email: webmaster @platetuning.org

 Last updated:

1st.. Jan. 2021

Copyright  (C)

 www.platetuning.org

Example 1: 1-piece back [factory] violin, Ref. MT1

     This is a factory, ex-school violin that was selected because it had such a bad tone, even after a new bassbar was fitted. It was worth working on because of the pretty slab-cut one piece back (see right).

As received:

    Data before work: total weight 467 gm., very poor tone, especially on G and D strings, with a ‘thin' A string sound.

   As received: A0 = 286 Hz, B1- = 454 Hz, A1 = 467 Hz, B1+ = 574 Hz.

 The B1- to B1+ delta was 120 Hz, and this delta should ideally be in the range 60 Hz -  95 Hz.  So a 120 Hz  B1- to B1+ delta is too big.

 Fingerboard and belly were removed. Belly weight as received was 90 gm. and its Mode 2 = ~165 Hz and Mode 5 = 361 Hz.

  Back and bouts had no corner blocks: Mode 2 = ~180 Hz, Mode 5 = ~364 Hz estimated from tap tones of the back in bouts with neck. The neck block is intrinsic with the neck as was standard German factory practice in the 1800's.

Work done:   The neck block was widened to the standard width with 2 small white-wood blocks either side.

  The method followed here is outlined on the page ‘Back still in bouts'.

   Four corner blocks were shaped and glued in, and the belly's bassbar height reduced slightly. After thinning the belly it now weighs 85 gm., and Mode 2 = 154 Hz, Mode 5 = 340 Hz by thinning in the 2 'lungs'. How this Mode 5 was chosen is shown on the diagram / chart on the right: click on it.

   After thinning the 'lungs' of the back too the back plate in the bouts with neck weighs 245 gm., so back plate alone is estimated at 125 gm. Now Mode 2 = ~187 Hz (still high), and Mode 5 = ~342 Hz again estimated from tap tones of the back in bouts with neck.

 The charts and tables on the ‘Plate tuning 4 dummies' page allow the B1-, B1+ and ‘Delta' body-mode frequencies for a violin to be predicted with  reasonable accuracy ¹.

 The predicted and actual body mode frequencies are shown for this Sample violin in the table below:-

Example 1: 1-piece back [factory] violin, Ref. MT1 Belly Mode 5 340 Hz Belly Mode 2 154 Hz - - Back Mode 5 ~342 Hz Back Mode 2 ~187 Hz (Mode 2 of back is rather high) - B1- predicted 447 Hz B1+ predicted 525 Hz ‘Delta' predicted 78 Hz B1- actual 451 Hz B1+ actual 540 Hz ‘Delta' actual 89 Hz A0 = 289 Hz A1 = 466 Hz CBR = 389 Hz

 Violin playing tone after the plate-thinning work :-

   The tone is now good, and certainly unrecognisable from what it was. Good G string, better D and A strings. Good projection. Quite acceptable instrument now for Grade 6 to 8 exams.

---------------------

Odd shaped violin examples can be found here.

Footnote 1 : 

            For the model used here, dated May 2015, the standard deviation sigma is estimated at 17 Hz in predicting the ‘Delta' (B1- to B1+ frequency gap) from the 2 violin plate Mode 5 frequencies.

 For the later April 2017 model the standard deviation is 11 Hz.

Example 2:  Hopf violin copy,  Klingenthal, c. 1830.

     This very cheap 200 year old bench-built violin was chosen for work because again its original  tone was so bad. It had a bassbar crack most of the length of the violin .... but then it did not actually have a bassbar, just thicker wood there. The neck was 10 mm. too short and at too low an angle, i.e it was a ‘transitional neck'. There was only ‘inked purfling', still visible on he back.

  I've begun tidying up the front to put on a good new ebony fingerboard as the original was a blackened hardwood. Now the quality of the tone justifies it.

   I took this instrument apart to individual pieces, so the plate weights and tap tones are measured on the actual plates themselves and are not estimated. The plates are only crudely arched: no machine was used in the making of this violin.

    Six new and shaped blocks were added to the garland or bouts.

   After the belly plate was reduced in thickness it was 83 gm., and with a new low, shaped bassbar Mode 2 = 155 Hz, Mode 5 = 303 Hz.

     After thinning the back plate weighed a very low 96.5 gm, with Mode 2 = 171.5 Hz, and now Mode 5 = 340.5 Hz: an ‘octave' plate

   After assembly the playing tone was very pleasant but it had no carrying power so I removed the belly again and put in a new flat-topped bassbar, 10.5  mm at the centre, tapering down to 2 mm from a point 60 mm from either end. I also inserted new channels in the belly plate at the C-bout areas as it had none before, and that took it down to 78 gm.

   This took the belly Mode 5 to 320 Hz, and Mode 2 to 159 Hz.

 The charts and tables on the ‘Plate tuning 4 dummies' page allow the B1-, B1+ and ‘Delta' body-mode frequencies for a violin to be predicted with  reasonable accuracy ¹. The predicted and actual body mode frequencies are shown for this Sample violin in the table below:-

Example 2: Hopf violin copy,  Klingenthal, c. 1830 Belly Mode 5 320 Hz Belly Mode 2 159 Hz - - Back Mode 5 340.5 Hz Back Mode 2 171.5 Hz - - B1- predicted 439.5 Hz B1+ predicted 513.5 Hz ‘Delta' predicted 74 Hz B1- actual 453 Hz B1+ actual 550 Hz ‘Delta' actual 97 Hz A0 = 289 Hz A1 = 477 Hz CBR = 380 Hz

Playing tone after plate tuning work :-

    Easy to play, and very good for dance music as it responds quickly - with a very good ‘bite' to the sound. G and D strings are good, but not as strong as some other violins: slightly baroque sound to it. A delight to play.

   Sound wise, as good a violin as I could possibly hope for, considering its very humble origins.

Changes to body resonances with a new bassbar:-

    These figures are quite interesting in their own right:

Footnote 1 :          For the model used here, dated May 2015, the standard deviation sigma is estimated at 17 Hz in predicting the ‘Delta' (B1- to B1+ frequency gap) from the 2 violin plate Mode 5 frequencies.

  For the later April 2017 model the standard deviation is 11 Hz.