Page 17 - Summer2019

concatenation of phonetic units and also the first synthesizer
n ,....e: """::i“ 1 mo - based on replicating the serial ordering of articulations seen in
_21.;;_;; .:‘\ the human vocal tract, using a fon-n of demisyllable synthesis.
.e L“—H_,,._4i.;. ‘V’-& j-H ' — .1 Finally, as the examples above show, they synthesized the first
, 1"”./, ,:,:::_“;._‘, ,’,.,‘,‘,.,.,., " "T17 71 dialogue between two machines. all of these advances
""""°""‘”"”“‘°"“"""“ Ii? 7 is the undeniable deficiency that they seem to have relied on
/ ‘_ N1 coordinating and switching between multiple separate mecha-
. .1‘ .
I/5., .' V} nisms to produce the illusion of speech instead of controlling a
l‘,’_f :_,._ -. | ll} single unified model of the vocal tract, which partly accounts
rm 7 'j [ for the poor and unnatural sound quality.
;' I: ‘ u--- M ‘ll l Faber's "Euphonia"
- 1- :..,.___ .... ‘ _ _ _ _
l , g M‘ ,,.,_: , V 1 ‘£3 l ‘ hi the early mneteenth century, an Austrian mathematician
.4 l . "' “"'_:'r“ ””" "T l “ and astronomer, Ioseph Faber, came across Kempelen’s book
l ,. '" ., l ‘ . _ . . .
l[ n ' ‘ ‘ '~- - ~--"" l E ‘ ml and built a rephca of the mechanical speaking machine for
l ‘ i- V . . . . . . . .
ill ' " n,., W J n_,_ ‘ l l himself. Quickly realizing the problems involved in playing it
_. " \_“'_ j_-:_~;:_ -5;“ ‘ "_ ‘ i _ by hand, he determined to improve on the original design by
; ,__ \ A "~ l turning it into a keyboard instrument to automate the means
_ . :,. _ of control. Like Kempelens machine, he used bellows and a
g "1; _ . ‘G vibrating reed as lungs and larynx to provide an airflow and
1",‘: , sound source. However, he replaced the simple hal.f-open
‘ A" ’ L}; s‘: ' I‘ mouth tube that Kempelen shaped by contortions of the hand
,5 { H. .
M“ A _ kv_v_r_>_~_r._«u;_-‘.ug_» 2 ' _ W L . ‘ y with a fully configurable model of the whole vocal tract. By
7 77 , , ,'j_“,”' ",f‘,,‘,,:fj’_':,:, , _, IN, , ,1 " ll mountingsixadjustable metal blocks back-to-backinasquare
73”» 7 /V " 7 re -13’ ‘ 4 3 7: L’ 7' 7» ' ' box behind a pivoting tongue resting on a rotating jaw that
terminated in a pair of moveable rubber lips, he was able to
Figure 4. The “talking heads” of the Abbe’ Mical, exhibited in Paris create tube shapes with front and back cavities that could be
in 1783. The dialogue spoken by the two heads is written on the easily and directly related back to actual vocal tract configura-
curtain hiding the mechanism (see text for actual French): "1st Head: tions. As in a real vocal tract, the tongue, jaw, and lips moved
‘The King gives peace to Europe’; 2nd Head: ‘Peace crowns the King to create simple opening and closing movements in the front
with glory’; 1st Head: I/ind peace makes the people happy’; and 1st cavity of the model, incorporating natural articulatory con-
Head: ‘O adorable King father of your people, their happiness shows straints, whereas the blocks behind resemble a six-section area
Europe the glory of your throne." At the bottom of the display is the function that could be used to create a more detailed shaping
caption: "Talking Heads: A problem solved in mechanics that up to of the back cavity. Following Kempelen, Faber also added a
this day had been considered unsolvable, or at least very dijficult. The more realistic nasal cavity, inserted a rotating vane to inter-
Academy of Sciences has said in its report that these talking heads rupt the airflow for trills, and used a moveable lever to alter
can throw the greatest light upon the mechanism of the vocal organ the effective length of the reed, raising and lowering the pitch
and on the mystery of speech. It added that this work was worthy of or cutting off the reed vibration completely.
its approval by its novelty, by its importance, and by its execution.’
Illustration reproduced from the Bibliothe'queNationale, Paris, France, The real ingenuity of Fabefs talking machine, however, lay
with permission. Copyright Bibliotheque Nationale, Paris, Gallicu, in the control mechanism. The jaw, tongue, lips, six blocks,
ark:/12148/btv1b8410437r. nasal opening, wne, and reed were all controlled by a set of
vertical rods such that the continuously adjustable heights
of the rods entirely determined the configuration of the
This mechanism is remarkable in the history of speech syn- vocal tract model at any point in time. Faber connected the
thesis for several original contributions. The Abbe Mical’s rods to akeyboard with 14 keys and 3 pedals, using a grid of
“talking heads” constituted the first automatic programmable cross-patched levers to transform each single key or pedal
speech synthesizer. They were the first synthesizer based on depression into a predetermined setting for multiple rods. He
Summer 2oi9 | Aisuuuueu-riusigy l 15