Jump to content
Maestronet Forums

Andreas Preuss Super-Light-Violin Project


Recommended Posts

  • Replies 319
  • Created
  • Last Reply

Top Posters In This Topic

In my initial guesses, I had assumed that the French Factory Fiddle would be the stiffest-sounding.  Knowing that the Pareschi is the stiff one, this second test I would order:

1 Guarneri

2 Factory French

3 Lightweight

4 Pareschi

The lightweight fiddle I have the most conficence in identifying, as I have built some and am familiar with that type of sound.  And it's the most different.

Link to comment
Share on other sites

  • 3 weeks later...
17 hours ago, Don Noon said:

How heavy is the bridge?  Although I think you could get some help with taking some mass off the top of the bridge, really major changes probably need amazing wood and arching tweaks.

The bridge is 1.9g and 29mm high. 
 

Before testing another top, I have still some less time consuming experiments on my list. One is switching back to a traditional bass bar, and the expanded top block is also still on the list.

But yes, in the end there are some minor issues on the top arching which might be not so ‘minor’ in the end. 

Link to comment
Share on other sites

  • 3 weeks later...

After the last changes

Brighter and louder.

First I couldn't really see any difference between tis and the previous graph. However on a closer look things are visible. The peaks of the lower resonances are now ascending. This seems to be audible in a less pronounced bass register.

Though the distribution of the high frequency didn't change a lot the average became slightly higher than before. Another interpretation is that the valleys seem to be not so deep any more. Though not so dramatic in the graph, there is clearly an improved soprano quality. 

Increased loudness is probably visible at the dB level of the graph. Before -37dB now -30dB. (After looking at previous graphs, this might be more due to recording setup though I am always standing in the same spot with the mic in the same spot.)

With the last changes I had hoped to get higher amplitudes in the frequencies between 2000 and 3000Hz which unfortunately didn't happen. I still think that the resistance to the forces trying to 'fold' the violin body is responsible for the whole high frequency range or at which frequency the highest center peak is located. (Looking through previous graphs I posted in this thread I had once peaks around 3000hz and this was when cross bars were installed. There seems to be a simple connection between top plate cross stiffness and the overtones in the 2-3khz bridge hill region. Basically this is confirmed by the research on the bridge hill. This means that the cross arching above the ff and its thickness are pretty important.)

Weight is now 305g without chinrest. 

Lines in orange indicate for orientation the levels in the previous graph.681731457_NCV141.thumb.png.1c5f0d2a3ff604524abcac4b86c3f900.png

For those interested I attached the recorded sound playing a g major scale at 5m distance to the mic.

NCV 142 g major scale.m4a

Link to comment
Share on other sites

  • 2 weeks later...

@Don Noon the report on the upper rib reinforcement (or you can call it also massive upper block enlargement.)

Here is the latest change in the construction. Because I think it is not good if the neck can wobble around too much I reinforced the upper ribs. Though one might think that the closed box is pretty solid there, I wanted really to know what happens if I make the upper ribs  muuuuuch stronger. 
This was also motivated by the thought that the new concept violin had compared to classically built instruments a weaker point there because of the imbalance between neck overstand (10mm) and rib height(20mm). This makes not only the neck leverage on the back much stronger but the whole area becomes structurally more unstable. 

This in mind, I started to rework the upper rib area. I left the spruce linings and filled everything with balsa wood (not to add too much weight). This technique follows the idea of composite materials. A Rubber foam square beam  wouldn’t be tough enough to withstand let’s say a heavy weight, but if the same beam is fitted precisely into a square paper tube it will take much more weight without deformation.

Then to imitate the paper I glued from the inside a second rib. (see technical drawing) 0A05C22D-908A-4AB7-BF09-CFB46EBC9E1A.thumb.jpeg.99ade0ad68342b3ad607f2c15f0844f5.jpeg

From the outside there are the following layers

1. maple veneer (0.2mm)

2. Linings 2.0mm x 8.0mm) and the ‘in between’ filled with balsa.

3. 3 layers of bent balsa (grain parallel to the ribs) 7.0-7.5mm

4. 1mm maple ribs anchored in the top block. 
 

This massive change added 20g or 7% which is a lot. 
 

For a ‘crash test’ I tried to string up the violin without back in this condition. It would hold the string forces pretty well but the neck would still almost go down the top. (Before I m made the upper rib reinforcement the e-string alone would pull the fingerboard right away flat down on the top.)

 However, the body was closed  normally keeping the projection at normal height of 27mm 

After stringing it up sound was noticeable different. For those interested in making a guess, go ahead and let me know by PM. 

Hint: Not what I wanted to achieve but still quite interesting.

total weight: 325g

technical data and sound graph coming up soon.

 

Link to comment
Share on other sites

Posted (edited)

So here is the graph shortly before I took the back off:

209974424_NCV146beforereinforcingupperribs.thumb.png.85cf4beebaa50f3d4fe9c9f10b37ae0f.png

And 24 hours after strining up the violin

931280510_NCV150(2)46mmbridge324mmstringlength.thumb.png.0eb4d310cfe934a09c1ddfb51b3209a8.png

I marked the difference I see in the graphs.

1. In the lower register the peak around 700 Hz is not sticking out so much any more. If this has for the overall sound any relevance seems doubtful to me because this peak, though it is always present and a little bit unusual for its frequency never had any massive disturbing effect. The first notes on the E string are in this region and have a little more ring than the other notes.

2. Biggest surprise was that the area around 1.0 - 1.5 kHz lowered. In higher resolution graphs this becomes more apparent. 

3. I had hoped that I could shift the center of the bridge hill into higher frequency region, which didn't happen. However the 3 small peaks (circled yellow) nicely stayed in their place. However the whole mountain got steeper flanks and this translates somehow more soprano like voice.

Altogether this massive change a least didn't worsen anything noticably. If anything was 'missing' it was the dark depth of the G string. The G string became 'flatter' but at the same time more sizzling or vibrant. A not too strong wolfwhich was there before became weaker almost to the degree that it disappeared.

D- string was never the weakest point on this violin. Now it blended better with the brighter G string.

A string was so far the most problematic string. With the rib reinforcement it improved the most of all 4 strings, gaining clarity and getting a better response than before.

E string didn't change too much, but got a little more ring (or maybe not)

Not too massive changes in the sound, altogether a bit lighter and brighter on the G string and  especially on the A string.

 

Summary: For the low ribs it seems to be good to reinforce the upper ribs. If the sound changes are due to a better anchoring of the neck or just the stiffening of the upper body portion is difficult to estimate. Because the top linings were already tripled before I would roughly say that the more radical reinforcement didn't have too much effect on the framing of the top. However, if I will do this reinforcment as a a part of the 'new concept' I will make the transition from thick to thin more smoothly. I might consider to cut the wall thickness of balsa woo down towards the corner blocks  in another experiment. 

total weight w/o chinrest 321g

String angle 159.5

Bridge height 31mm

Bridge feet width 46mm (sic!)

bridge weight 1.6g

String length 324mm

single tailgut tailpiece 

neck overstand 8mm (this was actually changed along with the upper rib reinforcement

Jargar superior strings for all strings

 

I will show more graphs soon of different sound adjustments which I think are highly interesting.

Edited by Andreas Preuss
add setup data
Link to comment
Share on other sites

So here is where I got after some serious adjustments. 1629474112_NCV150(7)46mmbridge327mmstringlengthsoundpostnottooclose.thumb.png.5b42e0da5e0aa2de975a135908f9288e.png

What I like most is marked in red. I was able to adjust the region between 2kHz and 3kHz. We played the violin repeatedly and the improvements were pretty audible when playing and listening.

There was more sizzle in the sound and altogether the sound got more 'density'. Both are things I find very desirable for a really good violin sound. 

Here are the interesting points for this setup:

First I switched the single tailgut tailpiece back to a normal Tempel boxwood tailpiece. All efforts to use instead of the 46mm bridge were negative for the sound. Last not least I set the bridge below the f nicks to gain 3mm in string length (making it 327mm) for the purpose to increase the string tension. The combined changes were able to increase the peaks in the 2-3kHz band. Repeated hear and play testing somehow confirmed that only the combination of the three changes brought the most desirable results.

My interpretation of what happened is that first the lengthening of the strings would add the amount of pressure to make everything work. We tried as well to go over the 327mm and this would diminish the effect. In those terms the 327mm string length could be regarded as 'focus string length'

The idea of augmented bridge feet width came from a paper on the bridge hill where the bridge feet width is given as a factor to trigger the frequencies in that region. However the bridge didn't bring any notable changes for the sound when the upper ribs were not reinforced. Now under the new conditions it worked very well and so far another 42mm could not get the peaks in that region up. I assume that it has to do with the leverage action pivoting from the soundpost on the bass bar. For a really thin top rather weak is strong enough. Secondly the wider bridge feet might send off the bending waves in a different direction with less interference. 

This explains as well why the single tailgut tailpiece doesn't work. It allows an easier pivoting motion of the tailpiece and hence for the bridge. 

For all the other graphs I made in different setups the peaks in the 2-3khz region would change to different patterns, In the transition hill region would be always lower but the width and depth would vary depending on the adjustment. Last not least i had an eye on the peak around 700 Hz because it was so far always pretty much sticking out. There were some adjustments which could make its amplitude lower than the peak around 500Hz. In the setup shown here I was able to align it on a slope with the peaks below which seemed to be good enough.

Summary

So far all the adjustments I made at previous stages in the experimental alterations to the new concept violin could by no means trigger anything in the bridge hill region. Without trying to build a explanation in physics, I would put it like this: Areas on the body under biggest stress must be strong enough to withstand it. Lack in stiffness in such areas will diminish the over tone spectrum band width and amplitudes. However I would say that on a violin with normal rib height this massive reinforcement might not be necessary.

The same graph in higher resolution. 

824072079_NCV150(7)highresolutiongraph.thumb.png.b226bfd2878d3e2cc44559810f30bfec.png

 

 

 

Link to comment
Share on other sites

I do not understand what you have done to the back and the details there. But will just mention that a strong 700Hz region resonance is usually a sign of a thin back. In normal violins th mode is called the C4 and it is a kind of ring mode in the back plate. The bridge feet act in phase for this mode so the resonance is stronger while tapping the bridge downwards instead of sideways. 

You might be playing the instrument. Then the bowing force is a mix of the two. 

However 700 Hz can also be something else. A dipole upper lower bout comes a bith higher usually in frequency. It does radiate sound as the bouts are different in size.  

Link to comment
Share on other sites

14 hours ago, Anders Buen said:

I do not understand what you have done to the back and the details there. But will just mention that a strong 700Hz region resonance is usually a sign of a thin back. In normal violins th mode is called the C4 and it is a kind of ring mode in the back plate. The bridge feet act in phase for this mode so the resonance is stronger while tapping the bridge downwards instead of sideways. 

You might be playing the instrument. Then the bowing force is a mix of the two. 

However 700 Hz can also be something else. A dipole upper lower bout comes a bith higher usually in frequency. It does radiate sound as the bouts are different in size.  

The back is pretty normal. Center around 4.2 upper and lower bouts down to around 3.0.

Yes, I take usually the graphs by bowing half note scale 5 times in a row. 
 
For the cause, I have no idea. It just seems that the signature modes of the new concept violin don’t fall into the usual regions because of some different construction elements. I don’t think that different signature modes are blocking the path to a good sounding instrument. Only if amplitudes are out of control this sets a red flag. 

Right now I am quite happy how far I got and hopefully with a few more constructional experiments I can finish the project completely. 

Link to comment
Share on other sites

Posted (edited)

Latest progress. The peak around 700 Hz disappeared, and around 3000Hz there is finally a stronger resonance zone. Altogether I would see it as a progress. 

 

Playing the instrument I would describe it as vibrant and with some bow pressure it gets an ear piercing sizzle. Some notes on the half tone scale became a little bit difficult to play, but I didn't try to adjust it in detail. I couldn't say that the weaker resonance around 700 Hz now changed a lot. Maybe it makes the sound more flexible and gives a bit more possibilities for sound colors. (Maybe) 

Since all is experimental I want to know now if I can get still more peaks between 3kHz and 4 kHz. Got some ideas....

 

technical data are 

total weight without chinrest 315g

string angle 160

strings Dominant forte ADG and Lenzner 0.28

46mm bridge 

Bridge weight 1.5g (unchanged from before)

single gut tailpiece

 

low resolution graph 

1042371614_NCV160.thumb.png.632203ab2a64ecf8279b926b4dd4a020.png

high resolution graph 

1346471636_NCV1608192resolution.thumb.png.12ae8c546d671a77da088b9ee0b01cbe.png

Edited by Andreas Preuss
Add bridge weight data
Link to comment
Share on other sites

How much does the bridge weigh?  The only way I can think of to increase the high frequencies on a given violin is with a lighter bridge.  

I can't tell for sure what the B modes are on your graphs.  At first glance, it looks like B1+ is ~470 Hz and B1- is 390 Hz, both very low.  However, it could be that B1+ is 520 Hz (normal-ish) and B1- is very high at 470 Hz, and the peaks in the 300 - 400 Hz range are all CBR related.  

Link to comment
Share on other sites

On 8/25/2021 at 9:25 PM, Don Noon said:

How much does the bridge weigh?  The only way I can think of to increase the high frequencies on a given violin is with a lighter bridge.  

I can't tell for sure what the B modes are on your graphs.  At first glance, it looks like B1+ is ~470 Hz and B1- is 390 Hz, both very low.  However, it could be that B1+ is 520 Hz (normal-ish) and B1- is very high at 470 Hz, and the peaks in the 300 - 400 Hz range are all CBR related.  

I added the bridge weight to the technical data. I used the same bridge as for the last setup, so it is the same weight (1.5g). 

I find it interesting that your predictions from structural changes are pretty exact, but when I present changes only as data in a graph your conclusions on the structural changes are not so precise. But this being said, it is really difficult to guess structural changes from a graph.

In the last structural change I was experimenting with was the bass bar outlay again. The x-shaped bar extended to the borders works overall very well, it has only the disadvantage that it does not 'connect' to the upper and lower blocks. With this thought in mind I reopened the top and installed two short bars from the edge of the plate pointing into the middle of the V-shaped slot.

61F453AD-B413-40FB-8F26-F077FE0CAE45.thumb.jpeg.6c18f96c13f8f8d16fc5da7c0fc826c6.jpeg

However, this aside, I think the last graph is really interesting because a small accident happened. The violin got a crack below the treble side f-hole half way down to the border. So first I was thought 'Himmelkruzidaibelnochmol' (=dammed!) but then just out of curiousity I made an audacity recording to see what is going on, when the apparent tension in that area is released.

So here we go, here is the new graph after closing the crack. It could shift the bridge hill center to 2.5 Khz but the resonance around 3kHz is definitely not there any more. Secondly the 700Hz peak reappeared. Something to think about. 460020946_NCV160aftercrackclosed.thumb.png.d5ad0d3b7a08eda014612c0170f3a306.png

 

And here is a recording at 5m distance. (I’d say not bad at all)

NCV 160 short bars from block end no crack.m4a

—-————/———-

2021/8/28
Got it played by a good musician and she was pretty impressed by the general loudness, balance and sound quality. Encourages me to continue.

Link to comment
Share on other sites

  • 2 weeks later...

So here comes the next constructional change. I made a much more solid lower block.87CF7D71-8E88-4FCF-AAAE-53AC38C8ABF0.thumb.jpeg.a8be5be40e94bef4a915cc63006b3a8b.jpeg

Though it didn't bring dierctly sound changes I had hoped for, there are interesting positive and negative differences.

To start with the negative effect, the violin became more difficult to play which means if no attention is made to control bow pressure and speed the sound will 'flip'. 

The positive effect is difficult to describe. There is a hue of smoky flavour paired with some ear piercing 'grittiness' which is strongest on the g string and gradually gets less when climbing up scales. One could also say the sound got more texture. I have to add that it took me considerable time to figure out the best adjustment. Checking in between by making a FFT graph things looked a little bit confusing because I couldn't detect great differences between different adjustments though under the bow there were clear differences in playability and response.

Adjustments I made could achieve quite different sound characteristics. For one of the adjustments I set the soundpost pretty much inside and at a great distance from the bridge with a good amount of pressure. This resulted in a pretty coarse and dark sound which I couldn't control any more with my bowing technique. (But could be just THE thing some crazy skilled players are looking for.) On a total i found that the current construction allowed more possibilities with sound adjustments. (Somehow the very first setup of the the super light violin had zero possibilities for sound adjustments with setup parameters) 

Anyway here is the last graph taken from an adjustment I feel (so far) most comfortable with.

Total weight 316g

String angle 160.5

bridge weight 1.4g

46mm bridge

Jargar superior strings 

string length 329mm

sound post normal position1266481213_NCV170smokysound.thumb.png.8afbfc77845697dba2f2b96c6b087830.png

 

For second reference a quite different setup giving the sound a maximum of texture. This needs however extremely good bow control. 

the only changes to the setup above are 

string angle 160.0

sound post position far inside with tension.688350774_NCV170texuredsound.thumb.png.37ee1f786b9e504669fdd231a483ada3.png

Thinking now about further alterations. 
 

One thing is now certainly necessary and can be done rather quickly. The enlargement of both end blocks was in total probably too massive causing problems in the playability or requiring very good bow control. So to check if this is related, I want to trim both blocks to some degree smaller.

After that the average level of frequencies 2-4kHz could be still I mproved and I want to re-bent the top plate into a higher arching, maybe about 2mm higher.  This sims also at correcting some deformations which developed in the meantime. 

Edited by Andreas Preuss
correction of the string angle figure (marked red now)
Link to comment
Share on other sites

10 hours ago, Don Noon said:

Are you really using a 46mm viola bridge and carving it down to the abnormally light weight of 1.4g?

AF0407D1-B749-41FE-B849-CFB9D295298E.thumb.jpeg.97e2a72386a5eb55573f586e9d45e825.jpeg

This picture is from the very first setup. Later I thinned it down on a sandpaper board just a little bit coming down to 1.5. Not necessary to say that I chose the lightest blank I found in my shop. Thickness is around 3.8 - 3.9mm. After rounding off the shoulders it came down to 1.4. 3C18A022-CCB8-4DF7-8E90-F5C5C6801C0F.thumb.jpeg.c77800ac7a094b7a82644031b4272010.jpeg

if I would make this bridge for a customer I would reinforce it with carbon rods to prevent bending or collapse. For a better design I’d need to ask Despiau to make me another custom made bridge.

This bridge, for what reason ever works definitely better. I tried several other standard (41.5 - 42.0mm) violin bridges, and somehow the sound was always more diffuse or kind of slightly muted.

The next top I am going to make will have 46mm f-hole spacing and set circa 3mm further down where I believe is a sort of ‘focus point’ for the bridge position on this particular outline. 

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.

Guest
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.

Loading...
 Share

  • Recently Browsing   0 members

    No registered users viewing this page.




×
×
  • Create New...