worse, so I corrected the neck angle by reglueing the back on the upper bouts. (And 'dang' the change kicked in)

Previous experiments with altering the string angle with my adjustable lower nut showed most of the time that the violin set up at a lower angle somehow sounds more resonant and therefore I decided to lower the bridge too. (Note: there might be a connection to the heavy and sturdy linings on the top side of the rib garland, so I can't say if this is a general rule)

With basically no changes in mass structure and its distribution I achieved now a pretty good sound (don't say this is perfect) which doesn't have any strong viola characteristics any longer. In between I tried to get a richer sound by reinforcing the bass side ribs only. Improvement was only very limited and maybe more self deception than anything else. Some experiments made me think that more weight in the bass side C bouts might give the bow more resistance without realy changing the sound.)

So here is the impact hammer graph


And the graph taken by playing half note scale with a strong bowing294871706_NCV107halfnotescalestrongbowing.thumb.png.254b58e2928509a012b44bcf5892037c.png

Actually under the bow the violin feels now pretty good. It is full and vibrant with a good bass support for all the notes. At least one thing I noticed from listening to Classical Cremonese instruments on YouTube (my favorite is the opening of ravels Tsigane) it seems to me that a common characteristic is this sort of depth always bordering to hollowness paired with a kind of smoky flavor)


In the end I am not ignoring your arguments, because now that after the setup achieved the best balance I could get, I am seriously thinking about better structural mass stiffness for the back. As usual with no scientific evidence I think that the back a kind of forms the bandwidth of the overtone spectrum and simultaneously can put the 'center' of the overtone range in the right spot. (when thinning down a back from the outside I have always had the impression (by ear) that overtones slowly descend and depending on how thin upper and lower bout get in comparison to the center, the bandwidth becomes broader) But I found out as well that there is literally almost no effect under the condition that the top is simply too thick.

In any case since recently I am testing the sound more on its flexibility and response with different bowings in terms of pressure speed and location between bridge and fingerboard. 

Technical data

Bridge height 30.5mm (reduced from 32) 

Average string angle 162 degrees

Overstand 10mm

Lower nut height 7mm

Total weight 282 

Weight with a special made chinrest 315g

For one of the next changes I am thinking to make a test with the Casein ground on the top. 

Link to comment
Share on other sites

I spent the whole day in readjusting the string angle of the violin. No structural changes.

So here is the result1496513209_NCV108Aftersecondsoundajustment(wettingexperiment).thumb.png.1be3719ab826773f800dede4ca73efd0.png


And the same graph in higher resolution784765382_NCV108Aftersecondsoundajustment(wettingexperiment)highresolutiongraph.thumb.png.d945d3f3caa9895ec64c4c68eeb2a341.png

The sound is now pretty clear and rich, quite satisfying. G and D string are still a bit too strong. 

Though this probably can't be demonstrated on a graph, only small changes in the pitch of the neck and/or the height of the lower nut made subtle but important changes in the coloration of the sound. In simple words I would describe it as drawing 'sound contours'. The final setup was most satisfying for overall balance, clarity and flexibility. 

The last addition I would like to get for this sound is more grit or spicy roughness.

Link to comment
Share on other sites

The graphs look abnormal for a violin, mostly in the extreme strength between 700 and 800 Hz, as well as the abnormal strength in the CBR/B1- area and abnormal weakness of the B1+.  I can only guess that it would sound somewhat abnormal as well, regardless of how satisfying you might think it is.

Try playing a good, "normal" violin for several minutes, and then switch to this one, and see if your impression changes.

Link to comment
Share on other sites

  • 3 weeks later...

So here we go for a comparison listening test.

I played g major scale on 4 different violins which are

1. French factory with good sound

2. Gaetano Pareschi 1943

3. Andrea Guarneri 1638

4. The new concept violin in its current state

recording was made with my iPhone at a distance of 3m in a different order.

so which is which? Or is the new concept violin sound wise so different? 

And here is a second comparison. The four instruments are played again in a different order (so everyone can ‘see’ if he/she can recognize the single instruments from the previous session.

This time I played first all instruments on the g string one octave then D string, A string, E string. Always in the same order. (Vl1 G string, Vl2 G string, Vl3 G string, Vl4 G string, Vl1 D string, Vl2 D string etc.)

I don’t think this makes a huge difference but it should be noted that the 4 violins had different strings but at least all synthetic from Thomastik. 

Will try to get a professional to play and compare the new concept violin soon. 


Link to comment
Share on other sites


Putting these large video files on the glacial MN server is not a great idea.  The first time I tried to download the samller file, it crashed; the second time it was successful after 10+ minutes downloading.  I didn't bother to try downloading the second file, which is nearly twice as big.

Either upload to YouTube and post a link, or strip out the video (we don't need to see carpet) and post just the audio file.

With the caveats that the recording quality is not very good, playing scales doesn't show off the real capabilities of the instruments, and I have no idea what to expect from the French and Pareschi, my guesses for the first video are as follows:

1) Super light

2) Guarneri

3) Factory French

4) Pareschi


Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.

  • Create New...