JPRS ID: 8234 TRANSLATIONS ONTELECOMMUNICATIONS POLICY, RESEARCH AND DEVELOPMENT

• Document Type: CREST
• Collection: JPRS
• Document Number (FOIA) /ESDN (CREST): CIA-RDP82-00850R000100010041-4
• Release Decision: RIF
• Original Classification: U
• Document Page Count: 35
• Document Creation Date: November 1, 2016
• Sequence Number: 41
• Content Type: REPORTS

APPROVE~ FOR RELEASE: 2007/02/08: CIA-R~P82-00850R000'1000'10049-4
;
i9 JANUARY ~i9t9 CFOUO 2179~ i OF i

APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
FOR OFf=1C1NL USE ON~Y
JPRS L/8234
19 Janu~ry 1979
~ ~ ~
TRAHSLATIONS ON TELECOMMUNICATIONS POLICY,. '
RESEARCH AND DEVELOPNENT ~
CFOUO 2/79) ~
.
U. S. JOINT PUBLICATIONS RESEARCH SERVICE
FOR OFFI~IA~. USE ONLY ~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02/48: CIA-RDP82-44850R000100014441-4
NOTE
JPRS publicntions contnin informaC'on prirt~trily from foreign
newspapers, periodicnls and books, but also from newa agency -
transmissions and bro~dcngt~. Materiels from foreign-language
sourcea are tranalated; those from ~nglish-language eources
are tranacribed or rEprinted, wiCh the original phr�sing and
oCher characteriatics reCained.
Headlines, editorial reports, and maeerial enclosed in bracketa
are supplied by JPRS. Procesaing indicators auch as (TexC] -
or [ExcerptJ in the firat line of each ieem~ or foLlowing rhe
lasC line of a brief, indicaCe how the original informaCion was
procesaed. Where no processing indicaCor is given, the infor-
mation was summarized or extracted. -
Unfamili~?r names rendered phonetically or CransliteraCed are
enclosed in parenthesea. Words or namea preceded by a ques-
tion mark and enclosed in parentheaes were not clear in the
_ original but have been aupplied as appropriate in conCexC.
Other unattributed parenthetical notea within the body of an
item originate with the source. Timea within itema are as
given by source. -
The conrenta of this publication in no way repreaent the poli-
cies, views or attitudes of the U.S. Government.
PROCUREMENT OF PUBLICATIONS
JPRS publications may be ordered frem the National Technical
Information Service, Springfield, Virginia 22151. In ordEr-
ing, it is recommended that the JPRS number, title, date and
author, if applicxble, of pL:~lication be cited.
Current JPR5 publications are annaunced in Government Reports
Announcements isaued semi-monthly by the National Technical
Information Service, and are listed in the MonChly Catalog of
U.S. Government Publications, issued by the Superintendent of
Documents, U.S. Government Printing Office, Washington, D.C.
_ 20402.
Indexes to this repor~ (by keyword, author, personal names,
title and series) art availab?~ through Bell & Howell, Old
Mansfield Road, Wooster, Ohio, 44691.
Correspondence pertaining to mattera other than pr~curement
may be addressed to Joint Publications Research Service,
1000 North Glebe Road, Arlington, Virginia 22201.
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02/48: CIA-RDP82-44850R000100014441-4
~ i_
~16LIOCRApHIC DATA 1~ Kepun No. 2 J, Hccipient'd Acees~ion No~
_ SNEEY JPR5 L/ 82 34
. t e~ t c e . xepcn aee
_ TRAN5LATIONS nN T~L~COt~4~fUNICAI'IONS ~ULiCY, RL"S~;ARCH 19 Janua 19 79
AND U~VELOPMEN'r , ( FOUO 2/79 )
- . Autba(~) 8. Ne~rfamin~ Ora~eis~tian Rcpt.
9. P~rlamio~ Or~~ni:ation N~me ~od Addre~~ 10. P}oject/Ta~lc/Work Unit No,
Joint PublicaCiong Etesearch Service '
1000 North Glebe Road 11. Con~r~et/Gr~nt No.
Arlington, Virginia 22201
lZ, Spoo~at(of Oraaaizuion N~me ~nd Addre~� 13. Type ot Repat & Period
Co~ered
AB BbOV@
t~.
tS. suppt.wue~er IV�e.
1 . Ab~u~e~~ ~
This serial report contains translations from the world presa and radio ~
_ relating to worldwide political~ economic and technical developments in
Celecommunications, computers, and saCellite communications. Coverage will
be worldwide with focus on rrance, Federal ~tepublic of Germany, United Kingdom,
Italy, Japan, the USSR, People's Republic of China, Sweden, and the Netherlands.
. Key iad~ ~od Doeumeoc Ao~lysia. 17a Deicriptor~
Worldwide
Computers ,
Satellite Communications
Electronics and Electrical Engineering
Telecommunicationa
Telemetry
176. Weocifiera/Opeo-Ended Tecma
17e. COSATI,FieW/Group ~9$~ C~ F~ 17B~ 22B
10. At~ilrbility St~temeot 19..Securi~y Cl~as (This Z1. o, of P~6es
For Official Use Only. R`Po") 31
Limited Number of Co ies Available From JPRS. ��C�r~~y ss z~. Pr~�
p Ps e
N L SSIF ED
?owa Ntiru fta~ol usco~.w~�oc ~o~:~�v~ ~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02/48: CIA-RDP82-44850R000100014441-4
I
FOR OFFICIAL USE ONLY '
JPRS L/823Q
19 January 1979
TRANSLATIONS ON TELECOMMUNICATIONS POLICY, ~
RESEARCH AND DEVELOPMENT
(FOUO 2/79)
CONTENTS PAGE
SUB-SAHAR~N AFRIC,A
GABON
Briefs
ADB Telecommunications Loan 1
USSR
Evaluation of Electroma.gneti.c Compe~tibility Conditi~ns in .
the imi,zstion of Radio Relay Line Construction '
~A. I. Kalinin, V. A. Shamahin; E7~QR03VYAZ~, Sep 78). 2
Complex Determination oP Tolerance ~a~ the Farametera of
a Co lete Color Telecasting Cha.nnel
. ~N. G. Deryugin, A. K. Kustarev; ELEIfl'R03VYAZ~ 'TS) 13
W~STERN EUROPE
T*PQT,Y
~Talki ~ Computer Developed by CSEIJr oP Turin
~Giuliano Ferrieri; L'EUROPEO, 24 Nov 78) 26
. ' 3 ' ~IrI - INT - 140 FOUO]
FOR OFFICIAL U5E ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02/48: CIA-RDP82-44850R000100014441-4
~OR OFFYCTAL USC ONLY
CABON
BRIEFS
ADB TELECOI~tUNICATIONS LOAN--Two loans amounting to 10 million unitis of
account, or approximately 2,857 million francs CFA have been graneed by
the Arabl Aevelopmenr Ba~k (BAD) to Gabon following an agreement signed
on 5 December in Abidjan ar this insCiCution's headquarters. These loans
wtll be used by the Gabon Office of Post and Telecommunications to finance r
a te.lecamnunicaCions development project in Port-Gentil, Franceville,
Moanda, Ndjole, Lambarene, Oyem, Bitam, Mouila and Kou1a-Moutou, as well
as to remodel and increase the Libreville Celephone neCwork. These loans,
which are guaranteed by the Gabonese governmenC, are fifth in a series of
AnB grants Co Gabon, now toCaling more rhan 23 milli.on units of account. ~
[Text] [Paris MARCtiES TROPICAUX ET MEDITERRANEENS in French 22 Dec 78 ~
p 3510~
' CSO: 5500 ~
1 .
- FOR OFFIGIAi. USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02/48: CIA-RDP82-44850R000100014441-4
FOR OFFICIAL USE ONLY
~ USSR
U1DC 621.396.43
EVALUATION OF ELECTROMAGNETIC COMPATIBILITY CONDITIONS IN THE OPTIMI2ATION
OF RADIO RELAY LINE CONSTRUCTION "
M> ~ c~~ :
~cK rn ~ ny a--- - srn ~ .
i ~
~g~ ~ ~y
~nc
. ;
Figure 2 ~
Rey: 1. ASR 4. TP ~
2. NLS 5. SLS
3. ISZ [artificial earth 6. GLS [hypothetical co~unication
satellite] line]
7. VTP
8. P ~ .
9. PU
~5
FOR OFFICIAL USE ONLY
~ F':ii. . . ~ . ' ~ . ~ .
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~Olt n~'~TCIAL US~ ONLY
Therefore, when egCabliehing eoleranceg, it seema pracrical Co proceed frnm
ehe presence of a hypothetical TsTV channel of average complexity (Figure 2).
In ie, thP hypotheCical communtcation ~ine GLS may consiaC, for example, of
a land radin relny or cable l~.ne of a total len th of 250Q km and a hypo-
CheCical sCandard a~Cellite c~mmunicaCion line ~4]~ In Chie aase, Che qual-
ity of the ourpuC TV image of a channel more complex than Che hypotheCical
~ channel will bQ aomewhat inferior th~n rhe quality sAt for th~ correapond:.ng
Cypes ot rec~ivers, and in Che case of a eimpler channel, i.t will be ensured
wiCh a margin of safPry.
Thue, tolerances for rhe paramete~s of a TsTV channel muaC be aelected in
such a way that, for n c,ertain peraenCage of time (leC ua say, 80ye), the
qualiCy of the color image of not leas than 4 pointa and 3 pointe would be
enaured aC the outpuC of the hypothetical channel for Che atationary and
portable TV receivers, reapective.ly.
Since the channel is calculated from the condition of obtaining a quality of
~ 4 pointa on stationary receivers, a quality of 3 pointa will be obCained for
porCable receivers only at the expenae of great tolerancea for Che parameCers _
of the receiver. Therefore, firat it ia neceseary Co establiah tolerancea
for the parametera of the TaTV channel with a sCationary receiver, and then
tolerances for the parameCers of a porCable receiver.
Classification of Distortions in a Color Televiaion Image. Distortione of
the color TV image may be divided inCo Cwo classea which we aha11 term con-
ventionally as achromaCic and chromatic.
AcromaCic distortiona show thenaelves in the color image in the same way as in
Che black-and-white TV image and are connecCed chiefly with errors in the
scanning and synthesis of the image. This group also includea brightness
noises. Various achromatic distortions, their causes and posaible places
of their originaCion are shown in Table 1. Chromatic distortions, in turn,
can be divided into two characteristic groups which are chiefly connected
with errora in color reproducCion in small and large areas of Che image.
Small areas can be convenCionally considered to be sectiona ~ith a minimum
size of up to 8-10 image elements, and large areas aections of over 8-10
elements.
We shall term color disCortion of amall image areas as heterogeneous distor-
tion. Although Chey shc~w themselves as changes in the color, these disCor-
tiona should not be treated as diaturbances in the fidelity of color trana-
miasion. The concept of the fidelity of color tranamission iCself refers~
strictly apeaking, only to the reproduction of color in large secti.ons of the
image. Heterogeneous distortions are perceiveli by the viewer not as inac-
curate reproduction of colors, but as various defects leading Co ttne deteri-
oration of the overall quality of the image analogous to achromatic distor-
tions. These are chrominance noise, color edging, eta. This also includes
drawn out prolongations which could occupy large areas, but, by the nature
of their perception, belong to heterogeneous distortions. The latter ahow
themselves in a color picture inde~.pendently from one another and are dynamic
in nature. The'data on various distortione are summariaed in Table 2.
16
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~OR O~~ICIAL US~ ONLY ~
Table 1 '
Typea of achromaric Overall Channel param NaCure of Sections of Chan-
No dieCortiong (grouped eters whose changes diatoreions el rhere param� .
channeX error) cauae distortioc~a eter deviationa
� are oesible
1 Coordinate Scanning nonlinear- Static
ites, etc ,
2 Die?.urbancee of in- Aeymmetry of inter- Dynamic ASK, P
. terle~Ced scanning la~ed scanning
3 Uneven sharpnese Aperture Di~tortfone Static
over the acanning of rhe camera and pic ~
patCern field ture tubea
G Repeated images Matching,Lcho signals "
5 Black-and-white Pulee interference Dynamic ,
doC~ and atrokea
6 BrighCnesa noise Itatio of Brightnes8 "
eignal to fluctuation _
noiae in 0.3-6.0 l~iz
band '
_ 7 Moving 1ow-frequen- Ratio of aignal to "
' cy interference background noise ASR, GLS,
S UnsaCiaf+actory re- Nonuniformity of AChKh " VTP, P
production of small [ampliCude-fre uency ! '
detaila and eharp characteristic~ in 0.5-
brightnesa changing 6.0 I~a frequency ban
9 Fine grid ~ High�frequency inter "
ference
1A Gradational Nonlinear dietortion " ' -
coefficient of the
brightneae channel
Diatortione of color in large areas ahow themaelves as diaturbances in the
fidelity of color Cranamiseion. Due to varioua cauaea, colora in the image
could change in different ways, and the resulting dietortione in color trane-
m~,ssion may inteneify or weaken, or even compeneate one anoCh~r. ~
The grouped parametere whose deviaCions from the rated values cause diatar- ~
tions in color transmieaion are liated in Table 3. Diatortione in color
tranamission arise as a reault of changea in the parameCera and character-
istics of the tranemisaion channel and deviationa from the rated values of
color eignals at varioua points of the channel. ;
The TsTV channel has two main groupa of signals which corre8pond to the
color coordinates in two color syatems, namely, the systems of the receiver
and the tranemissic~n channel. Consequently, it ie poasible to isolate two ~
grouped parametera. The first one includes various types of disturbancee
leading to independent changes in the signals corresponding to the color co- '
ordinatea R, G, B in the system of the receiver, and the second parameter
17
FOR OFFICIAL USE ONLY
_ ~
~
. i
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~Ott d~~ICIAL USE ONLY
Tgble 2
Typee of heterogeneoue Qenerul channel parumeCere 5~ctione of ahannel
nolor dietortione whoee ehangea caus~ dieeorCi.one where deviatione of _
(grouped chann~l error) peramerere ere poe-
eible
Chrominance noiee Ratio of the brightness eigr~al
to fluctuation noise in tha ASK~ GLS,
chrominance channel VTP, P
Diacontinuous colored Differential phaee S~me
fringe DifferenGial gain "
FluctuaCion noise in ehe chro-
minance band "
AChKh nonuniformiCy in the chro
minance band "
Miamatch of VCh [high frequency
characteriatica of prediaCor- ASK, P
tion and correction
Colored drawn out DifferenCial phaee ASK, G.LS, VTP, P
prolongationa ' Transient characterietic Same
fn the area eE low ~rid medium
~ frequenciea
� Miamatch of predistortion and ASK, P
and correctian aiCh reapect to
video fzequeney
Colored edging InaccuraCe matching of rasters ASK, P
Tranaient characteristic in
the area of high frequenCips ASK, GLS, VTP, P
Timing diacrepancy between the
brightnesa and chrominance Same
, signals
Partial or total TraAeient characteriatic of ASR, P
abaence of color in the chrominance channel
small image areae (color aharpnesa)
Difference in the re- Miematch of the direct and de- Same .
production of adjacent luyed aignals
linee or groupe of Difference in tt~e sensitivity o ASR
lines the heads of the video tape
recorder
leads to changes in the signals corresponding to the color coordinates DR,
Y, DB in the system of the tranamission channel. Changea in the signals ~
- cause diatortions which are dqnamic in nature.
18
~ FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~OR U~'F2CIAL US~ ONLY
Table 3 '
Grouped channel Generai channel parameeern Nature of Sections of chan-
~ No paYameter whose changea cguee dietior di.eCor- nel aheYe devia-
Cions in color rr~nemis- tion eion~ df parem-
I ~ miasion eeer~ are oesible
1 Color coordin- Amplitude of Che eignal of Dynamic ASK '
ates R, G, B primary color
Amplieude o� gamma-predis- " " +
torCed signal of primary '
calor
Value of protective inter-
vel (level of black) " � ASR, P .
Transmiseion Cheraceeriatic "
of chnnnels of primary col
ore from light to 1ighC
Nonuniformity of color over Static, � ASK, GLS,
~ the raeti~r field dynamic �VTP, P
Spectral Characterisei~s of Static ASK .
the seneor of main eignals
Primary colore of receiving " P
tube
2 Color coordin- Amplitude of the brightnees Dynamic ASK, GLS, P
ates DR, Y, eignal
DB Amplitude of the color dif-
ference signal " ASR, P
Changes of auch parametere as the primary colore of the receiver and the
apectral characterietics of the eensor of color signals cause dietortione
in color tranemieeion which are, chiefly, atatic in nature. It is possible
to conaider that deviations of theee two parameters reault in conetant errore
in the reproduction of all colora in the image (but different for differe~nC
receivers and aeneors) With the exception of the reference ahite color. The
latter will not Change iF the correct colot balance is aiaintained both in
the receiver and in the aenaor of the aignals.
It ia assutned that the color eyetem of the receiver ie realiaed in the senaor
of aignals. If it is not so, then it ie necessary to examine the deviations
of both spectral characterieticg and output signals from the rated valuee in
the coordinate ayetem of the eensor and to introduce a third grouped param-
eter.
~
Theoretical Study on Separate Effects of Changee in the Parametere of the
TsTV Channel. When breaking the overall tolerance for a grouped parameter
into individ~al parameters, it is necessary to know what dietortiona occur
in the TV image due to changea in only one parameter ahen all remaining param-
eters are unchanged. It is relatively simple to atudy this theoretically
for the parameters ahoee deviatione fraa the rated values cauae dietortions
in color tranemiseion (Table 3). Distortions of colore occurring during
19
FOR OFFICIAL USB ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
FOR n~rICLAL U5~ ONLY
various changes within prescribed limita oF aome pargmeCer of this group can
be represented in Che form of spaCiA1 areas in a color apace, And the pos-
eible changes in the chrominance can be repreaenked in Che form of varia-
~ Cion aonee in the chrominance chart [5, 6]. IC ie sufficienC to plot auch
variaCion zonea for several (reference) chraminancea.
The effecC of changes in individual parameters causing herterageneous dieCor-
tiona can also be studied theoretically. However, in this case, it ia more
expedienC Co represent the obtained computaCion data not in Che form of
changea in the chrominance, but in Che form oF distortions of various Ceat
eignals (for example, in ~he form of tolerance zones for tranaient, pulse,
and other characCarieCics).
The deCerminaCion of permieaible diatortione of the color TV image can be
done only experimentally. Since the qualiCy of the image dependa on the pre-
aence of dieCortions of all three groupa in it (achromatic, heterogeneoua
color disCorCiona, and diatortions in cnlor Cransmission), permiasible dis-
torCiona should be determined when Chey ahow Chemaelvea eimultaneoualy in
the image. However, it ia very difficult Co do this due to a large number
of possible combinatians of the values and types of various diaCortions. It
ie more realietic to find permiasible diatorCions separately for each group.
Since the diatoxCiona o:E Che three groups ahow themselvea in the TV image in
in different ways, the viewer, psychologically, perceives them separaCely.
For example, achomatic and heterogeneous color distortiona are perceived by~
the viewer as various defects in Che image, and disCortions in color Crans-
miasion as incorrect reproduction of colora which he evaluatea by the re~
production of colora familiar Co him (fur example, the color of flesh, the
color of vegetar.ion, etc).
Simultaneous manifestation of distortions of all three groups can be ~stimated .
approximately if the criteria for the evaluation of the quality of the image
in determining permiasible distortions are eomewhaC higher than thoae accepted
above for the resulting quality of the image. Inatead of the evaluations 4
and 3 for stationary and portable sete, it is poasible to take the evalua-
tiona of 4.2 and 3.5 poinrs, reapectively.
Permissible valuea of achromatic distorCions in a color image can be eatab-
liahed by the same method as for the black-and-white image, o: Co borrow them
directly from the experience of black-and-whiCe televiaion.
Permissible valuea of heterogeneous color distortions are found experimenCally
separately for each type of distorCion (for each groupe~~ error shown in
Table 2) during aimultaneoua changes in all parameters causing this distor-
tion, i.e., Which make up the grouped parameter. For example, the permisaible
value of the distortion "discontinues colored fringe" (Table 2) must be estab-
liahed when the following overall parametera of the channel deviate from the
notm simultaneously: diff.erential phase, differential gain, fluctuation noise
in the band of the chrominance signal, irregularities in the AChKh in the
20
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
,
FOR 0~'FICIAL U5E ONLY
band of the chrominance aignal, and Che miamaCch of the chaCacteriaCic of _
high-frgquency predistortion and ita correrrion, Tolerance values for the
~ppropriaCe overall parameCera of the channel are obtained directly from the
experimental data on permissible diatorCions~
Permieaible diatortions of color tranemiasion can be deCer~nined Chrough eub-
~ective evaluation of Che quality of color transmission by the viewers on
real aubjects when various diaCortions of colors are introduced inCo the
i~~nage. In order to create them, it ia not neceaeary eo change aimultaneously _
all of the overall parameCera of Che channel lisCed in Table 3. For example,
diatoreions in color transmiasions can be created by linear trana�ormation
of t}~ree color aignals [7]. During auch transformations, there will occur
changes boCh in the chrominance and in Che brightness of the image. Changes
in the overall brightnesa of the image, Changes in the overall brigh~ness
of the image can be diaregarded becauae, firatly, it ia not Che goal of tele-
vision Cranamission to reproduce Che abaolute value of Che brightneas of the
acenery and, eecondly, changea in the brightnesa durin~ real Cranamission '
progreas elowly and can be eliminaCed Co a certain degree by the viewer by
means of the control devicea which are at his diaposal.
As a result of subjective evaluation of the qualiry of color tranemission, -
permisaible variation zonea must be obtained on the chrominance chart for ~
the aelected reference chrominances and for eanh quality class. WiCh any
changea in ti~~ parameters leading to the appearance of diatortions in color
transmisaion, each of the xeference chrominancea muat not go beyond the limits
of the zone esCabliahed for it. It ia sufficient to find Che zones of permis-
sible variation for four poinCs of ~reference chrominancea, the reference
wtiite col~r and chrominances which are close to the chrominances of the pri- '
mary colors of the receiver. Such zones for two evaluations of quality were '
deCermined experin~enCally [8,9].
Determination of Tolerances for the Overall Channel Parametera. As tolerances
for the overall parameters 1-5 of Table 1, it is possible to Cake tolerances
for analogous parameters of black-and-white television. Tolerances for the
parameters 6-10 muat be refined wiCh consideraCion for the peculiarities of
color television. ~
Tolerancesfor each of the grouped parameters of Table 2 and parameters 6-10
of Table 1 are esCablished, as was mentioned above, through experimental
determination of permissible diatortion. The combined effect of all grouped ;
parameters causing heterogeneous color distortions is taken into considera-
tion by using the evaluations of 4.2 and ~.5 units as a criCerion of the image
quality during a change in ~nly one grouped parameter.
During simultaneoua independent changea in the overall parameters of'group 1
of Table 3, Che values of color signals ER, EG, and EB will fluctuate. As
a result of this, the point of the color reproduced on the screen of the
picture tube will fluctuate in the color space within the limits of a certain
apatial zone. If we assume that the changea of the three color coordinates ~
follow Che normal laws, then the surface of equal probability denaity in the ~ ~
21 ~
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
_
FOR OFFTCIAL U5E ONLY
color space limiCing the zone within which Che fluceuating poinr of tYie given
color is located with prescribed probability will be an ellinsoid. This el-
lipsoid of scaCCering can be easily calculated for preacribed tolerances for
changes in the grouped paramerer and Che probability of the pene~ration af
the point of wrong color wiChin Che e111psoid: The major axes of the elllp-
soid are parallel to Che axea of coordinates R, G, B~ Analogous computation
for ~he grouped ~~aram~ter 2 of Table 3 will also yield an ellipaoid of scat-
_ tering in the color apace, but wieh ma~or axes directed alang the axes of
coordinatea DR, Y= Dg.
The spatial acaCCering zone of Che color poinC as a resulC of aimultaneoua
independent changes of both grouped parametere can be found by convoluting
, the two determined diaCribution laws. The reaulting ell~~paoid obCained through
- convoluCion is pro~ecCed onCo a unit plane of the color spacp; which givea -
Che zone of Che poasible variaCion of chrominance~in the chrorninance chart
in the form of an ellipse of equal probability density. This operaCion is
performed for ail four reference chrominancea. The aummaCion of vartou~ errora
af�ecting the fi~3elity of color transmig4ion is shown in Figure 3.
dfv ~ . ~ afR
dDe � ~DiP dE~
~ d Fe Cnekmp. ,~a0oxme0ucmuivu r n
~N ~
amputro DR , De
2 YoCmomN Mnd~namnn ~ dD ~ GEe d Eq - X�90Nmev~rcmurra nepedavu g
3 tR cu:.v. uBrm nndHec ut e~i ~ d fG _ ycunum. ocHOdN. curNn~md . 9
-dDn-� A�~ 9cu~um. tnnno-nCedacx.c�rtN 10
~p 'lacino~ni demcn�ma �dEy * dEe-- ~
5 Yconum. DR ; DB ' dEe y OBeHO V?~7N020 11
Mam u~o� cunum. Ey OcxaBN. uBema ~~2j
~ lIER dE~ d�s
Rx
R.
iAo Ro-fRo Ee ~ ~--ER --~R* iR, ~~w
~0 ~~BO Go-f~o ifRo E~ �~--?B"' rt_~o_y'.Bo .
B_.v Ba �e~ Ee _Ea!..i~' fe, -EB . B
L(Bem o6aeKma (13 ) (14 ) BacnparaQoDunari qBem
F igure 3
Key: 1. Matrix
2. Frequency modulator
~ 3. Limiter of signal of color
subcarrier
4. Frequency detector 10. Amplifier of gamma-predistorted
5. Amplifier signal
6. Matrix: amplifter 11. Level of black color~
7. Camera spectral characteristics 12. Primary colors
8. Tranamission characreristic 13. Color of the object
9. Amplifier of fundamental 14. Reproduced color
signals
22
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~~t~ o~~~ic cnt. U5~ ONtY
1'he oneratinne dedcri.bed ahovc mnke iC pdg~Sble td find Che di~tribution of
~h~ r.eeulting chromingn~ce etirnr, i.e.~ the xon~ of po~~ible rhromtnance v~ri-
uei~n Edr th~ preecrib~d Cdlergncee for group~d par~m~Cer~. Now~ver, in ord~r
tv derermin~ Cha Colergnce, ir, ig neceee~ry Cn enlve gn inverge problem,
numely, ro bregk down ehp resulCing error ineo $roup~d ~rrore. Thte c~n be
done by tl~e trinl-nnd-Qrror meeh~d when the resulCing ~one of pa~eibl~ chro-
minance vnriatiori ig found having been given Che Calerance for Che grouppd
pnrumeCere. Tolergnc~ for ch~r~ges in each of the Chree color coordinat~~
(dignglg) Far twd group~d pgr~m~C~rg ~f Tgble 3 mu~t b~ ~~l~ceed in rueh ~
wuy ~hgt the ellipses of pn~stble chromin~nce v~ri~Cinn for gll four refer-
~nce chrominancea would noe go beyond ehe limie~ of the ~~t~bliehed zone~ of
permiesible vurfation and would be ga cloee Co Ch~m gg pogeibl~.
The eetabliehed tolerances for grouped parameCere ere broken down in an
analogouo manner into individual generel parameCere. In Chis caee, Che daCa
on the effects of changea in individual parametere on color transmission are
uaQd.
Breaking Tolerances into CenerAl ParameCera by Che Sectione of the Channel.
It ia mathematically difficult to perform this. IC is poseible to eolve an
invNree problem, namely~ Co deCetmine the resulting tolerance by the knrnm
tolerencee for invividual eecCione. The aummaCion lawe of tolerancps can
be found theoretically or on Che baeie of experimental daCa. Tolerancee for
indivLdual gectione are estgbliehed by eelecCion in euch a vay Chat the ob-
tained resulting Colerance would coiricide witih Che ov~rall taler~nce ~or the
entire channel. In ao doing, iC is necaseary to coneider th~ eeonomie f~ctor~
for example, a lese severe tolerence should be prescribed for g televieion re-
cetver, and a more severe Colerance for ASK.
The calculattoi~ of reaulting tolerancea for hererogeneous color distortione
are often done by pmpirical arithmetic or quadratic leag of sumnaCion of _
distortiona of one kind. In individual inatancea~ both the arithmetic and
quadratic aummationa are uaed("intermediate" law), i.e., the so-called 3/2
law. For differential phase distortiona, the author of [10] euggested to
divide by 1~ the reeulte of the computation of the overall tolerance with
the use of the quadraCic law.
Table 4
JI�nc.~.~ ~u a~..~,.;.,�p~�rma,.u..
Int~d~,1, ry � i. ~urrn r~
J.I~AII 4 _ ij~c\f~.__
~~'MNll~~ue~un~ .
c i~ 2~ ~/~~8,
. , .
(2) ApnQ~~~cri~vec� f2 w4 t3 ~6
Kui1
(3) Kua.~patiiyu~fi t4 t8~5 t5 tll
Key: 1. Summation law 4. Tolerances for the differential phase~ degrees,
2. Arithmetic for channel aections 6. GLS
3. Quadratic 5. ASK 7. VTp g. p
23
FOR OFFICIAL USE ONLY
,
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~Ot~ 0~~'ICIAL U~~ dNLY
Trybl~ 4 givQ~ ~n ~x~mpl~ of the r~aults df bre~ktng Che nv~r~11 toler~nc~ inCo
g diff~renCigl ph~~e (;~15 degr~~a) by tih~ ~eeeidn~ df Ch~ '~~'i'V ch~nnel for
v~rioug eummaeion 1~we~
Tgb1e 4 showe the dependence nf Che valuee nf Culer~nces for individugl ~ec-
tinne of rhe ch~nnel on the summaCion lawe~ In Chie connecCton, it io neceg-
sary to continu~ vork on ~he determineeion of aummation la~.d for varioue die-
CorCiong.
The runmation of Colergncee for pargmeterg af�ecting color tranemieeion can
be done by the quadratic law, since ie was assumed gbove ChAC Che dtstribu-
tion of errore from Che Eluctuatione nf the pargmeCera n� Table 3 ae ehe ch~n-
npl outpuC ie normal. This law can glen be used in determining the reeulCing
Colprgnce of a s~ction by the prescribed Colergncesfor individual agsemblies~
units~ gnd other elemente of Che section.
Bibliography
1. Dery~gin, N. G., gnd KueCarev, A. K. "U~eerminatinn o� Percnieaible bis-
tortinne of Chrominance in Color Televiaion," TEKHNI1tA KINO I TEI.EVIDEN-
IYA [Cinema and Television Engineer~, No 2, 1967.
2, "Technical Perfot~uance Targete for a PAL Colour Televieion Broadcaeting
Chain," THE RADIO AND ~L~CTROII2C ENGINEERING, Vol 38, No 4, October 1969.
3. Gofayzen, 0. V.~ end Yepifanov, N. I. "Evaluation of the TV Image Qual-
ity by Che Set of Parameters," T~IaIIVIKA KINO I T~LEVIDEATIYA, No 6, 1976.
4. M[QQt [International Radio ConeultaCive Commi.Ctee]. "Requirements for
the Characterietics of Televiaion Channels Ueed for International Con-
nectiona. Report 486-1, XIII 1'lenary Asaembly (G~neva, 1974."
~
5. Kustarev, A. K., and Deryugin, N. G. "Effects of Random Signal Changea
in the Color TV Tranemisaion Channel on the Fidelity of Color Reproduc-
tion," EI.EKTROSVYAZ', No 11, 1965.
6. Deryugin, N. G. "Effecta of Changea in the Frequency CharacterisCica of
Modulator8 and Diecriminators on Color Tranamiasion in the SECAM SysCem,"
ELEKTROSVYAZ', No 10, 1969.
7. Deryugin~ N. G., and Kuatarev, A. K. "Changea in Color Transmission
During Linear Traneformation of Color Signals," TEiCE~1IKA RIN~O I TELEVI-
DENIYA, No 2, 1973.
8. Deryugin, N. G., and Kuatarev, A. K. "Determination of Permissible Die-
tortions in Color Television," TBKHNIItA RINO I TELEVIDENIYA, No 2, 1967. '
9. Kuatarev, A. K., and Deryugin, N. G. "Permiasible Chrominance Vsriation
With Invariable White Reference C olor," TEKHI~IIRA RINO I TELEVIDEtdIYA,
No 10, 1969.
24
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
FOR OF~ICIAL U5L ONLY
10. Kopylov~ P. M~ "Differentiai Uietoretoaa tn Co1or Tel~avtaton," LLEK-
TR03VYAZ', No 3, 1968.
COP'YRY(~H'~: I~datel'eevo ~~gvya$~,~~ ~~~lektro~vya~','~ 19~8
10,233
C30: 5500
~
I
i
~
i
i
~
. ~
~
+
~5 i
FOR OFPICIAL USE ONLY
~
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
NOR O1~ICIAt U8~ ONLY
I'CALY
' TALKINC' COMPUT~R D~{iBtAp~b BY CSBLT OI~ TURIN
Milan L'LUROP~O in Iteli~n 24 Nov 78 pp 154-156
[Articl~ by Ciullano Ferrieri: "Th~ Con~uter Takes the Floor"j
[Text~ Turin, November--"It sounde l~ke a Germ~n voice," one perean e~y~.
Mother saye, "Not gC al~; I sny it's the voice of a south~rner." Still
oth~ra "Dieturbiag." "Inhwmn." "ACtractive buC distu~bing."
The~e are eome of Che comments gatherea ati random among a group of pereong
who were ligteaiag to, for the fireC Cime, the voice of $n electronic
calculator. It is e talking computer, reading the vrittea ~rd. The
voice ie varm, full; it no ],onger hes Che menoConou~ tone that characterized
the first talkiag machine8 attempted (eucept far some 8yllabic interruptions
and brief perioda of~'confueioa" in the vovels). But vhat ie etriking ie
that the machiae has within itself, in an inCernal etorehouse, huadreds of
"sounds," which aa auch are neutral. But it is he, the cos~uter, vtw gives
tone, meaning and var~th tu the sentence~ through a eeriea of inechanical
operations. The raw waterial in the caiculator comes from man. It ie the
machine that makes the dialog "human." Md therefore, as one hearer eaid
justifiably, it ie "disturbing."
The calculator vas developed by Turin's CSELT (The Center for Teleco~mnunica-
tions Research and Laboratorir.u, ons of t~e fon~ard-looking Italian reaearch
centera) and it was studied by a multidiecipline team headed by electro-
acoustic engineer Ciulio Nodena. Is this a firet? Is it a toy or an in-
atrument vith a future? How does the telking calculator work? What pro-
bleas Were solved in order to develop it? And vhat vill ita uaes be?
Naturally, there is nothing new in the ~rld. In ancient 5yracuse, a
talking statue varned the people in tiAes of emergency; and, after a17,
today in mas?y netions the telephone alarm adviaes us ~rith a voice (""at is
7 o'clock, sir") ~+ithout the need for an operator on the other end of the
line. But in both these ca8es there ia a"trick": Within the statue of
Ephesus a man yas hidden vho aas employed by the tyranC Dionysius to
frighten his subjects; the voice on the telephone alarm is merely recorded.
2li
FOR 0lFICIAL USE OM.Y
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
.~oK o~~tcint. vs~ orrt,Y
In 1~91 ~ Hungnrinn barc~n, von K~mp~l~n, invented ~ tglking n~chin~ Ch~t
eri~d tn r~produce Ch~ voe~l cord~, eh~ lung~ gnd tn~n'~ nrgl cgviCy with
bellowe and wiree. In 1931, eh~ "~eh~ni~~i ~p~gk~r" wgg built by Ripgz
and in 1939 th~ Voder voiC~ dpmongCr~tor wag built by Be11 I.gb~r~Cdri~~.
gut in e~ch of the~~.cgg~~ it ie gCi11 mgn who makee th~ machine emi.C the
dp~ired ~oundg ~nd "wordg"--by contrulling Ch~ gir ~topg in eh~ Von
K~mpelen ~pp~ratug with hie fing~re ~nd op~r~ting th~ deh~rg ~l~ctric~lly
thrnugh a k~ybd~rd ~nd p~da1~.
'Ch~ gc~n~ chgng~d r~dicglly wh~n in Chp po~tiwgr p~riod gcienc~ w~g able
~n uCili~e ~lectronic cglc~latnr~. To br~gk dot~m and sCore ~yll.~bic
grnup~ wg~ rplarively ~~gy gnd m~ehin~g of thi~ type multipl~~d. Bue a~
Ch~ proverb ggys, "'~he wgy nuC i~ eg~iegt." The difficulties b~gan after
th~ thr~~hold wne Crdeeed. Mgchinem rhat read 1~tterg gnd trgn~let~d them
inCn ~ounde nnt only produced a rumbling monoCone ChgC wge Cireeome and
annoying (this would be the 1eg~t of the probl~n~) but had the defeet of
noe regnlving ch~ elpmentary prnblem~ of und~rgCanding the me~egge: "If
i h~~r DIAMANTiNLEIA~ANTi," profes~or Cecc~to ~~yg ~okingly, "I ddn't
und~rgeand wh~ther the reference ig td ~n gdulee~oua woman or one who tieg
many precious etones..." These limitationg bord~red on~the "ggm~e depart-
ment" of etudiea ott t~lking m~chines gnd eome decades passed without
revolutionary developmenes in the gector.
As we eaid, Che main characteristic of the new machine is its ability to
give phrase~ cadence and "Conality," for example, to take into account the
proper degree of eu~peneion in a question ahich is to come leter. How vas
thi~ arrived at?
"The preliminary work was that of the linguisC~," explains Engiriper Moden~
(~ttd he mentione Profeseor Bertinetto, of the University of Turin), "Who
studied for ug Ch~ rule of gtructur~ of the Italian language. They estab-
lished that, from one point to the following point, a phrase with a certain
number of syllables, so many punctuation elemenCs (commas, e!c) and certain
pauses of separation between worda, necessarily involves a kind of 'tone'
and 'breath' that are characteriatic of the melodic progression. After
the linguists (and naturally also phyeicians and paychologists) finiahed
their work, it was up to us, the experta in electroacoustics and in calcu-
latora, to translate their regults into numerical values, the only onea
the computer undcrstands." .
In other words, the talking computer acang with lightening speed (in- ,
finitesimal fractione of a second are enough) the entire "nature" of a ~
sentence from beginning to end, the punctuation pointa and breaks betaeen ,
the words. The result of the calculation autometically determines that
the computer aill choose, from amoag a certain number of alternatives, the �
right tonality with Which to proceed s~rith the reading. Regarding reading ,
per se, this is done with "stored" sounds (the first part~ resolved some
time ago~ of the problem. Even here th~ Turin machine, hoarever, has in- ,
novations, in the sense that the current principle of syllabic division
was replaced by stored sounds based on groupa called "diphonics." These
number 150 and each is compoaed of a vrn+el and one or more coasonants
preceding it.) -
2?
FOR OPFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
n~~t~ini. us~ ~rrLY
'('hi~s IH c:crtr~inly nn ftj~linn f
tr~t whiCh outr~tripg thc~ cx~ell~nC pr~d~~e~~or~
(c:cir~Cully H~ud1c~J hy CSLL'I' whtch mc~cttinn~ them with r~Hpc~re) Huch ~g th~
"mech~niC~l r~pnrt~r" by Crr.c~tn ~t the Univer~iCy df Milan, dnd thp np-
par~Cug renlixpd in th~ Univer~ity d~ pgdu~ el~ceroni~~ 1~bdr~Cdry. ig
thig ~l~n n world fir~C7 "We gr~ glined dn Che world 9Cglp~~~ sgys engineer
pierino `Tonetto, C5~LT co-director. He adds, "aC the beCter end of the
~r~1~." Thi~ ig rnnsidprgble for g nation wiCh IC~ly's rpgearch deficir,
even Chough it ig nnly experimentgl. But ag Che ~nglish s~y, it ip algo
~n "und~r~t~C~m~nt": ~Tho~~ whd hgv~ g~~n Ch~ "b~~t" ~f thp Am~riean navy
gnd ~ir forc~ (ahieh ~YC giving c~reful gteention Co thee~ gtudies) tngin-
C~itt thaC Chp Turin ~.,lking machine is beCter.
Wiehin limieg. $ecguge it hg~ them, The C5~LT cnmpuCpr needg, for pxampl~,
a"prepgred" text with liCtle indicators that Coach ie (oeherwise i.t would
r~~d ~11 Che word~ flat). ImprovemenCs can be made, and thig ia already
done in cnnneCCidn wiCh the fluency df regding tdneg. IC do~g not yee
hav~ g~~~ntt~r far the gn~lygig df gny writt~n ti~xe (thig will come). And
ngCurally it ig not being produced con~merci~lly (bue the "explnitgCion" of
the prdtotype ig not g maCeer Chat cdncerns CS~LT, which is a research
cettCer).
To whgt use will it be puC7 ~motionally, Che advenrage thae serikes one
at first i~ that nf freCing the blind from th~: slavery of Braille. Those
wl~o cannoC aee cnn buy a book, a weekly magdzine, a daily newapaper, and
put it into the machine to hear its contenta.
In less dramatic tera~s, but economically and numerically much broader, the
talking calculator will be useful to us all, particularly through employcn~nC
over telephone lineg: Por bank reportg, on the weather, on daily news
(whieh can be obtained from n telephone number. it is true, but through the
u~e of recorded tnpes that are replaced after intiervals of time; instead,
the calculator would instantly include every nev+ event and would be able
to continually give up-to-date, clear replies ct the preciae inatant when
the user calls upon it).
Furthermore, the new computer will give us a verbal report of information
contained in the encyclopedia and the library and reports from data centers
where other macroscopic computers function. Today all these reports are .
handled by the "high priests of the calculators" and they can be collected
by the small number of the elite who poasess a"terminal" on whose screens
the reports appear. In the future, with the subsCitution of the new talking
cerminal for the textual terminal, a few coins will be sufficient so that
anyone, from a Kmall town and without going elsewhere, can have available
to him the most minute and up-to-date information from the great university
centers of knowledge (at least in theory; in practice, certainly, the problem
of "coded" reports aill have to be solved. These are the reports that
rightly or arongly are withheld from everyone so that Chey can be reserved
for privileged categories.)
28
FOR OPFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
~OR 0~'~ICIAL U5~ ONLY
An Additi~nnl S~rvic~
And what ~boue Ch~ ~rhicgl aepece, the perpnn3al (buC g].wgys drgm~Cic)
quegCion of th~ denger that the machine wi11 "replgce" m~n raCher than
s~rv~ him? This weg raised recpnCly ~t p~ris in an inCernaeiongl eympo~ium
of ~peci~ligCg on eh~ gub~ece "Th~ Comput~r and b~C~ Cnn~rol." We de~lt
wiCh the problem ~epar~rely, in a talk with Professor Ceccato which ~oes
beyond Che Turin c~lculeCnr.
i~ut ~H fnr as the CS~LT mnchine ie cnncerned, it mugC be s~id immediately ~
thgt ehe quegCion hgs le~g weight. This ig not g question of ehe atom
bnmb nr g~n~Cic~, which can be used �or gond or for ~vil (after all, ehaC
line ig also valid for th~ invention of the knife) but for ~n imporCant -
and limited discover. ToneCto says, "W~ plan only to aupply the naoat
useful end le~st ~~sely services that will multiply contacts between men."
It ie gignificnnC thgt the Turin machine evolved, beginning wiCh the rele-
typ~, wiCh Cechniqueg that limit the parameters of inemorization to Che in-
dispensnble (and "poor memnry" means use of smgll and therefore cheap cal-
cul~tnrs). Engineer Scagliola, anoeher source in the group, addg: "At
first information wag tr~nsmiCted orally; it mulCiplied with the advent of
printing. That the calculator today can make the page ralk cannot frighten
people: On the contrary, it can only bring Chem closer Cogether."
i
29
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4

APPROVED FOR RELEASE: 2007/02148: CIA-RDP82-44850R000100014441-4
FOR O~rICtAL US~ ONi.Y
; � - ~ ~.~w
' .r..,,v
� ~
~ . ~ ~ ~
<
i ~ ~
~ ,
~
i'1 ~ ~ ~ -
i'~ ~ 1
. ~ 1
`~`c:, ~r,..
~ .
~ .ti n � .
'
~~.x~~ ' , � ~ ~ ~
I %ir~,'~`; .w/ r
i ~ y'-, ' . . ~ a
~ ~ ,
e!
,
. 1
r ~ ~ q~
! T
~ ~ 11 . ~ - I~
~
�
L.. ~
_~1 ~
~l I .A
, . \
w~ ~ 1
. � . i
i ~~y~n
~ , ~
~ ~ ~~~y
,
, 4
�
r'':~ ~ ' l
~ ~ . ,f
i.~~ . .~~s~~~
~ ~
( { tt~' r - i
i ~ � iy �
i
+ ~
/
~ ' � ~
f ~e,i .
EirX ~ ~ ~
~ ~ ~ ~
1~ n ~ ~
~
J ~ ~
..r 'I , ~
f',t. . f ~SI~~� !
F
: 2i . . ~
~ Ji: , '
f �i . ~ ' . -
S ,'ti� , � ~~r
y. ~
. , . j ,
�4 � �
~ y~ I t r~ j ' `
~ � ~ r
~ + ' -
' :I~ ~~;e-4~�~.
_ t y} ~y# t:=..W,' ~y~
. t .t i~( ~ t �.4~ ~y ?t.~. �
? ~ I, ~ ~ S~:�Y~`~'
t ~ ' 7~�'~ ~},yt~ � i ~
~ ~~f ? . . . ~ ~ .4 -~i/.St~�w. a
COPYRIGNT: 1978 RIZZOLI EDITORE
END
6034
cso: 5500 30
FOR OFFICIAL USE ONLY
APPROVED FOR RELEASE: 2007/02/08: CIA-RDP82-00850R000100010041-4