CA2052929A1 - Printing apparatus and method - Google Patents

Abstract

Abstract of the Disclosure Printing apparatus and method particularly suited to provide a hardcopy of an image which is produced by medical imaging equipment or the like. The apparatus produces hardcopy consisting of pixels whose size can be changed by area modulation to suit image tonal content and detail while still maintaining a large number of gray levels per area modulation cell (pixel). Preferably, the same area modulation patterns representing pixel tonal values are used for printing pixels of different sizes.

Description

20~2~2~

PRiNTlNG APPARATUS AND METHOD
CROSS REFERENCE TO RELATED APPLICATIONS
This application is related to United States Patent Application Serial Nos. (Our Case Nos. 7581, 7650, 7651, 7653, and 7654) filed on the same date herewith and commonly assigned .

BACKGROUND OF THE INVENTION
1. Field of the Invention The present invention generally relates to method and apparatus for providing a copy of an image available in electronic form and, in particular, to method(s) and apparatus for providing a hardcopy of an image which has been produced by, for purposes of illustration and without limitation, medical imaging equipment such as x--ray equipment, CAT scan equipment, MR equipment, ultrasound equipment, and the like.

2. Description of the Prior Art A hardcopy has been defined, for example, in an article by D. G. Her20g entitled "Hardcopy Output of Reconstructed Imagery," J. Imaqinq Teçhnoloqv, Vol. 13, No.
5, October, 1 g87, pp. 167-178, as "an image that is visible to the human observer, that has a degree of permanence, and can be transported and handled without deterioration of the irnage.
Hardcopy normally is an image imprinted on transparencies where the image is viewed by passing light through ~he medium "': ;
:;
,, ;

2 ~ 2 9 or on opaque material where ~he ima~e is viewed by reflectin~
light off the image." Many attempts have been made by workers in the field to fabricate apparatus which can make a hardcopy of an electronically generated or stored image.
It is well known that devices for providing hardcopios typically rocoiYo ima~e information as outr)ut frorn an image data source such as, for example, a group of sensors, a computer image processing system, or storage devices hardcopy services. Although such may receive imaye dala in oill)or analo~3 or d;gital form, the general trend in the art today is to receive image data in digital form. Further, such devices typically comr)riso b~lrrors, Inornorio~s, lool<-ur) tahlos, and so forth for: ~a) electronic proe~ssir)~ ar.cl/or formaltin~ input ima~o dala ~ncl ~b) y~ lr~ rur r."~ O,. t~ lr~ n~ r~r ,, J 15 effects such as, for example, print medium nonlinearities or to ;~ compensato for, or to provide, ima~e contrast enhanccmor1~.
. Still further, such hard devices typically comprise an image generator subsystem which includes enercJy shaping rnochanisrns and supportin~ eloc~ronics to convert an ener~y source such as, for examplel a laser beam or a CRT beam into focused spots for scanning onto a medium.
There are certain important image quality , parameters which must be taker~ into account when designing a hardcopy dcvicc. A first, important irna~o clualily paramolor i~s resol~ltion. Most ima~in~ devices have the capability of recording many thousands of picture elements (pixels) across the medium. The abilily lo distin~uish individual pixols or to ., , ;

20~2~29 smooth the image between pixels is determined by the resolution specification. A second, important image quality parameter is raster and banding. Raster and banding are artifacts that usually appear in pixel by pixel recording systems.
Raster is caused by incomplete merging of scan lines and appears as a regular pattern of density modulation at the pixel spacing whereas banding is caused by nonuniformity of pixel placement on the medium and may appear as regular or random patterns of density variation in across-scan or along-scan dircctions. The appearance of bandin~ depends on the source of placement errors, and since the human visual system is very sensi~ive ~o placemerll ~rrors, ,~ coll~el~l urr(>rs on the order of 1% can be discerned. As a result, banding requirements must be carefully considered due to the cost implica~ions of providing procise pixel and scan lino placomcnt.
A third, important image quality parameter is geometric fidelity. Geometric fidelity specifications define the precision with which pixels are located on the medium and relate to how the medium will ultimately be used.
A fourth, important ima~o quality parameter is density fidelity. The density fidelity specification defines the j~ transfer function of the input digital value ~or analog voltage) to output density. This specification encompasses the transfer function of value to c~ensity and the transfcr function of any ~5 duplicating process utilized. The transfer function is depcn~lo on processing variables as well as Qn the nature of the specific medium used. The density fidelity specification can be ': :

- - - . . . . . :. . . - ~ . .
,. .... , -, , - . .. .

20~2~29 s~p~r~ d il~(O Ic)ur pLlrls: (n) Elbsolulo don~;ily ro~ atability; (b) relative-densi~y v~r~us input-si~nal transfor function; ~c) aro~
modulation v~rsus continuous tono rocordin~; and (d) d~nsity uniformity. The first of these parts, absolute density 5 repeatability, is the ability of the hardcopy lJevic~l lo consis~ul~lly produce the same density values for given input signals. The second of these parts, relative-density versus input-signal transfer function, i.e., lone ~cal~:, is r~ lod lo lho f~ct ~h~l in some applications a linear-density versus input-signal transfer 10 ~unction is ulilized while in olhers a dcliberalo dislorlion o~ e lran~ r ful~cliol~ i~ ulili~o~l ~o provido contr~t ~djustmont, compensation, or enhancement in certain parts of the density range. The shape of the relative-d~nsity versus input-signal lral~lt3r lul~clion ~ (Jju~locJ usir~l c~ ralion lool<-ull tablo~ -~
15 located in a digital input signai processing path, and these tables can be either fixed, locally adjusted via panel controls, or remotely loaded via a control interface. Further, if the shape of the relative-density ~lersus input-signal transfer function is critical, an operational scenario involving media processor 20 control, periodic transfer function measurement, and periodic , calibration look-up table updating will be required. The third of these parts, area modulation versus continuous tone recording, will to be described in more detail below. Lastly, the fourth of these parts, density uniformity, refers to the ability of a 25 hardcopy device to generate a uniform, flat field over the entire image area. ~
.. , 20~929 A continuous tone recording has an apparent continuum of ~ray scalc Icvels 5uch as ar~ obsorv~d, for oxamplo, in photo~raphs and in naturai sconos. This is contrasted with an area modulation recordiny which i~ typically comprised of geometric patterns of, fol exarnple, ptlnt~ lo~
please note that prir~tin~ with patterns of vatiable-sized dols is frequently referred to as halftone recording in the art. In l~lr~ rl~corLlil~ rir.lo~J ~lol ~ o in ;l r~u~ r ~Irr~v i~
varied to provide a range of tones perceived as a ~ray scale by the human eye.
As is well known to those of ordinary skill in the art, a continuous gray scale may be approximated in halftone recording because variations in printed dot size yield, for example, a varying percentage of light reflection from a printed image and, as a result, creat~ an illusion o~ a gray scalo.
Although halftone recording is basically binary, at first blush, one woul(~ ~xpect a i)alf~one recording ima~c to be lik~ that of a line copy.
However, halftone recording is complicated by the presence of spatial frequencies which are not contained in the original image, which spatial frequencies may result in unwanted Moiré patterns or other artifacts in the halftone recording image.
As disclosed in the prior art, in one halftone } record;ng method for achieving gray scale representations by binary devices, i.e., devices which display or print fixed size dots having no gray scale capability, each halftone cell, herein denoted as a pixei, is comprised of one or more clusters of .
.

; . - - : : ........................... .. . : - .

- , ~ . . . . i.

-205292~

individual print or display units, herein denoted as pels. The most common form of halftone pixel is an N by N square pel matrix of binary, fixed sized pels. The general concept of the method is to print or display a computed number of pels within a 5 halftone pixel to achieve an average gray scale level which approximates the averaged density value of a corresponding portion of the original image. For example, in Qne such prior art halftone recording method, pels in a pixel are clustered to imitate the formation of a single halftone pixel and, in another 10 such prior art halftone recording method, pels are dispersed in a predetermined manner. Further, in still another such prior art halftone recording method, referred to as "error d;ffusion," a decision to print or not to print a pel is made on the basis of local scanned density information from the original image as well 15 as on gray scale density errors committed by already processed neighbors in the recording. In ~ddition to the above, those of ordinary skill in the art appreciate that while halftone recording reproduces gray scale levels for a pixel in an averaged sense, there may be a loss of fine detail resolution in an image if the 20 size of the pixel is too large.
' All of the above-mentioned prior art halftone recording methods disclose the use of binary, fixed size, print or dispiay dots. In contrast to this, U.S. Patent No. 4,651,287 discloses a halftone recording method in which each picture 25 element to be printed or displayed is programmably adjusted to have one of a fixed number of ~ray scale levels. The patenl discloses a halftone recording apparatus which includes: (a) ., .
"_ ~
~i .

, ~ . . ,. . ~ ~ . . , -.

-7- 20~6~929 image data input apparatus such as, for exampl~, a CCD scanner for scanning an original image and for producing an array of image inpu~ data corresponding to gray scale levels of picture olemonts of thc ori~inal image; (b) processin~ apparatus for 5 receiviny the array of image input data and for computing an array of print values wherein each print value corresponds to one of a fixed number of gray scale ievels; and ~c) printin~
apparatus capable of printing picture elements havin~ a dot size that corresponds to one of the fixed gray scale levels.
In addition, the patent discloses that a printer which is capable of printing picture elements wherein each picture element has a dot size that corresponds to one of a fixed number of gray scale levels may include apparatus which varies the energy necessary for the production of a printed dot.
15 Further, the patent discloses that the energy necessary for the production of a printed dot is generally prescribed in the form of an electrical signal puise havlng a predetermined time duration and a predetermined voltage level. Lastly, the paten~ discloses ~hat varia~ions of tho encr~y can be affected by changin~ thc 20 following patarn~l~rs ()~ clric~l si~nal pul~ o on ~ e polllon (~l~lty oyclo); 11~ v~lt;~ vnl; or tl~o nlnclrical c~lrront flow.
U.S. Palcnl 4,661,859 ~Jisclosos ~ln a~)r)ara(~ls which produces a pixel having a variable gray scale. In 25 particular, it discloses a one-dimensional electronic halftone ~or~l~r~ J systoln which is coml~rised of ~ so-lrce of di~ital data representative of pixel gray scale, a counter to store the clkJilal . . - . ~ , . ~-.~ ,. . ;

, 2~2~9 data, and pulse producing logic responsive to the counter to activate a laser modulator in accordance with the digital data representative of each pixel. More particularly, a six bit data word is used to represent one of 64 gray scale levels for a pixel, and the pulse producing logic responds to the data word by producing a pulse of a predetermined duration or width which drives the laser for a predetermined time duration to produce a predetermined gray scale ievel for the pixel.
Notwithstanding the above prior art halftone recordin~ methods and apparatus, there still remains a need in tho art for mothod(s) and apparahls which can r)rovido a faithFul reproduction of an image rapidly, which method and apparatus include strong ~ray scale sensitivily Wil~OUl sacri~icinsJ
resolutlon and which method and apparatus are particularly suitable for providing a reproduction of an image which is generated or acquired ~rom medical imaging equipmen~ such as x-ray equipment, CAT scan equipment, MR equipment, ultrasound equipment, and the like.

SUMMARY OF THE INVENTION
Elnbodilr~orlts o~ thc prosent invcntion sa~isfy (hc above-identified need by providing method(s) and apparatus for providing a copy of an image and, in particular, for providing a hardcopy of an image which is generated or acquired from, for purposes of illustration and without limitation, medical imaging equiprnent such as x-ray equipment, CAT scan equipment, MR
equipm0nt, ultrasound equipment, or the like. In particular, .:

-2~292~

9.

embodiments of the present invention produce an area modulated hardcopy of the image, which hardcopy has a large number of gray levels per area modulation cell (pixel) and a strong density sensi~ivity, for example, a large number of ~ray 5 level steps. This is accomplished by pulse width modulating two different-sized, prin~ing radiation beams.

Specifically, in accordance with a preferred embodiment of the prt~sent invenlion, lhe prinlt~r comprisos:

means for obtaining or measuring as digital input image data 10 in~ensity levels of radiation reflected by or lransmil~ocl lhrou~h an image; means for interpolating and/or processing the digital input imago data to providc di~ital intensity levels which correspond to areas on a medium, which areas are referred to as area modulation pixels which, in turn, pixels are cornpri~ecl o~

15 subunits referred ~o as pels; mearls lor ma,u,oin!J ea(:ll ol ll~(,`

digital intensity levels into a predetermined pattern of pels;

moans ~or providirll a drivo siunal ~o a so~lreo oF las~r radialior ror nctivrllio~ o :~ourco ~- ,orint tho r)rodntnrminod pattern of pels on the mecliurn, wh~r~in ~h~ ~ourco compris~s a s~urco of are formed by pulse width modulating the source of the two different sized beams.

In a furthcr embodiment oF the prcsent invention, the printer "writes white" to enhance the accuracy cf the copy 25 at high densities where the term "write white" denotes the use of a medium wherein an unwritten medium has the highest density, i.e., all black, and a beam of radiation, for example, ..... . . ~........ .

- . . - : . . ~. .. . . , , ., . . . , .

--` ` 20~29~9 laser radiation, ca~lses portions of the black to be reduced as one provides lower densities.

DESCRIPTIQN OF THE DRAWIN(~;S
The novel features that are considered characteris~ic of the present invention are set forth with particularity herein, both as to their or~anization and method of operation, to~ether wil~ o~hor objoc~s and advantaDos th~reof, and will be hest understood from the following description of the illustrated embodiments when read in connection with the accompanying drawin~s wherein:
FIG. 1 shows, in pictorial form, a "paintbrush" of las~3r l)eal11s used lo wri~o an area modulalion pixol in an embodiment of the present invention;
FlGs. 2A-2T show, in pictorial form, pel conri~3~lr.~ion ~a~torns for vario~ls 90,um x 90~m pixel ~ray scale levels in accordance with a pr~fcrrod ombodimont of the pr~s~n~
invention;
FIG. 3 shows a block diagram of an embodiment of the present invention;
FIG. 4 shows a block diagram of a pixel generator which is fabrica~ed in accordance with the present invention;
FIG. 5 shows, in pictorial form, a comparison between an arrangement of 60,um x 60,um pixels and 90,um x 90~m pixels; and FIG. 6 shows how laser dri~e data is arran~ed for a 90,um x 90,um pixel. ;

.. - .. , ., . .,, . , , .......... , .. ,, ..... , ., ~ ........ , .. . ~
,. . : ~ - . ; ..................................... ., . . ........ ~ .

. - ,. - ~; . . . ; ; . :

-0~292~

DETAILED DESÇRIPTION
A printer fabricated in accordance wi-th the present invention produces a hardcopy of an ima0e, which image may be any one of a large number of different types of images which are well known to those of ordinary skill in the art. For example, the image may be, without limitation, a medical image produced by equipment such as x-ray equipmenl:, CT scan equipment, MR
equipment, ultrasound equipment, or the like. In the alternative, the image may be an image which is stored in, for example, digital or anaiog form, on a storage medium such as, for example, video tape, optical disk, ma~netic disk, and so forth.
A hardcopy produced by an embodiment of the pr~sent invontion is produced in a medium which is a high resolution, thermal imaging medium that forms images in respons;e to intense radiation such as, for example, laser radiation .
Suitable medium materials for preparing hardcopy images using an embodiment of th~ present invention include the thermal imaging materials disclosed and claimed in International Patent Application No. PCT/US 87/03249 of M.R.
Etzel (published June 16, 1988, as International Publication No.
W88/04237). A detaill,~d descrip~ior) of a r~edium malorial preferred from th~ standpoint of producin~ an irna~o havirlg desired durability is found in the patent appiication of K.C.
Chang, entitled, `'Thermal Imaging Medium", Altorney Docke~
No. 7620, filed of even date and assigned to the assi~nee of the present patent application.

,,.. , .. ,..... ., ~ . , ,.......... ~ . .

A preferred binary thermal imaging medium is a laminar medium incJuding a pair of sheets, at least one of which is transparent. The sheets have image forming material sandwiched between their interior surfaces and, initially, 5 preferentially adhered to one of them. When exposed to pulses of thermal radiation, the initial preferential adhesion is reversed so that, when said pair of shee~s are separated, unexposed portions of image forming material adhere to the sheet for which there is initial preferential adhesion while exposed portions 10 adhere ~o ~ sl~o~ for wl~icl~ ~hcro is tho rovorsod proforontial adhesion whereby complimentary images can be formed on respective ones of the sheets. A prelerred ilna~ lamil~al(3 medium, actuatablc in response to intense image-forming radiation for production of images in colorantlbinder material of 15 the type ~or uses with the present printer, comprises, in order:
~ 1 ) a first sheet-like web rnaterial, said web material being transparent to said image-forming radiation and having at least a surface zone or layer of polymeric material heat-activatable upon subjection of said thermal imaging medium 20 to briof and inlollso radialion;
~ 2) an optional thermoplastic intermediats layer havin~ cohesivity in excess of its adhesivily lor said surface zone or layer of heat-activatable polymeric material;
(3) a layor of porous or particulato irna~-formin~
~5 s~ll)slan(,o ol~ s.li(J tl~(lrll~o~ stic intorlnodialo l;lyor, saicl porotls or particulate image-forr~ing substance having a(Ji~esivily ~or said thcrrnoplastic intermediate layer in excoss of the adhesivity 2~2~2~

of said thermoplastic intermediate layer for said surface zone or Iayer of heat-activatable polymeric material; and (4) a second sheet-like web ma-terial coverin0 saicl layer of porous or particulate image-forming substance and 5 laminated directly or indirectly to said image-forming substanoe.
The thermal ima~in~ medium is capable of absorbing radiation at or near the interface of said surface zon~ or layer of heat-activatable polymeric material and the thermoplastic intormodi~to l~yor, at tho wavolongth of tho oxr)osinn sourcn 10 and o~ convorlin~ ubsorbod onor~y inlo ~horlr)al ()noroy o~
~u~icionl inton~ily to ho~l ~ctivoto tho r.~lrfoco zon~ or layor rapidly. The heat-activated surface zone or layer, upon rapid cc~olil1g, allacl~s Ih~ lhormopla~l;c intormucli~lo layor firlnly lo the first sheet-iike web material.
The thermal image medium is thus adapted to image formation by imagewise exposure of portions of it to radiation of sufficient intensity to attach exposed portions of the thermoplastic intermediate layer and image-forming substance firmly to the first sheet-like web material, and by romoval to ~ e second sheet-like web r~ateriai, upon separation of the first and second sheet-like web materials after ima~ewise exposuro, ol portions of the ima~e-forming substance and the thermoplastic int(~rm()(Jklll) layor, therel)y lo ~)rovil~e lirsl and ~;oco~ a~Jos, respectively, on ths first and second sheet-like web materials.
2 5 Thc optional thermoplastic intormodiatfl lay~r provklo~; s~lrf~c(~ protoction Flnd c!~lr~bilitY for the s~cond ima~e on the second sheet-like web material.

;- ., . ,. :
.,: ~ . ~ ;. . ; .;

2~2~29 Thus, two steps are required to form a harclcopy with the thermal hardcopy medium. One step comprises exposing the medium to the proper amoùnt of heat to form a latent image and the other step comprises processin~ the latent 5 copy by a peeling process whereby the second sheet carries with it the unexposed parts of the image forming substance and, in a preferred embodiment, as will be explained in further detail below, the hardccpy.
Even though the preferred medium is a laminated 10 structure, it will be clear that two unlaminated sheets with equivalent functicJns can also be used in prac~icing the invenlion.
Lasers are particularly suitable for exposin0 the medium because the medium is terrned a threshold or binary ly,ou t~ . Tll;~t i~ o s~y, it poss~ssos hi~l1 contrast ~nd, if 15 exposed beyond a certain threshold value, it will yield maximum density change, whereas no density at all is obtained below l~)is threshold .
A hardcopy produced by an embodiment of the present invention is comprised of a multiplicity of pixels. In 20 particular, in a preferred embodiment of the present invention, each pixel is about 60,um x 60~um, about 90,um x 90,um, or some variation of these sizes. Further, the hardcopy is produced by digital area modulation, also referred to as spatial dithering in lhe ~rl. Ar~3~ mo(3ul~liol) is ;~ ~oclhc~ w~ rcin ~.~cl~ pixcl is ;~
25 comprised of a predetermined number o~ pels and a particular tone, density, or gray scale level for a pixel is produced as a precletormined pattern of pels. As is well known in the art, area 2~a29~

modulation provides an illusion of a continuous tone image ;n a medium which is capable of producin0 only black and white p~ls since the area modulation tones appear to have different densities when viewed at an appropriate distance.
The following describes the criteria ~I)at are us~
determinin~ pixel size, pel si~e, and pel configuration patterns for pr~f~rr~d ~mbodimcnts of th~ pr~sflnt inventior).
Il ix w~ll l<r~own il~ arl ll~a~ ol~or;~ rn is ;~
tr~do~o~f botwoon co~y rosolution and th~ numb~r of ~fay scalc I(~V~?I'; wl~is,l~ aro no(~(lo(l 10 I)roducl) a ~III<llily cor)y ol an ima~l0.
For example, the use of an area modulation pixel comprised of n x m pels allows reproduction of nm -~ 1 dislinct gray scalo levels for a binary medium. Further, using the same pel size, an increased number of gray scale levels can be obtained by increasing the size of an area modula~ion pixel. ~low~v~3r, il ll~o size of a pixel is increased, there is a loss of resolution in the hardcopy. On the other hand, if too few gray scale levels are~
available for prin~in~, i.e., loo few sl()ps il~ e lol~t~ scalu ca occur. This is the appearance of a contour in the hardcopy that was not present in the original image and often occurs when a roproducliol-l is madc of a lar~o, smootllly varying, gray scak~.
transition .
Thus, in general, at least two measures are important in assessing the quality of a hardcopy made on a printer using a binary medium: ~1; the area modulation frequency, i.e., the number of area modulation pixels per linear inch, and ~2) the number of distinguishable gray scale levels.

: - . - , . ,. . . :; . - . .

2052~2~

The required number of distinct gray scale levels in a hardcopy depends on the ability of the unaided eye to distinguish closc~
spaced ~ray scale levels. For example, it has been -found that, at normal reading distance, the human eye can detect a refleclance 5 modulation of abou~ 0.5% at a spatial (requency near 1 cycles/mm. The inverse of this "just perceptible" modulation l~as t)een inl()rpr(~le~l us (h() muxilr)um numl)(~r ol ~ruy ~oul~

the printing industry is that a "just acceptable" picture should 10 contain about 65 aray scale levels and, for a ~ood quality copy, 100 or moro l~vols is dosired but, for modical applications, 200 or more levels are more appropriate. In addition to this, it is also known that a substantial improvement in copy quality can be achieved when pels have more than two gray scale levels.
In view of the above, the follo~,ving criteria were used in arriving at choices for the size of a pixel and a pel for preferred embodiments of the present invention: ~1 ) a pixel should be as small as is requi,ed to be invisible to the naked human eye and to produce a high quality copy; (~ for a given 20 pel size, a pixel should be as large as is required to comprise large enougl~ number of pels lo provkll) a suilaL)lo numL)ur o~
distinguishable gray scale levels and to provide a suitable mapping of density levels from the image to the copy (As will be explained below, although the ratio of the size of a pixel to the 25 size of a pel determines the number of pels which comprise a pixel and this, in turn, determines the number of ~ray scale levels which can be achieved, this ratio alone does not provide - - - . . . ; ,: - . ,. , -. ,., , ., . . -.-................ ~ . . .
- , , : ~ , " ~. .. ,:: ::

2~52~29 the capability for a on~-to-on~ mapping of density from an ima~e lo ~ c~y); un~ (3) ll~ pol puttorn ~houkl not conltih~lto to texluring or contourin0 in tho copy.
In ~dcJi~ion lo ~ho ~bovo, wo h~vn ~ volor)cd ~n additional çriterion which is derived from the fact that a perceived gray scale level of a pixel is not linearly related to-the ratio of black and white areas therein because tl~e human eye does not perceive ~ray scale level as a linear function but as a logarithmic function of intensity. One implieation of this is that the gray scale level of a pixel whose density is one density unit from the maximum pixel density is determined by the size of a pel and, as a result, the jump in density from the highes~ density on the gray scale level, i.e., Dm~,X, to the next highest density on the gray scale level, i.e., DmaX." must be small. Lastly, the choice of pixel size, pel size, and pel configuration patterns is made in light of the fact that thP. number of gray scale levels ~ -which are detectable by the human eye, i.e., the least detectable contras~, docr(!;ls~s rar)idly with spatial frenuoncy~ Thus, at lhe resolution limit of the eye, one need only represent black and white.
In accordan~e with the above-stated criteria, we have determined that a pixel size of about 60,um x 60,~Jm provides hi~h resolution copies and solves the problem of pixel visibility for a copy page of generally available sizes such as, for example, 8" x 10", 11" x 14", 14" x 17", or the like. In addition, due to considerations re~ardin~ copy speed, a "print"

2 0 ~ 2 9 2 9 pixel of about 90,um x 90~m is also wtitten in a preferred embodiment .
Initial at~empts to make coples using a "prinl" t~ixel of about 90,urn x 90,um entaited the use of three laser beams, ~ach of which provided a pcl havin~ a spot size of about 30,um x 3,um on the medium. However, as was explained above, such an arran~ement can provide only 91 linear transrnission incremen~s and thls, It wa~ ;cov~r~ l, provi~ l an inadequate number of gray scale levels for certain applications.
In fact, a far lar~er number of transmission increments are needed to provide a more suitable numbor of gray scale lovels.
A larger number of transmission increments is provided, in accordance with thc present invention, b~ pulse width modulation ol tl1o drivin~ si~nal for tho radialion boams, in this embodiment, the driving signal for the laser sources, to produce variable sized peis.
In accordance with the present invention, a pixel is ~ `
"painted" with a predetermined area modulation pattern of pels, which predetermined area modulation pattern of pels corresponds to a predetermined intensity level in the ori~inal image or to a predetermined intensity level computed by the printer. In this context, the term "painted" refers to the exposure of a pixel of heat sensitive medium to beams of laser radiation. In a preferred embodiment of the present invention, a pixel is chosen to be substantially 60,um x 60,um or -90,um x 90,um in area and a "paintbrush," i.e., the beams of laser radiation, which is used to "pain~" the pixel with pels is 20~2~29 compris(,~l o~ ~our s(:l~ar~l~3 be~ s o~ l~sor radio~iol-. As ~hown in FIG. 1, each of the first three beams oF radiation 200, 210, and 220 in "paintbrush" 250 provides a spo~ on lhe modiurr whosc smallest footprint thereon is an area which is sub~tantially oqual to 30,um x 3,um. r3~ams 200~220 ~)ro alignod in an offsot configuration so that a stroko of "paintbrush", lab~l~d as 250 cov~rs 011~ an~ on~half (1.5 ~0~JIII x ~~ I l)ix~ Ol OI~U ao,u~l~ x ao/~", pixol. Al: w.~
discussed above, the choice for the size of beams 200, 210, and 220 was determined by the criteria set forth above as well as factors such as tl~o comploxily and oxponsc roquirod to provide a smaller sized pel, the additional print time required in producin~ a hardcopy with a small~r si~l3d pol, ar)d ~I)e complnxi~y, ox~onso an~ print timu involvod in utili~in 15 additional laser beams.
As also sllowl~ In I tC~. 1, in ~ ion lo be~
200--220, "paintbrush" 250 is comprised of a fourth beam ol radiation, bearrl 230. B~3am 230 provi~les a s,ool on ll1e mo~liuln whose smallest footprint thereon is an area which is substantially equal to 5IJm x 3,um, and beam 230 is alignelJ st) thal il lr~vor~os a linl) whiel~ passos rou~l)ly lhroll~h tho conlor of beam 210.
As doscribod abovu, in ll~is prelerro(J ~mboclirr~orlt, oach of boams 200--230 has a minimum footprint width on the medium, i.e., distance from top ~o bottom of a footprint, of subs~ lly 3~m. I-lowovor, in ~ccordanct3 wi~ tho prosont invention, the footprint width is variable for each of the four 2~2~29 -~o-beams, i.e., beams 200--230. The footprint width is varied by allowinu a l~oarr~ ~o ir~pinUl) u~)ol) ll)o In~Jium lor a varlal)k) ;lrno~n~ of time as the medium passes under the bearn. The variable amount ol timo for allowin~ a boarrl ~v impin~o upon lhe 5 modi~Jm i~ providod in tho preforred embodiment by pulse width modulating each laser bcam so that the footprint width can vary from the thickness of the laser beam, i.e., approximately 3.0,um or more, to roughly 60.0,um or 90.0,um in incremonts of .375,um. This method of pulse width modulating the laser beam 10 radiation will be referred to below as slicin~.
In accordance with the present invention, slicincJ is achieved by modulating the writing fr8quency of a laser drive signal such that a laser is turnad on ror a minirnurn writin~ lir~o (t) to write, for example, 3~m or for lon~er times ~t + x~d~
15 where dt is lhe time to wrile a slice an(l x is lllo nurnbor ol dosirod slices. The use of siicincJ increases the effective number of pels in a pixel.
In a par~icular ernl)o~Jimurll Or ~hu ~ro~ invonlion~ tho cl~oico of slice size is determined by balancin~ the need to provide an 20 adequate number of ~ray scale Isvels and the complexity involve~ in providing vory small slicos. Vory small sliccs place great demands on both hardware and medium. Hardware needs to become more complex whilo mediurn musl l)e capabl~ ol generating small spots. As a result, in the preferred 25 embodiment, we have chosen a slice of about .375,um.
l-lowov~r, it shouk~ l)o clu~)r lo Ihoso of orclinary skill in the art that the particular choice of the nurnber of slices and the .~

2~2929 minimum and maximum widths for a pel is a rnatter of design choice and does no~ limit the scope o~ the present invention.
The following describes the advantageous results which are obtained from the use of laser beams which have different footprin-ts on the medium, i.e., laser beams 200--220 each have a rninimum footprint of about 30,um x 3,um and laser beam 230 has a minirnum footprint of about 5,um x 3,um. If copies were printed on the above-described medium the highest gray scale level for the above-described medium corresponds to a density value, DmaX~ approximately equal to 3Ø
Using 90,um x 90~m pixels and laser beams wi-th a minimum footprint, i.e., pel size, of about 3~m x 3,um, the next highest gray scale level in the copies would correspond to a density value DmaX.l approximately equal to 2. Another way of understanding this result is to appreciate that if one were to produce copies using pels having a minimum footprint of about 30,um x 3IJm, one would make the densi~y range between 2 and 3 inaccessible in the copies. This, of course, is unacceptable for a printer which has to produce copies of images provided by medical imaging equipment where vital information is recorded by donsity variations. Sp~cifically~ as stat~d in Neblette's Handbook of Photography and Reprography, Seventh Edition, Edited by John M. Sturge, Van Nostrand and Reinhold Company, at p. 558-559: "The most important sensitometric difference between x-ray films and films for yeneral photography is the contrast. X-ray films are designed to produce high contrast because the density differences of the subject are usually low , . : ~

2~2929 and increasin~ thsse differenc~s in th~ radio~raph adds to its diagnos~ic valuo.
Radiographs ordinarily contain densities ranging from 0.5 to over 3.0 and are most effectively examined on an 5 illuminator with adjustable light intensityUnless applied to a vury limited donsity ran~ the printin~ of radio~r~phs on photo~raphic paper is in~ff~ctive bec~use of the narrow ran~e of densities in the density scale of papers."
As a result, the printor needs to b~ ablo to vvritc a pel having a substantially smaller size than 30,um x 3,um. This capability is provided, in accordance with the present invention and as was described above with respect to the preferred embodiment, by using laser beam 230. Although, in principle, laser beàm 230 could be added to "paintbrush" 250 in any one of several ways, the placernent shown in FIG. 1 provides a preferred placement wherein laser beam 210 is replaced with Iaser beam 230 at predetermined times. In the preferred embodiment, the minimum size of the small pel is about 5,um x 3,um and, as a result, DmJX_1 is about 2.7 for a 90,um x 90,um pixel. Since the depth of focus required to provide a pel of a particular size is inversely proportional to the square of the pel size, a pel size of about 5,um x 3,um is reasonablo in torms of tho comploxity and expcnso involved in providing a smaller sized pel.
Further, as described above, slicing is also applied to pels written by the fourth and smallest laser beam and, as a result, the number of gray scale levels is dramatically increased, :' . - , ~ " ' '. : ~

~3 2~2~2~

and small increments between 0ray scale levels are realizable.
The increase in the number of ~ray scale lev~ls is most advanta~eous at high densities bscause th0 human oyc is most sensitive to transmi-ttance or reflectance chan0es which occur at 5 hi~h density. Sp~cifically, th~ human oy~ is sonsitiv~ to rolativo chang~ in lumir1ar1c~ as a ~unction of dL/L whoro dL is th~
change in luminance and L is the average luminanc~. Thus, when the clensily is ~ ., L i~ ~m~ll, lhu ~o~ ivily i~ Jh for a given dL whereas if the density is low, i.e., L is large, then 10 the sensitivity is low for a ~iven dL. In accordance with this, embodiments of the present invention preferably provida small steps between gray scale levels at the high density end of the ~ray scale. Further, in accordanc~ with this, it is also pr~f~rrod to wri~ci (he l)ilh (Jensity ,oarl ol (1~ 3ray seaiu as accuralllly ax 15 possible because the human eye is more sensitive to in-tensity dilf()rerlces wl)ich oc(,ur in Ihal parl ol lho ~ray scak~. In accordance with a preferred embodiment of the present invention, this is accomplished, as was described above, by writing "white" on the medium. As was clescribed above, in lhu 20 preferred embodiment, the medium i5 such that, in an unprinted or virgin state, the medium is black. The making of a copy entails the use of radiation from laser beams 200--230 to cause the copy formin~ substance on the medium to adhere to the surface of the web. Then, when the cover is peeled, the 25 exposed regions remain on the web and the unexposed regions remain with cover and form the hardcopy. Since the hardcopy is written by using laser beams 200--230 to denote areas on the 2~2929 ultimato copy whoroin black is r~movod, th~ formation of th~
hardcopy is referred to as a process where one "writ~s whito."
This is advantageous, as can be seen from the above, since laser bearn 230 which produces the small pel is used to provide 5 gray scale levels which corresponds to high density. The advantage is derived from the fact that the accuracy of the specification of the high density gray scale levels depsnds on the positioning of a single lasar beam, narrl~31y, last~r- L)oam 230 which is responsihle for writing the small pel. If the medium 10 were written "black" the high density gray scale levels would be written by the interaction of sev~ral, il nol all, o~ lasor boarns 200--230 and provide more opportunity for positioning error.
As a r¢sult, a prinlQr would havo to bo more complox and expl3r)sivu ~o acl~ieve a colnparal)lo lovol o~ accuracy as lhat 15 achieved by a printer that utilizes a "write white" process. This is because, as was set forth above, intensity differences are ;
more readily detected in the high density portion of the gray scale levels, and medical images typically are darker than picture photographs. Notwithstanding the above, it should be 20 understood that the present invention is not restricted to "write white" embodiments and that the present invention also encompasses "write black" embodiments.
In a preferred embodiment o~ the present invention, pel configuration patterns for "painting" a ~O,um x 90,um print 25 pixel are designed to meet several objectives which are necessary for repeatable imaging of high quality. A first objective in developing pel configuration patterns for the : ., ~

- : ~ . , ~ . ...................... ~ :
- , ; ~, . , .. . . . ~ ,. . ., ~25- 20~2~2~

preferred embodiment which "writes white" is ~o make as few chan~es in an area modulation pixel as is possible -for hi0her density gray scale levels because the most critical inforrnation in most medicai images is in the darker areas of an image. In addition, a second objective in developing pel configuration patterns is to minimize the effect of bridging in the medium on image quality. Brid~ing is a phenomenon that occurs in the nbovo-~lo~;~,rit)oll rr~o~liuln w~ novor n ~:t~vor i~ r~lntl ~n~l t,lo~ioly spac~d oxpos~d mat(3rial brid~s, i.~., pulls unoxposod mat~rial between them, flom the cover. As one can readily appreciatc, bridging will result in density variations and, hence, lower quality copies. Bridginy can be prevented by ulilizir1~ pol con~i~uralion patterns which maintain minimum distances of unexposed material in the medium between clusters of exposed material.
For example, we have determined that the probability of bridging, i.e., the probability that two clusters of exposed material will bridge, is reduced substantially if there is a minimum unexposed distance between the clusters of ~bou-t 1 0,um to 1 2,l~m.
FlGs. 2A-2T show various pel con-figuration pat~erns for various 90,um x 90,um pixel gray scale levels in accordance with a preferred embodiment of the present invention. These figures are best understood when they are viewed in conjunction with TABLE I (see page 57).
The ~rid in FlGs. 2A-2T represents one 90,L/m x 90,um pixel which is comprised of 3 columns, each column being comprised of 30 rows. Pels in the first column are ,, ., ,. . ~ ;... .... . . .

20~929 "painted" by wide laser 3; pels in the second or middle column ar~ "painted" by wide las~r 1 or by narrow laser 4; and p~ls in the third or lasl column aro "paintod" by wido lasor 2. Tho coordinates of a particular pel in the ~rid are desi~nated, as to 5 row, by a number from 0-29 and, as to column, by the laser numbar which "paints" that pel. In viewin~ FlGs. 2A-2T, keep in mind that we are viewing a "negative" of a "write white"
medium, i.e., the white ar~as in ~h~ uro~ ar~ ur~oxposolJ aro.J~
and the black areas are "painted" areas. Thus, the hardcopy will 10 be the reverse of the figures. For example~ FIG. 2A shows a "negative" of a completely unexposed medium and, as a result, represonts a pixel havin~ th~ darkest ~ray scale level.
In providing pel configuration patterns for the preferred embodiment in accordance with the above-described 15 criteria, we have divided the pel configuration patterns into groups A through J an~ we have specified certain "painting"
rules for the various groups. The rules are displayecl in TABLE I
and are illustrated in FlGs. 2B-2T. In particular, pairs of ~i~ures from FlGs. 2~-2T show tho startin~ p~l configuration pattorn for 20 a group and the last pel configuration pattern in a group, respectively. Specilically, Wilh r~crenc~ lo TABLE l~ e column headed "GROUP" refers to pel configuration patterns in the various ~roups A-J the column headed "BEGINNING OF
CLUSTER LOCATION" gives ~rid coordinates in ~erms of row 25 and laser for pels in the first pel configuration pattern in a group;
and the column headed "CLUSTER SIZE RANGE IN SLICES"
~ives the minimum and maximum number of slices for each of - . . ~: ~ ~ - . ............. . .
- ~

-2~ 2 ~ 2 9 the lasers used to produce pel configuration patterns in a group.
TABLE I and FIG. 2B show that the first pel configuration pattetn in group A comprises 6 slices "painted" by laser 4 startin~ in row 5. Further, TABLE I and FIG. 2C show that the last pel 5 confi~uration pattern in ~roup A comprises 200 slices "painted"
by laser 4 startin~ in row 5. TABLE I and FIG. 2D show thal lhc first pcl confi~uration patt~rn in aroup B compriscs 110 slir.,es "pnint~l" I.y Inn~r ~1 r.l.~tr~inU in row 5 and 1 ? r.licn~ "r)aint~3d" by laser 3 starting in row 0. Further, TA~LE I and FIG. 2E show 10 that the last pel configuration pattsrn in ~roup B comprises 20() slices "painted" by laser 4 startin0 in row 5~and 12 slices "painted" by laser 3 starting in row 0. TABLE I and FIG. 2F
show that the first pel confi~uration pattern in group C
comprises 1 10 slices "painted" by laser 4 starting in row ~, 12 15 slices "painted" by laser 3 starting in row 0, and 12 siices "painted" by laser 2 starting in row 15. Further, TABLE I and FIG. 2G show that the last pel confi~uration pattern in group C
comprises 200 slices "painted" by laser 4 starting in row 5, 12 slices "painted" by laser 3 starting in row 0, and 12 slices 20 "painted" by laser 2 starting in row 15. The remaining ones of FlGs. 2B-2T can be similarly undorstood with rof~ronco to l-ABLE 1.
Note that ~roups F-J which correspond ~o lower densitics do not use small laser 4. However, this is not a 25 drawback since, as was descrihed above, the logarithmic human visual response rneans tl)al lar~ur lransmission or rofleclion ~28- 2 ~ 2 ~

differences in regions of low density can still be nearly invisible to the human eye.
As one can readily appreciate from the above, FlGs, 2A-2T and TABLE I provide more pel configuration patterns than 5 would bQ usod to provido, for oxamplo, 256 ~ray scale levels.
Thus, in ptacticc, an appropriat~ subs~ of th~ various p~l confi~uralion pallorns provir~od in FiGYi. 2~-2T ancl TABLE I ~or use in a specific case depends on the particular requir~merl~s o~
the specific case and, an appropriate subset therefor, is selected 10 to approximate the specific tone scale desired. However, one may consider the following methodology for choosing pel configuration pattorns from amon~ the various possibilities in a group. First, consider the first pel configuration pattern for a ~roup and, for each laser, determine the amount of area that can 15 be "painted" to reach the last pel confi~uration pattern for the group. Second, pel configuration patterns from that group, other than the first pel configuration pattern, are first selected as bein~ those which are obtained by "painting" with the laser that has th~ lar~est arca that can be "painted." Howover, as the 20 selnet~d lasr~r "paints" to provide selected pol confiç1uration pattotn~, tho amount of aroa that can bo "pain~od" for that lass)r is docro~sod. Third, wh~n tho amount of aroa that can b~
"painted" by the first laser equals the arnount of area tha~ can be "painted" by another laser, pel con~i~uralion patterns are 25 then chosen which alternately "paint" these two lasers.
The laser source which is uscd to provide a beam to write the small pel may be similar to ~hose used to provide a . . . . .;; . ~ . - ~ .. . ~. . .

--i. . .. ..
-, ,: ; . ; ;. ~ ~:

2~32~29 -2~-beam to write the large pels, but with its radiation output clipped using mirrors of appropriate dimensions. Alternativel~/, one could utilize a laser having a smaller emitting region.
FIG. 3 shows a block diagram of inventive printer 10 which produces a hardcopy of irnage 50 on medium 205. As shown in FIG. 3, printer 10 comprises: ~a) Image Scan and Acquisition Module 100 which acquires image data in el~ctronic form corresponding to image 50; (b) Image Frame Slor~ 110 which stores the image data provided by Image Scan and Acquisition Module 100; (c) System Cor)~roiler 115 whicl): (i) processes the image data stored in Image Frame Store 1 10 in a manner whiGh will be described in detail below, (ii) causes the processed image data, an~ other inforrnation that will be described in detail below, to be transferred to other portions of printer 10, and, in certain embodiments, (iii) receives inpul information from a user to provide printing format information an(~ Ih~ lik~: ((J) Pix~)l G~norator 700 which roc~ivos ima~o dnla lloln ~ U~J rr;llnl1 Slol-~ 1'10 ;~n(l ~,onlr~ [~ linl~ froll-System Controller 115 and, in response thereto, produc~s OUl~Ul ~o La~r Mo~lul~ 0; al)~ ) Lu~u~ M~-lul~ 750 wl~h:l-comprises Lasers 195, which lasers produce a hardcopy of image 50 on medium 205 in response to the output from Pixel Generator 700.
Image Scan and Acquisition Module 100 is apparatus which is well-known to those of ordinary skill in the art for scanning image 50, for acquiring image data from image 50 in analog or digital form, and for converting the acquired 20~2929 image data into digital form; if necessary. Embodiments of Image Scan and Ac~uisition Module 100 are well-known to those of ordinary skill in the art and comprise, for example, apparatus: (a) for scanning image 50 with radiation output from, for example, a CRT; (b) for measuring the amount of radiation which is reflected from ima~e 50 and/or which is transmitted by im;l~l) 50 wi~ olo(loloclors il~ a Inallnor wl~ Iso woll-known to those of ordinary skill in the art; and lc) for cor-ver(in~, ~or exampl~, outpul Irom ll-t~ piIolo~ loclors lo digital image data by sending the output through, for example, analog-to-digital converters in a manner which is also well-known to those of ordinary skill in th~ art. Altornativnly, Imago Scan and Acquisition Module 100 may be a CCD scanner. In the embodiment described below, for purposes of illustration 1~ only and wilhout limitation, it is assurnod that tho di~ital ima~o data output from Image Scan ancl Acquisition Module 100 comprises eight (8) bit data, each of which corresponds ~o a 256 step gray scale. Further, also for purposes of illustration only and without limitation, each eight-bit image datum corresponds to the intensity of the radiation which was reflected from a predetermined area of image 50 or which was transmitted by a predetermined area of image 50. In addition, it should be clear to those of ordinary skill in the art that image data which is output from Image Scan and Acquisition Module 100 and which is applied as inp~t to Image Frame Store 110 under the control of System Controller 115 could just as well have been read from a storage medium such as, for example, a video tape, an optical disk, a magnelic cJisk, ar)(J so lorll~ ~n~J, in such an embodimen-t, the output from the stora~e device woulcJ
be applied as input to Image Frame Store 110. Alternatively, the digital image data could also be generated at a remote 5 location and transferred to Ima~e Frame Store 110 over a Local Area Network (LAN) or tnrough a small computer system interface (SCSI), and so forth. It should be understood that the image does not have be stored in any one particular digital or analog format, and it is well within the spirit of the present 10 invention to accept image information in any type of format.
It should be understood that each image datum output from Image Scan and Acquisition Module 100 can be displayed on an area which could be larger than, equal to, or less than the ~ize of a pixel. For example, the particular choice 15 may ~e made on the basis of format versus content the term "format" referring to, for example, the aspect ratio of the copy and the t~rm "content" referrin~ to the resolution and tone of the copy. As shown in FIG. 3, in certain embodiments, such choices may be entered ~y us~r input to Sys~em Controller 115.
20 However, in the embodiment described below, for purposes of illustration only and without limitation, an area correspor1dinl lo ~In im;l~Io tlal~lm (111 imnSI(~ 50 i~; or(linarily l;lrrl(?r Ih.1rl a. pixol an~l thus of lower spatial resolution. As a r~sult, th~r~ arr3 moro pixels in ~he hardcopy produced t~y invel~live ,uril~ler 1~ n 25 there are areas in image 50. Further, for purposes of illustration only and without limitation, medium 205 is affixed to the outer surfaco of a drum ~not shQwn), which drum, as is w~ known to : . , . . , ~ . . . ~ . .

-32- 2052~29 those of ordinary skill in the art, is cylindrical in shape. In a typical such implcrn~n~ation, as is also w~ known to those of ordinary skill in tho art, as tho drum and the m~dium affixed thereto rotate, radiation output from Lasers 195 in Las~r Modulo 750 impinges upon medium 205 alon~ a lin~. Stili further, a sufficient number of lines are formed on m~diurn 205 ~o provi~
~he hardcopy ol ima~l3 50 on m~dlum 205 as ra~lialion oulpul from Lasers 195 of Laser Module 750 is moved in a direction which is transverse to ~he direction of a line. `~et still fur~her, a page of hardcopy output may comprise several images which are reproduced on, for example, an 8 x 10 inch hardcopy and the pixel size, pixel aspect ratio, number of active lines per page in, for example, the 8--inch direction, and the number of active pixels per page in the 10--inch direction are programmably variable and embodiments of the present inv~ntion are not limited to any one particular set of such parameters.
Ima~ Framo S~or~ 110 is any appatat~s which i~
woll-kJ~own t~ II)oso of or~in~lry ~kill in ~ho art which will sorvo as temporary storage for image data obtained from image 50 or Irom a mulliplicily ol such iloa~o~. Sys~ l Col~lrvllor 115 composes and formats a "page" which "pa~e" is to be produced as a hardcopy image on medium 205 in Image Fram~ S~ore 110 in a manrler wl~icl~ i~ well-kllowl~ lo ll)oso ol ordinary skill in the art. As a result, a "page" may be comprised of a single image like image 50 or it may be comprised of a multiplicity o~ ima~os liko ima~o 50.

.-. -: . . . ~ . . : .-, - . . , ;.. ~ . .~ : . . . , . -2~2~29 System Controller then transfers the following to pixel ~enerator 7V0 preferably over a VME Bus 695 as setup data which is used by Pixel ~ienerator 700 in performin~ its function: (a) values for sertain programmable parameters of Pixel 5 Generator 700--such as, for example: (i) number of lines per page; (ii) number of pixels per line in the direction of rotation of the drum; (iii) number of pels per pixel in the direction of the rolu~ (.'?~ U Ir~ iv) ~iX~'?I ~ ?Cl r~ ; Jll~l ?~ O lor~
look-up table data which is used to generate signals for driving 10 Laser Module 750 in a manner which will be doscrib~d in dolail below; and (c) software for use by a digital signal processor (DSP) which comprises a portion of Pixel Generator 700. It should be clear to those of ordinary skill in the art that, in some embodimen~s, such data and software can be transferred prior 15 to making each hardcopy image whereas, in other embodiments, portions of such data and software may be transferred whenever the relevant data and software or portions thereof nood ~o chan~o for various portions of tho hardcopy.
As shown in FIG. 4, Pixel Generator 700 is 20 comprised of the following components: (a) VME Interface 119 VME Interface 119 receives input over VME Bus 695 and provides an interface between the internal circuitry of Pixel Generator 700 and VME Bus 695; (b) DSP 120--DSP 120 receives parameter data, software, and image data frorn System 25 Controller 1 15 (this data and information is sent from System Controller 1 15 to VME Ir:terface 119 over VME Bus 6?95 and is relayed by VME Interface 1 19 to DSP 120); (c) DSP Memory ": .- ? ; ." . .

%~29~9 121 DSP l\/lemory 121: (i) receives parameter data and softwaro from Systom Controllor 115 (this data and inforrnation is sent from System Controller 1 15 to VME Interface 119 over VME Bus 695, is relayed by VME Interface 1 19 to DSP 120, and is finally relayed to DSP 121 by DSP 120) and (ii) transfers parameter data and software to DSP 120; (d) INX Memory 130 INX Memory 130: ~i) receives imaye data from System Controller 115 (this image data is sent from System Controller 1 15 to VME Interface 119 ov~r VME Bus 695, is rolaycd by VME Interface 1 19 to DSP 120, and is finally relayed to INX
125 by DSP 120) and (ii) transfers image data to DSP 120 in response to commands from DSP 120; (e) Out Buffer 140 Out Buffer 140: (i) receives image data from DSP 120; (ii) receives addressing information from Pixel Size 163; and (iii) transfers image data to LUT Processor 170; (f) Pixel Size 163 Pixel Size 163: (i) receives parameter data (such as, for example, number of lines per page, the number of pixels per line in the direction of drum rotation, and the number of pels p¢r pixel in the direction of drum rota(ion) ~rom Sysl~)m Cor1trollcr 1 15 (this data is sent from System Controller 1 15 to VME
Interface 119 over VME Bus 695 and is relayed by VME
Interface 1 19 to Pixel Size 163); and (ii) transfers pixel address information to Out Buffer 140 and pel address information to LUT Processor 170; (g) LUT Processor 170 which is cornprised of look-up table memories LUT0 and LUT1 ~it should be clear to those of ordinary skill in the art that LUT Processor 170 is not restricted to two memories and can be comprised of oniy one .~.
: . ~ ~ ~ , ,;, .. . ;.. . .. .

35 2~52~29 mernory or even more than two memories), each of which memories contain look-up tables which provide a mapping of intensity level to pel configuration pattern for use in digital area modulation printing on medium 205 -LUT Process~r 170~
receives mapping data from System Controller 115 (this data is sent from System Controller 115 to VME Interface 1 19 over VME Bus 695 and is relayed by VME Interface 119 to LUT
Processor 170); (ii) intensity level input from Out Buffer 140;
and (iii) pel address information from Pixel Size 163; (h) Multiploxer and Delay 180 - Multiplexer and Delay 180: (i) receives input from LUT Processor 170 which contains laser drive information in 16 bit words, which 16 bit words are comprised of four 4-bit values for each of the four lasers which comprise Lasers 195 and lii) receives input from DSP 120 which contains information which is used to determine how to convert the mapping information in the two 16 bit words from LUT0 and LUT1 of LUT Processor 170 to 16 bits oF informa~ion appropriate for specific ones of Lasers 195; ti) Slice 190 ~
Slice 190 (i) receives ir,put from PLL 185; ~ii) receives 16 bit -input from Multiplexer and Delay 180; and (iii) transforms the 16 bit input signals ir,to signals for use in driving the lasers of Lasers 195; (j) PLL 185 PLL 185 is a phase-locked loop clock which: (i) receives input from Drum Encoder 187 and (ii) outputs a clock which is synchronized to the rotating drum; and (k~ Drum Encoder 187 which receives a signal when the drum rotation reaches a predetermined position.

`.

2~52929 The following describes the operation of Pixel Generator 700 in more detail. System Controller 115 obtains data which corresponds to a portion of an image which has formatted and stored in Image Frame Store 110. System 5 Controller 115 transfers the eight bit data corresponding to the portion to Pixel Generator 700 over VME Bus 695 in real time.
The term "real time" means that, for example, data corresponding to the portion such as one or two lines of the formatted image in Image Frame Store 110 are transferred 10 to and processed by Pixel Generator 700 per drum revolution.
Specifically, for an 8 x 10 inch copy printed using 60~ x 60,u pixels, the maximum number of eight bit pixels which are transf~rr~d per line in the preferred embodiment is 4096.
The eight bit pixel data which is transferred trom System Controller 115 to Pixel Generator 700 is transferred over VME bus 695, through VME Interface 119, and is applied as input to digital signal processor 120 (DSP 120). DSP 120 then transfers the data, in turn, to INX Memory 125. INX Memory 125 is apparatus which is well-known to those of ordinaty skill 20 in the art for storing digitized image data. For examplc, INX
Memory 125 may be a random access memory. INX Memory 125 is used as input buffer memory to store image data which is waiting to be processed by DSP 12(). INX Memory 125 may hold several lines of image data but typically it does not hold an 25 entire"page."
In due course, DSP 120 obtains image data from INX Memnry 125, processes it, and stores the processed data in , ~..
- - . ~ , .. ;:,, ... .,,: ~. - . ~ ......... ., , ~.

; . - ~

2~2~29 Out Buffer 140. Embodiments of DSP 120 are well-known to those of ordinary skill in the art. For example, in a preferred embodiment of the present invention, DSP 120 is a Motorola 56001 digital signal processor. DSP 1 2û accesses DSP Program 5 Mernory 121, for example, ~ RAM memory device, to obtain software which guides l)SP 120 in converting the input digitized image data into a ~orm which is compatible with the output format required for making ~he ~lardcopy, i.o., to convort lho "area-sized" image data into "pixel-sized" "print" data, and/or to 10 enhance the quality of the hardcopy by the process oi "sharpening." For example, for purposes of illustration and without limitation, in one embodiment of the present invention, DSP 120 performs a two-dimensiional interpolation on the digital image data by using two one--dimensional interpolation steps.
1 5 Specifically, DSP 1 20 performs: (a) a one--dimensional interpolation step to provide digitized image data for an "interpolated" line on image 50 which is disposed between two actual lines acquired by Image Scan and Aco,uisition Moduie 100 and (b) a second one--dimensional interpolation step on each of 20 the scan lines, actual or in~erpola~od, lo produco di~ ed imag~
data for "intorpolatod" data points which arc disposed betwecn the input data points. In particular, such interpolation steps may comprise, but are not limited to the following interpolation steps which are well-known to those of ordinary skill in the art:
25 nearest neighbor interpolation; bilinear interpolation; cubic convolution; and so forth. Furtl)er, as was mentioned above, the digitized image data, includin~ any interpolated digitized ,, image data, may be sharpened in a manner which is known to those of ordinary skill in the art. Still further, specific embodiments of ths present invention can apply different mothods of interpolation to different parts o~ ;ma~e ~0. Yet still further, as was indicatsd above, the software which is stored in DSP Pro~3r.3ln Mul~ory 121 w~s lr~nsforrod ll-~oroto from Syslem Controller 115. It should bf3 nol~l tha~ ) in som~
~rnbOLIilr)U~ lW~lf~ ly l.)l~ I(.)~I(JU~ >ri~)r ~ ri~
page to provide for the use of different imaging algorithms for different images; ~b) in other embodiments the software may be loadod prior to printin~ different portions of an ima~3e; or ~c~ in still other embodim~nts th~ softwaro is loadocl onc~, at th~ timc the system is powered up.
The output of the image processing provided by 15 DSP 120, for this embodiment, comprises eight bit numbers which correspond to ~ray levels of the processed pixals.
However, it should be understood that the present invention is not limited to the use of eight bit intensity levels. I he imasJt~
procossin~ out~ t is stored in Out Buffer 140. The embodiment ~scribod h~r~in which ontails transforrin~ ima~ data t~ PjXQI
Gonoralor 700, slorin~ il in INX Mornt)ry 125, and porformil-~
image processing upon the imase data in real time is advanta~eous because it reduces mernory cos~s lor the invenlive printer. I
In the preferred embodiment, while the drum rotates througl1 one revolution, image data necessary to crcate two output lines on medium 205 is input to Pixel Generator 700, - ............ . ~ - ~ . . ., ~:

. - - . ................... ,. .
- , . . . : .
.-, . ,.

2~29~9 where an output line is defir,ed to extend in the direction of rota~ion. Vurin~ lhe noxl rt3volulloll ol ~ rUIII/ ~WO ~ r~
are transferred while the two lines th~t w~r~ lr~n~(3rr~ urir the previous revolution are irna~e processecl arl(J oulpul lo Out Buffer 140. On the third revolution, two more lines are inpul, the lines on the previous revolution ar~ processo(J and s~or~
and the lines that were processed during the second revolution are output to be printed on the rotating drum. This continues until the entire "pa~e" has been printed. Howovor, somo ima~os 10 do not require two lines on every rotation for every outpu~ line.
In the case of in~erpolations transferring of input lines to the Pixel Generator 700 rnay be less frequent.
As described above, image processing, DSP 120 transfers eight bit digitized output ima~e ~Jala lo Oul ~uller l~lC) l5 ror ~lor~o. Oul B~lrror 1~10 is ~ppar~tus wllich is well-known to those of ordinary skill in the art for storing digitized image data.
For example, in the preferred embodiment of the present invention, Out Buffer 140 is a dual ported buffer, for example, dual ported RAM, with read/write capability through one port by 20 DSP 120 at a first rate and with read capability by LUT
Processor 170 through the second por~. This enables the data to be accessed by the remainder of the output path of Pixel Generator 700 at a rate which is commensurate with the rate at which the ima~e is to be written and the speed of rotation of the 25 drum. Further, in Ihe pr~rr~d ~mbodim~ , Oul Bu~ror 140 is configurable so that one or two lines of pixels may be oulpul from different sections thereof, and DSP 120 stores up to 41<

. - . , : . ., . , ., i ., ~ . - . . ~ , 20~2929 pixels per line therein. However, Out Buffer 140 is no~ re~uired to be a dual ported RAM and may be, for example, a FIF0.
In accordance with the present invention, LUT
processor 170 receives pixel data from Out Buffer 140 in the 5 form of pixel values and pel address information, referred to below as row addresses, from Pixel Size 163. LUT Processor 170 uses the input to retrieve pel configuration pattern information from among a multiplicity of predetermined pel configuration patterns. The pixel data from Out Buffer 140 the buffer is selected by DSP 1Z0 is transferred to LUT
Processor 170 in response to address information received from Pixel Size 1 63 .
The manner in which LUT processor 170 conver-~s a pixel datum, i.e., the digitized ou~put image data for an area 15 modulation pixel, into pel information which is derived from a multiplicity of predetermined pel configuration patterns will be explained in further detail below. However, at this point the structure of LUT Processor 170 is described in further detail.
Specifically, LUT Processor 170 is comprised of look-up table 20 memories LUT0 and LUTt. In the preferred embodiment of the present invention, eaeh memory contains the same look-up table data for use in mapping from intensity level, i.e., pixel datum, to pel configuration pattern. LUT0 and LUT1 are comprised, in a rr~anner which is wcll-known to IhOso ol ordinary skill in ~ho ar~, 25 Iroll~ n)emory sloraye devices which are wull-l<nowl) (o ~l~ose (jl ordinary skill in the art. A pel configuration pattern which corrcsponc~s lo cacll possible in~er)sily lev~:l (Jal-~m i~;

- ~ . . -., :, ............... ~,... .... ., . , --: , . : . , .................... . - ~ . ~ . -. . , .. ~, . : . . . ~ . :

20~2929 predetermined from, for example, the results of psycho-physical testing. However, the present invention is not limited to the use of one particular mappin~. Specifically, it is within the spirit of the present invention that, in some embodiments thereof, the 5 tone scale mapping between a particular intensity level and a pel configuration pattern may be varied by varying the initial configuration of printer 10 or by storing several sets of mappings and by receiving manual input from a user, as illustrated in FIG. 3, as to which of the predetermined tone scale 10 mappings is to be used for making a particuiar copy. For example, the manual input may be received by means of a user setting an indicator or depressing a button or by means of a user providing input to a user interactive system. The tone scale may be varied for use in a particular application for the purpose of, 15 for example, brightness and/or contrast adjustment.
The ~utput from LUT processor 170 is data which is used to control the behavior of Lasers 195 of Laser Module 750.
Specifically, in a preferred embodiment of the present invention, LUT Processor 170 provides 16 bit numbers which are 20 comprised of four, hex-coded bits for each of four lasers which comprise Lasers 195. For purposes of this description, and that set forth below, we designate lasers 1, 2, and 3 of Lasers 195 as being capable of providing a substantially 30,um x 3,urn pel and laser 4 of Lasers 195 as being capable of providing a 25 substantially 5,um x 3,um pel. The four, hex-coded bits are encoded so as to effectuate the slice method which has been described above, which sl;ce method divides up the time during , .,. ~ , "
- ", , ., - : ~;;

20~2929 which a laser is activated so as to be able to illuminate medium 205 in areas which comprise fractions of a pel size.
Multiplexer ~nd Delay 180 may be fabricated in a manner which should be readily understood by those of ordinary 5 skill in the art from commercially available shift registers or from programmable gate arrays. In particular, Multiplexer and Delay 180 receives the above-described 16 bit numbers output From LUT Processor 170 as well as information from DSP 120 which indicates whether a 80,um x 60,um or a 90~m x 90/um pixel is 10 being printed. This information is used, in a manner which is described in detail below, to select 4 bits per laser. The 4 bits pcr lasor aro usod lo dovolop si~nals which ar~ us~d, in turn, to develop further signals that drive lasers 1-4. The signals corresponding to lhe four bits for specific oncs of lasers 1-4 are 15 also delayed relative to each other by Multiplexer and Delay 180.
The relative delay of the various laser drive signals is understood as follows. As has been described above, the preferred embodiment of inventive printer 10 utilizes four lasers 20 in Lasers 195 to provide a "paintbrush" for printing lines of hardcopy on medium 205. In accordance with that, to prevent interference between the edges of the beams by, for example, diffraction and beam irregularities, from causing inadvertent print errors, wl~icl~ irre~ulari(ios occur l~osl ollell a~ l~oam edu~s, ~
25 laser beams which make up the "paint brush" are not physically disposed side-by-side in a line. The beam irregulari~ies resul~
from the fact that the intensity of a focused Gaussian laser .. . ; . .. .. ;: .. .... ~ ,.... :. . - ., 20~2929 beam gradually decreases from a maximum in the center o~ the beam. Thus, since focused laser beams cannot produce a uniformly intense spot, some areas of the medium may be well under or well over its exposure threshold. To avoid problems at 5 the edges, the lasers are spatially offset in the direction ot scanning. Thus, the firing of the lasers must be delayed relative to each other such that the pels generated by lasers 1, 2, 3, and 4 aro nliunod wilh oach oth(lr whon Ihny oxposn Iho mt~cJi(lm.
As such, Multiplexer and D~lay 180 adds or subtracts, as the 10 c;l~;o m.~y ho, r)rodo~otmin(ld dolays in tho firin~l timos for the lasers which genetate th~ "paintbrush" to comp~nsata fot their spatial offset. For example, in the prererrc~l olnboclim~nl, lascrs 2 and 3 are delayed 64~ relative to laser 1 and laser 4 is delayed 1 28,u relative to laser 1.
Mullipluxor and ~olay 180 transmi~s ~ho 4 bit numbers for each of the ~our lasers to Slice 19~. In the preferred embodiment, each four bit number is a four bit hex number from 7 to 15 which determines how many slices of a pel the laser is to be energized over, a pel having a rnaximum length 20 of 3,urn in the direction of rotation of the drum.
Slice 190 may be fabricated in a manner which should be readily understood by those of ordinary skill in ~he art from commercially available pro~rammal~lo array logic or rrom programmable gate arrays. In particular, Slice 190 converts the 25 input from Multiplexer and Delay 180 into four digital signals, one per laser, that are applied as input to laser drivers in Laser ;

20~292~

Module 750, which di~ital signak;, are hi0h or Inw when a l~ser is on or off, respectively.
Phase-Locked Loop 185 ~PLL 185) receives input from Drum Encoder 187 which detects rola~ion of thc drum ancJ
5 generates a signal which is input to Slice 190 so that the output from Slice 190 is synchronized to the rotatin0 drum. In lh~
preferred ernbodiment, one tick of the slice clock corresponds to .375,um at 2150 rpm or any other suitable speed.
In response to the di~ital signals output from Slice 190, laser drivers in Laser Module 750 produce high current drive signals which are applied to drive Lasers 195. In response to the drive signals, Lasers 1 as output timed beams of radiation which impinge upon medium 205 and produce therein a copy of image 50. It will, of course, be clear to those of ordinary skill in 15 the arl that further lines are printed upon rnedium 205 as the radiation output from Lasers 195 is moved across medium 20iS
in a direction transverse to the direction of a line when the optical head (not shown) in Laser Modulè 750, which holds Lasers 195, is moved in the transverse direction. An example of 20 a suitable optical head is shown for example in U.S. Patent Application (Our Case No. 7584) entitled "Printer Optical Head"
filed on the same date herewith and commonly assigned.
Further, the lasers are only driven when their beams would impinge on medium 205 and they are not driven when their 25 beams would impinge, for example, on drum clamps. In addition, it should be clear to those of ordinary skill in the art that inventive printer 10 further comprises apparatus which are 2~2929 well-known in the art but which have been omitted for ease of understanding the present invention. For example, inventive printer 10 includes, without limitation, the following types of modules: drurn drivers, synchronizin~ rneans for drum 5 positioning, laser autofocus apparatus, medium transport, and the like.
We will now describe th~? manrlor ir~ which dala stored in Out Buffer 140 is applied as input to LUT processor 170 to ~enerate laser drive si~nals. -rhe eigl~l bil proc(3~ ul 10 data in Out Buffer 140 are output as the upper address of LUT0 and LUTl. The address of the eight bit data in Out Buffer 140 is determined by a signal transferred thereto from Pixel Size 163 and is the address of the printer side of the dual ported P~AM of Out Buffer 140. This address signal is updated at a pixel rate.
15 For example, for a 60~m x 60,um pixel, the address is updated every 6Q,um, whereas, for a 60,um x 80,um pixel, the address is updated every 80~m. The lower part of the address of LUT0 and LUT1, i.e., the row address, is generated in response to an output signal from Pixel Size 163 which is applied as input to 20 LUT Processor 170. The row addr~ss countor coun~s from 0 to 29 at a pel tate and rolls over at a rate corresponding to 3,um pels.
In a partic~lar embodiment, the pixel and pel rates can be determ;ned from the following information: the length of 25 the page, for example, 10 inches; the size of the pixel, for example, 60,um x 60~m, 90,um x 90,um, and so forth; the size of the pel; and the rotation speed of the drum. For example, the ., :.
; ~

- .: . . . , : , . .......................... . .
, . . . , ~ , ~ . ., ,~ .

2~2929 pel rate is equal to tslice clock)l8 and, in an embodiment where the drum rotation speed is 2400 rpm ancl a pel is 3~m, ll)t~ ,u(31 rate is 30MI Iz/8. Further, the pixel rate is the ~pel rats)/~number of pels in a pixel). Las~ly, for a 60,um x 60~m pixel, there are 20 pels/pixel and, for a 90~m x 90,um pixel, there are 30 pels/pixel.
We now turn to describe, in detail, the manner in which data is retrieved from LUT Proeessor 170 with reference to FlGs. 5 and 6. FIG. 5 helps to show how data stored in LUT0 and LUT1 is retricv~d to supply r~l informalion which is us~d ~o drive Lasers 195 in Laser Module 750. Specifically, FIG. 5 helps to show how data is retrieved to supply pel information for a 60JJm x 60,um pixel and for a 90,um x 90/um pixel in accordance with our discovery that the mapping for a 9~m x 90,um pix81 may also be used to provide a 60,um x 60,um pixel and other pixel sizes as well.
In particular, first consider the case o~ a 90,um x 90~m pixel. As was described above with respect to the preferred embodiment of the present invention, a paintbrush for Lascrs 195, as shown abovo arrQw 2000 in FIG. 5, is comprised of laser 3, lasers 1 and 4, and laser 2. The footprint of each of lasers 1, 2, and 3 is 30,um and the footprint of laser 4 is 5um along the diroction indicated by arrow 2000. Thus, as lasers 1-4 are excited and impinge upon medium 205 alon~ the path between lines 1003 and 1004, ~hey "pain~" with a brushstroke which is 90,um across. Further, as shown in FIG. 5, the distance between arrows 2000 and 2002 are 90,um. Thus, ~i : - . i , . . : ; . ,, . , .

2~2929 there are 30 pels in the 90,um x 90,um pixel whose borders are lines 1003 and 1004, and the linus indicatcd by arrows 2000 and 2002.
The data which are stored in LUT0 and LUT1 are 5 identical and these data correspond to the 90~m x 90,um pixel just described. As a result, ~or a 90~m x 90,um pixel, one only needs to retrieve data which is stored in LUT0. FIG. 6 shows a matrix of data corresponding to a 90~m x 90,um pixel. The rows 0--29 correspond to pels for lasers 1--4 and each row, i.e., rows 0--29, contains a 16 bit number which has four bit, hex coded values for each of lasers 1--4.
In order to re~rieve this da~a, one needs to present LUT Processor 170 with two pieces of information, i.e., the intensity level of the pixel in the preferred embodiment this is an eight bit number between 0 and 255--and a pel number in this embodiment a pel number is a row address between 0 and 29 which corresponds to the pels which ar0 paintcd as thc laser beams impinge upon medium 205 between arrows 2000 and 2002. In response to this information, LUT Processor 170 retrieves a 16 bit number from LUT0 where bits 0--3 are used for laser 2; bits 4--7 are used for laser 1; bits 8--11 are used for laser 3; and bits 12--15 are used for laser 4. Of course, those of ordinary skill in the art understand that this choice of bits is arbitrary and may be changed in other embodiments. For example, this chuice of bits may be changed in software or in cabling.

.~ :...... . :......... . ...

20~2929 -4~-The inputs to LUT Processor 170 which correspond to th~ int~nsity l~v~ls of the pix~ls and th~ row ad(Jr~ss~ ol ll~o pels are obtained from Out Buffer 140 and Pixei Size 163, respectively. Pixel Size 163 has three re~islers whlch cor~lain the following information, respectively: the number of pels/pixel;
the number of pixels/line; and the number of lines/page. As such, Pixel Size 163 ~ransmits a number to Out Buffer 140 which corresponds to llle locatlon or Ih~ pi~ in a lin~ lo bu printed. Out Buffer 140 uses this number to address ~he pixels which are stored therein and which correspond to a line. Out Buffer 140 retrieves the value in its memory which corresponds to the intensity level of the pixel and applies it as input to LUT
Processor 170. At the same time, Pixel Size 163 applies the value of a row counter which cycles betwsen 0 and 29 as input to LUT Processor 170.
As one can readily appreciate, as Out Buffer 140 cycles through the pixels stored in its memory and, for each such pixel, as Pixel Size 163 cycles through 0-29, a line of data is retrieved for use in firing Lasers 195 in Laser Module 750.
We now turn to the case of a 60~m x 60,um pixel.
This case is complicaled by Iwo fac~s. Firs~, in or~or to ~al<o advantage of all four lasers, a 6~m x 60,um pixel requires the simultaneous printing of one and one-half such pixels. Second, due to the real time constraints on the system, there is not enough time available to retrieve the necessary data from a single look-up table memory.

:

.

,:` .. , `, ;,~ `. ,.. , .. : - .. ' ~ :, '' -2~929 With reference to FIG. 5, LUT Processor 170 retrieves the necessary laser drive data as follows. First, consider the region denoted by AI between lines 1003 and 1005 and arrows 2000 and 2001 to be pixel 1; the re0ion denoted by A2 between lines 1005 and 1006 and arrows 2000 and 2001 to be pixel 2; and the re~ion denotsd by A3 between lines 1006 and 1007 and arrows 2000 and ~001 to bc pixcl 3. Tho pixcls in the line of pixel 1 are painted with laser 3 and lasers 1 and 4 usin0 data obtained from LUT0; the pixels in the line of pixel 2 1Q are painted with laser 2 and laser 3 using data obtained from LUT1; and the pixels in the line of pixel 3 are painted with lasers t and 4 and lasor 2 usinç~ data obtaincd from LUT0. As ono can readily appreciate, the lines of pixels across a page, i.e~, the direction transverse to the direction in which lines are painled, obtained data to drive the lasers alternatively from LUT0 and LUT1 in a variety of sequences.
In addition to the above, since a "paintbrush"
utilizes laser 3, lasers 1 and/or 4, and laser 2, the paintbrush covers one and one-half of a 60,um x 60,um pixel simultaneously.
The ~ata to accomplish this task is retrieved as follows. ~ l ) The data for laser 3 and lasers 1 and 4 for the pixel between lines 1003 and 1005 and arrows 2000 and 2001 are obtained -from LUT0 by providin~ intensity level Al and row addresses 0--19 to LUT Processor 170. For each 16 bit number retrieved ~herefroln: bits 8--11 are for laser 3; bils 4--7 arc for lascr 1;
and bits 12--15 are for laser 4. ~2) The data for laser 2 for one-half of the pixel between lines 1005 and 10û4 and arrows 2~2929 2000 and 2001 are obtairled from LUT1 by providin~ intensity level A2 and row addresses 0--19 to LUT Processor 170. For each 16 bit number retrieved therefrom: bits 0--3 are used for laser 2. (3) The data for laser 3 and lasers 1 and 4 for the pixel between lines 1003 and 1005 and arrows 2001 and 2003 are obtained from LUT0 by providing intensity level B1 and row addresses 20--29 to LUT Processor 170 for the portion b(3lwoun ~rrow~ 2001 an~l 2002 un~J by provi(linu intonsil~/ lovol B1 and row addresses 0--9 to LUT Processor 170 for the portion betwcon arrows 2002 and 2003. For (J~cl~ 16 L)il nurrlb~r retrieved therefrom: bits 8--11 are for laser 3; bits 4--7 are for laser 1; and bits 12--15 are for laser 4. (4) l he data for las~r for one-half of the pixel between lines 1005 and 1004 and arrows 2001 and 2003 are obtzined from LUT1 by providin~
Intensily lev~l B2 an(J row ad~r~s~s 2C)~29 to LUT Procossor 170 for the portion between arrows 2001 ancl 2002 and by providin~ intonsity levol Bz and row addrosses 0--9 to LUT
Processor 170 for ~he port,on between arrows 2002 and 2003.
For each 16 bit number retrieved therefrom: bits 0--3 are used for laser 2. (5) The data for laser 3 and lasers l and 4 for the pixel between lines 1003 and 1005 and arrows 2003 and 2004 are obtained from LUT0 by providing intensity level C1 and row addresses 10--29 to LUT Processor 170. For each 16 bit number retrieved therefrom: bits 8--11 are for laser 3; bits 4--7 are for laser 1; and bits 12--15 are for laser 4. (6) The data for laser 2 for one-half of the pixel '~etween lines 1005 and 10û4 and arrows 2003 and 2004 are obtained from LUT1 by , :

' 2~929 providing intensity level C2 and row acldresses 10--29 to LU T
Processor 170. For each 16 bit numbe! retrieved therefrom: bits 0--3 are used for laser 2.
We will now describe the manner in which the laser drive data for the second half of the line of pixel 2 and the line of pixel 3 ar~ obtained. (1 ) Th~ dala for tasor 3 for ono-half of the pixel between lines 1 û04 and 1006 and arrows 200Q and 2001 are obtained from LUT1 by providin~ inlensily l~vol A2 and row addr~sses 0--19 to LUT Processor 170. For each 16 bit number retrieved therefrom: bits 8--11 are used for laser 3.
~2) The data for lasers 1 and 4 and last3r 2 lor th(3 pixel L~elww lines 1006 and 1007 and arrows 2000 and 2001 are obtained from LUT0 by providing intansity level A3and row address~s 0--19 to LUT Processor 170. For each 16 bit number retrieved therefrom: bits 4--7 are for laser 1; bits 12--15 are for laser 4;
and bits 0--3 are for laser 2. ~3) The data for laser 3 for one-half Qf the pixel between lines 1004 and 1006 and arrows 2001 and 2003 are obtained from LUTl by providing intensity level Bz and row addresses 20--29 to LUT Processor 170 for the portion between arrows 2001 and 2002 and by providing intensity level B2 and row addresses 0--~ to LUT Processor 170 for tho portion b~tw~on arrows 2002 and 2Q03. For each 16 bit number retrieved therefrom: bits 8--11 are used for laser 3.
~4) The data for lasers 1 and 4 and laser 2 For the pixel between . 1 25 lines 1006 and 1007 and arrows 2001 and 20û3 are obtained from LUT0 by providing intensity level 533 and row addresses 20--29 to LUT Processor 170 for the portion between arrows .~

. . -. :: .

20~2929 2001 and 2002 and by ptovidin~ intensity level B3 and row addresses 0--9 to LUT Processor 170 for the portion between arrows 2002 and 2003. For each 16 bit nurnber retrieved therefrom: bits 4--7 are for laser 1; bits 12--15 ar~ for l~ser 4;
$ and bits 0--3 are for laser 2. ~5) The data for laset 3 lor one-half of the pixel between lines 1004 and 1006 and arrows 2003 ~n~J 200~ ~ro obl~in~ rom LIJ 1 1 I)y provi~lin~ inl~nxily l~v~11 C~ l r~w ~ ro~lo~ D lo I.UT Prnr:n~ nr 17t). Fnr ~ach 1 B bl~ r)umb~r r~lrl~v~d thorofrom: bits 8 ~11 Dro us~d for lu4~r 3. ~1;3) Tl~ lulu for In~or~ r~cl ~1 nn~l In~or ~ for Iho l-ixol between lines 1006 an;l 1007 and arrows 2003 and 2004 are obtained from LUT0 by providing intensity level C3and row addresses 10--29 to LUT Processor 170. For each 16 bit num~er retrieved therefrom: bits 4--7 are for laser 1; bits 15 12--15 are for laser 4; and bits 0--3 are for laser 2.
As above, the inputs to LUT Processor 170 which correspond to intensity levels and row addresses are obtained from Out Buffer 140 and Pixel Size 163, respectively. However, in this case, instoad of scquoncing through 3 single line of pixel 20 intensity level data, Out ~uffer 140 sequences through two lines at the same time. As was indicated above, this enables LUT
Processor 170 to apply the intensity level from one line to LUT0 while the intensity level from the other line is being applied to '. LUT1. Specifically, as was shown above, The intensity level 25 from pixels in the line of pixel 1 are applied to LUT0 and th~
intensity level from pixels in the line of pixel 2 are applied to LUT1. Then, after the l ne of pixel 1 and the first one-half of the 2~29~

line of pixel 2 have been printed, the intensi~y level ~rom pixels in the line of pixel 2 are apptied to LUT1, and the intensity level from pixels in the line of pixel 3 are applied to LUT0 to print the second half of the line of pixel 2 and the line of pixel 3. This alternatin~ technique continues untii all of the lin~s on th~ pa0e aro print~d.
In addition to the above, it should be understoocl that l3mbodim~3nts of th~ pr~r~nt invontion also apply to situations which utilize pixel replication and magnification. For example, usin~ repeat factors for lines and/or for pixols, an ima~e may be magnified in either dlrec~lon in in~el~r incrolnonls with the smallest size being such that one pixel is mapped into a sin~le output pix~l as has been described in re~ard to th~
preferred embodiment set forth above. In addition, as a special case, shading characters are realized when the replication factors are such that each input pixel produces an inte5lral numbor of outp-lt pixels. In this case, the intensity level is represer~ted by a whole matrix and nev~r by a fraction of a malrix. Furll~er in acltJiliorl, lho aspocl ralio o~ ll)o pixols may l~e adjusted by using non-equal pixel and line replications to correct for non-square input pixels, output pixels, and/or both. Such various embodiments may be provided by appropriately pro~rammin~ DSP 120 in a manner which should b~ cloar to those ot ordinary skill ir) tl~e ar~.
It should be noted that, in the preferred embodiment, the pixel to pel configuration pattern mappin~ was a parlicular typo of mappin~. Howovor, it should b~ notod, that . , . . ~ - - ~ - . -20~2929 the present invention is not limited to the use of the mapping of the preferred embodiment. In ~eneral, ~he present invention applies to embodiments wherein the pixel to pel confi~uration pattern mapping is a whole host of different mapping functions 5 such as, for example and without limitation, area modulation imaging produced by clustered threshold arrays, dispersed dot ordered dither mapping, rectangular or hexagonal array structures, non-monotonic pel configuration patterns wherein pels that are used in a lower gray scale level do not have to be used in higher gray scale levels, and so forth.
Embodiments of the present invention which utilize such variations in pixel to pel configuration pattern mappings may be fabricated by fabricating LUT Processor 170, in a manner which should be clear to those of ordinary skill in the art, to retrieve the appropriate data from matrices which comprise such mapping data. For example, in an ernbodiment of the printer wherein DSP 120 provides pixel intensity levels that are buffered in Out Buffer 140 so as to print multiple lines in a single pass of a multiple writing elernent print head comprised of Lasers 195, the iines in Out Buffer 140 may be double buffered so that, while one group of lines is being printed, the next group of lines can be read therein.
For example, in such an embodiment, Pixel Generàtor 700 is initialized with: the number of lines; pixel intensity levels to be printed per line; and the number of pels in a pixel. Further, space is allocated for buffers in Out Buffer 140;
pointors to tho curront printin~ and loadin~ buFf()rs in Out BufFor .
'' 2~2929 140 are initialized; and corresponding flags for these two conditions ar~ sot.
Tho first step in print;nl is to load a line into -the buffer. There is 3 signal, PGactive, that indicates the position o~
5 tho rola~in0 drum. PGactivc indicatos whon lho lasors aro active, i.e., printiny a lin~, and w~en l~ r~ ar~ ovor a clul~p, i.e., the lasers are off. As the drum rotates throu~h one revolution, DSP 120 fills a buffer in Out Buffer 140, beginning when the lasers are over the clamp, with 1 or 2 lines of ei~ht bit 10 pixols, dnpcndin~ on whother a 60,um or a 9û,um pixel is bein~
printed. During the same revolution, 1 or 2 lines of eight bit pixels that where written to Out Buffer 140 from DSP 120 during the previous revolution are output from Out 8uffer 140 to LUT 170 to be printed on a page. During the next revol~ltion, 15 DSP 120 fills the buffers that were previously use~3 lot ptir~lin~
and Out Buffer 140 outputs fron) the buffers that were filled by DSP 120 during the previous revolution.
uul~ ul~ , ,oril~ J ~ l~ix(JI ruquiru~ o retrieval from a memory such as LUT Processor 170 of the pixel 20 to pel mapping. For example, the inputs for the mapping are intensity level, column pointer, and row pointer for the pel at a particular column and row of the ma~rix corresponding to the intensity level. The manner in which such mapping matrices may be stored and retrieved from storage is well known to those 25 of ordinary skill in the art.
Other embodim~nts of the invention, includin~ ~-additions, subtractions, deletions and other modifications of the - . .. , , , . . :

:

~5.~9~ ~

-56~

preferred disclosed embodiments of the invention will be obvious to those skilled in the art and are within the scope of the following claims.

2~2929 SiL~QUJ~GINNIN~ 01- Cl.~)SII-Il SIZI~ llAN
Row ~r ~ M~x, lj ~ r~ J~)o 0 3 ~2 12 0 3 12 l:~

0 1 24 9~

4 9C~

o 1 192 216

Claims (9)

1. Printing apparatus for providing drive signals to an energy source whose output is utilized in forming an image in hardcopy form as a plurality of pixels on or over a medium, said apparatus comprising:
means for receiving input image signals representing at least part or an original input to be printed;
means responsive to the receipt of the image signals for producing at least one pixel signal that represents at least one area modulation pattern within a pixel comprised of subpixels called pels, said pixel corresponding in brightness to a predetermined portion of the original image;
an energy source for modulating pixel area with patterns of pels; and drive means responsive to said pixel signal for generating at least one set of predetermined pel configuration pattern signals to activate said energy source to produce the required pel pattern within a pixel where said configuration signals are adapted to drive the energy source to generate output which provides pixels having at least two different sizes.

2. The apparatus of claim 1 wherein the drive means comprises memory means which stores data relating information derived from pixel signals to information relating to predetermined pel configuration patterns and means for generating predetermined pel configuration pattern signals in response to the information relating to predetermined pel configuration patterns.

3. The apparatus of claim 2 wherein the data stored in the memory means provides predetermined pel configurations for pixels of different sizes.

4. The apparatus of claim 3 wherein the same data is used to provide predetermined pel configurations for the pixels of different sizes.

5. The apparatus of claim 1 wherein said energy source comprises a plurality of lasers arranged to create pels in columns substantially adjacent one another in the direction of print scanning.

6. A method for providing drive signals to an energy source whose output is utilized in forming an image in hardcopy form as a plurality of pixels on or over a medium, said method comprising the steps of:
receiving input image signals representing at least part of an original image to be printed;
responsive to receiving the image signals, producing at least one pixel signal that represents at least one area modulation pattern within a pixel comprised of subpixels called pels where said pixel area corresponds in brightness to a predetermined portion of the original image; and in response to receiving said pixel signal, generating at least one set of predetermined pel configuration pattern signals and activating an energy source with them to produce the required pel pattern within a pixel by modulating pixel area with patterns of pels where the predetermined pel configuration pattern signals are adapted to drive the energy source to generate output which provides pixels having at least two different sizes.

7. The method of claim 6 further including the steps of storing in a memory data relating information derived from pixel signals to information relating to predetermined pel configuration patterns and generating predetermined pel configuration pattern signals in response to the information relating to predetermined pel configuration patterns.

8. The apparatus of claim 7 wherein the data stored in the memory provides predetermined pel configurations for pixels of different sizes.

9. The apparatus of claim 8 wherein the same data is used to provide predetermined pel configurations for the pixels of different sizes.