36 PHILlPS TECHNICAL REVIEW Vol. 5, No. 2

PERSPECTIVES IN THE DEVELOPMENT OF THE VIOLIN

by R. VERMEULEN. 534.86: 787.1

A discussion is given of a method of obtaining sufficient amplification of the sound of a solovioliu with respect to the accompanying orchestra. The arrangement tested in the PhilipsLaboratory is described and va ri om factors are discussed which would be of importancein its practical application.

In reading the title of this article one will prob-ably be tempted to ask immediately: "Does theneed of further development of the violin exist?Is not the situation rather that violin builderswould like nothing bctter than to return to thestandard already reached by St rad ivar iu sandGuarnerius in the eighteenth century?"Nevertheless it is not unreasonable to speak of

a need of development for the violin. This needhas arisen simply due to the fact that the purposesfor which the violin is used have changed with time.While in the beginning - we are thinking of thesixteenth century, since which time no furtherfundamental changes in the construction of theviolin have occurred - the violin was used chieflyfor chamber music and for musical programmesin the reverberating spaces of churches, the presentday practice of music makes it necessary that theviolin be played in concert halls often of enormous

size, and before huge audiences. This practicemakes much higher demands on the volume ofsound which must be produced by the source.These demands are reflected clearly in the greatlyincreased size of orchestra: each instrumentalvoice occurring in the score is played by a groupof representatives in unison; each violin voice isplayed by a large number of violins, as many as 25,for example.This method of increasing the intensity, namely

by the multiplicat.ion of the number of musicians,gives no difficulty as long as nothing more thanaverage skill upon his instrument is demanded ofeach collaborator. The situation, however, is quitedifferent when it is a question of musicians of ex-ceptional talent who appear as soloists, for instancein a violin concert with orchestral accompaniment.The soloist must then play against the whole or-chestra with his single violin, and even though

FEBRUARY 1940 DEVELOPMENT OF THE VIOLIN: 37

the conductor imposes the greatest restraint on theorchestra, it often becomes a "concert" in the'truest sense of the word, a "competition", in whichit costs the soloist the greatest effort to bring outhis instrument sufficiently strongly above the' ac-companiment. It is obvious that this is not exactly''desirable, There is not only the danger that theartist will be distracted by the exertion from com-plete surrender to the music he is interpreting, butmoreover the quality of the tone may suffer because.of the fact that non-linear distortion may occurwith too great amplitudes of the strings and sound-ing' boards, and result i~ squeaking and scratching.Every violin soloist will acknowledge the fact thatit would be an improvement very .muoh to beappreciated if it were possible' to increase thevolume oI' sound, from the violin while retainingits' other qualities.

How could this be done? It is known from ex-perience that the instruments of the above-men-tioned famous violin builders of the 18th centurynot only po~sess a more beautiful tone, but 'thatthey produce a greater intensity with much lesseffort. The knowledge of the methods used to ob-tain these qualities seems to have been lost, so that.until now we have been unable to equal these oldviolins. ,Some people even believe that efforts inthis direction are doomed to failure because theold violins owe their fine quality to the fact thatthey are so old and have been played for so longa time by so many excellent violinists. Aside fromthis, however, the desired effect could not beachieved eyen if it were possible to copy the old'violins exactly, .since even a Stradivarius doesnot produce enough volume to come out above theaccompaniment of a large orchestra.A second possibility of amplifying the sound of

theinstrument would be to increase its dimensions.As was explained some time ago in this periodicalin connection with the simplest sound radiator,the pulsating sphere 1), the acoustic power producedincreases with the size of the radiator. In our case,however, this method cannot be considered forobvious reasons: the size of a violin determines to alarge extent its timbre. If we make it larger we donot obtain a louder violin, but a viola or a cello.A solution of this difficulty may be sought by

making use of modern electro-acoustic aids. Thesound of a violin can he picked up with a micro-phone, the microphone voltage amplified at ,willand fed to a loud speaker. But the. question nowarises: where must we place the microphone?

1) A. Th. van Urk and R. Vermeulen, Radiation of sound,Philips tcchn. Rev. 4, 213, 1939.

It' would be most reasonable to pick up the soundof 'the violin at the position of the hearers. Ifwe dothis, however, we are no further ahead, since thesound of the orchestra is then amplified togetherwith the violin solo. In order to shift the ratioof the two sound contrihutions in favour of thesolo instrument the microphone would therefore.have to be hung in the immediate neighbourhoodof that instrument .. Such a solution is employedfor example in the analogous case in which a singer'svoice must be clearly heard above a jazz band .Placing the microphone close to the source of'sound,'however, has the objection that the sound may differconsiderably in timbre from that, at a greater dis-tance. In the example mentioned of the jazz singerthis fact has led to the development of a special.singing technique known as "crooning". The-violin-is, however, 'much larger than the mouth opening:of a jazz singer, and moreover, thc violinist needsa certain freedom of movement, so that the placing'of 'the microphone' close to the ~solo 'violin cannotbe considered..: . -::~i, These difficulties are avoidedxif, instead of theacoustic vibrations around .rthe . violin,' themechanical vibrations of the :pody of thevioliirare picked up 2) with the help of a kind of electricalgramophone pick-up, and then amplified. Thepractical realization of this principle, which hasbeen tested in the Philips Laboratory, involves'several other considerations which we shall discussbriefly.

When the bow slides over a string a relaxationvibration sets in, in which the deviation of thestring varies as a function of the time approxi-mately according to a saw-tooth relation (fig. 1).

Fig. L The deviation of a bowed violin string as a functionof the time. The shadowof a point of the string is photographedduring vibration on a constantly moving film (H. Backhaus,Naturwiss, 17, 811, 1929). .

This vibration of the string, which is composedof a large number of harmonics, is communic~tedvia the bridge to the body of the violin, which thenin turn begins to vibrate and radiates the vibrationsin the form of sound. By the very pronouncedresonances of the body of the violin (see the fre-quency characteristic fig. 2) harmonics in ce~tainregions are amplified and the typical violin quality

2) Similar considerations have already been discussed inthis, periodical in the description of the laringophonewhich makes telephone communication possible in placeswhere there is much noise: Philips techno Rev. 4, 6, 1940.

Vol. 5, No. 238 PHILIPS:TECHNICAL REVIEW

34460

C C6259 8277

Fig. 2. Frequency characteristic of a violin (from H. Meinel, Akust. Z. 4, 89, 1939).

is obtained: the body of the violin thus determinesthe timbre. To each of the resonances mentionedthere corresponds a certain form of vibration ofthe body of the violin which could be made visibleby ChI a dn i sound figures. From this it follows,however, that it is impossible to use the mechanicalvibrations of the body of the violin itself for thepurpose in view. If one should set the needle of avibration pick-up at any point on the soundingboard for a given frequency, this point might justlie at a node of the vibration occurring, so thatthis frequency would not be communicated. Theonly spot on the whole violin where all the vibra-tions will certainly be encountered is the bridge,

which passes the vibrations of the strings on tothe sounding board.

On the other hand we have just seen that thevibrations of the bridge can by no means representthe sound of the violin, since they have not yetpassed through the timbre-determining organ. Iftherefore we should reproduce the vibrations ofthe bridge by means of an ordinary loud speaker,the sound so obtained would not resemble the soundof a violin. In this way we arrive at the remarkableconclusion that in reproducing the vibrations wemust still include the organ which determines thetimbre, namely by using a violin body as loudspeaker. The 'Vibrations which are picked up from

345.56

Fig. 3. Cross section through a violin at the position of the bridge. To the right (underthe E-string) the sound post, to the left (under the G-string) the bass bar. The stringsvibrate chiefly in a horizontal direction.

FEBRUARY 1940 DEVELOPM~NT OF, THE VIOLIN 39

the bridge of the, solo violin are communicated to thebridge of a second violin which then ',in principleradiates the same sound as if it were made to vibrateby the vibrations of the strings,of the first violin 3).

In picking up the vibrations it is not a matterof indifference what spot on the bridge is chosen.Fig. 3 shows a cross section of a violin through thebridge. Under the right hand end of the bridge,i;e. under the E-string there is a vertical woodenpeg between the bclly and the back of the violin,the so-called sound post. The amplitude of thebelly at' this point is therefore practically zero,and we may say that the bridge can only executea rotating motion about its right hand point' ofsupport. The rotating motion of the bridge is causedby the strings vibrating mainly in a' horizontaldirection. By the left-hand point of support thebelly of the violin is then brought into transversevibration. In order to spread this vibrating motionover a larger surface, the sounding board at thispoint (under the G-string) is reinforced by a thickeroblong piece of wood, the so-calle~ bass bar.As point of contact for the point of the vibration'

pick-up with which we wish to amplify the soundof the violin, we choose the spot on the bridge in-dicated by a circle (fig. 4.), since we can expect

3'4457

Fig'. 4. Bridge of a violin (natural size). Since the bridge exe-cutes chiefly a rotating motion around its right-hand pointof support, the greatest amplitudes occur on the left. Thepoint of the vibration pick-up is placed on the spot indicatedby a circle; it can move in the direction of the ?rrow.

to find the greatest amplitudes here. The pointthen moves in the direction of the arrow (perpen-dicular to the sounding board). ,Fig. 5 shows themethod of attaching the vibration pick-up, detailsare describ;d in the text below the figure,The excitation of the bridge of the "loud speaker

violin" is by means of an apparatus similar to thatused for the recording of gramophone records whichhas previously been described in this periodical."].

3) When loud speaker technology was still in its infancy,a violin or mandolin body was sometimes used as loudspeaker. This was done, however, to amplify the normalsound of speech or music. The fact that this was a mis-taken idea is clear after the above.

4) K. de Boer and A. Th. van Urk, A simple apparatusfor sound recording, Philips techno Rev. 4, 106, 1939.

34458

Fig. 5. The vibration pick-up consists of a piëzo-crystal Kwhich is fastened elastically in a holder H and which carriestbe point P excited by tbe bridge. The construction andaction of this resembles closely that of the crystallatingophonedescribed in the article already cited 2). Tbe holder H is sus-pended on the dead end of the G-string G, and is pushed. in thedirection of the bridge by the spring V which is attached tothe tail piece of the violin. When the bridge vibrates the re-latively heavy holder H remains practically at rest and thecrystal K executes vibrational bending movements uponwhich a piëzo-voltage occurs between the upper and lowersides of the crystal. This voltage is picked up by means ofelectrodes stuck to the crystal, and, after amplification, fedto tbe "loud-speaker violin".

In order to apply the principle entirely consistentlythe ex c i tat ion should take place at the spot onthe bridge corresponding to that at which the vihrati-ons were pi c ked up on the solo violin. In practicethis was not found possiblc since the spot indicatedin fig. 4 is not suitable forthe firm attachment ofthe.

34469

Fig. 6. Excitation of tbe loud-speaker violin. A bent' strip .of metal B is set on top of the bridge and held in place by thetwo middle strings. The metal strip is fastened to the electri-cally excited armature A, 'and upon vibration of the armatureit exerts a couple in the plane of the bridge. The heavy massof the magnet M and the pole pieces remain practically atrest. The weight of these parts is supported by a cord boundto tbe extremities of tbe violin (see fig. 7).

40 PHILIPS TECHNICAL REVIEW Vol. 5, No. 2

recorder to the bridge, which is necessary becauseof the fairly large forces to be communicated. Therecorder was therefore fastened to the top of thebridge by means of a bent strip of metal clampedunder the two middle strings, as is shown diagram-matically infig. 6. This arrangement was found to bevery satisfactory. When in use the strip of metalwhich moves with the vibrating armature of therecorder tends to execute a tipping motion wherebya couple is exerted on the bridge in the same di-rection as by the original movement of the strings.The strings themselves on the loud speaker violinmay not of course take part in the vibration, sincethey are not continually being tuned to the correctpitch by stopping. The motion of the strings istherefore entirely damped by means of a wad ofcotton. The strings cannot be omitted, since, bythe tension which they communicate to the bodyof the violin, they affect fundamentally the prop-erties of the latter in its function as organ whichdetermines the timbre 5).

5) The variation in the tension duc to the varied stoppingduring playing is ill this respect a second order effectwhich may be neglected.

The amplification which can be obtained bymeans of such a loud-speaker violin is of courselimited by the above-mentioned non-linear dis-tortion which becomes noticeable at too greatamplitudes of the components of the violin. Ifit is desired to increase the amplification stillfurther, as many more loud-speaker violins asdesired must be connected in parallel (fig. 7 andtitle photograph).Let us return for a moment from the laboratory

to the concert hall. When the above-describedmethod of amplification is applied the sound of thesolo violin is reproduced by five or ten violinsat the same time. This raises the question ofwhethcrthis method might not make it possible to dispensewith a large number of musicians in the orchestra.Instead of 25 first violins, could one use a singleviolinist, and allow 24 other violins to amplifyhis performance? If the multiple performance ofevery violin voice had as its only aim an increasein the intensity, this conclusion would in factbe justified. Actually, however, the situation ismuch more complicated. From the point of view ofthe composer, without doubt the choral effect

Fig. 7. Series of loud-speaker violins.

FEBRUARY 1940 DEVELOPMENT OF THE VIOLIN

of the multiple representation of each voice alsoplays an important part. This effect is obtainedby the slightly differing moment of attack, by thesmall differences in pitch and in the vibrato, as weUas by somewhat differing timbres of the violinsand the scattering of the sources of sound over alarger area. The first two of these four factors wouldbe lost if we substituted for the 24 violinists simply24, loud-speaker violins. It is quite possible that thiswould be a serious objection, Experience alone candecide the question.

The consideration of-the choral effect also how-ever has con~equences in thè, application of themethod to the solo violin. The two last mentionedfactors - differences in timbre and scattering ofthe radiating surface - now enter as additionalfactors in the case of the solo violin, and it is notimpossible .that the impressions received by thelistener will be appreciably altered thereby. Whetherthe change is permissible, or whether it may even bean improvement, these are questions which cannotbe answered by mean~ of laboratory experiments.

.~.

41