Philco Corporation v. Radio Corporation of America, 276 F. Supp. 24 (D. Del. 1967)

276 F. Supp. 24 (1967)

PHILCO CORPORATION, Plaintiff,
v.
RADIO CORPORATION OF AMERICA, Defendant.

Civ. A. No. 2369.

United States District Court D. Delaware.

September 1, 1967.

*25 Alexander L. Nichols, of Morris, Nichols, Arsht & Tunnell, Wilmington, Del., Thomas M. Ferrill, Jr., Allen V. Hazeltine and Roger N. Coe, Philadelphia, Pa., of counsel, for plaintiff.

Rodney M. Layton, of Richards, Layton & Finger, Wilmington, Del., Richard E. Lyon, Los Angeles, Cal., and A. Russinoff, Princeton, N. J., of counsel, for defendant.

 
OPINION

CALEB M. WRIGHT, Chief Judge.

This action is brought by the Philco Corporation (Philco) against the Radio Corporation of America (RCA) in order to secure the reversal of the Board of Patent Interferences' award of priority to RCA in Interference 87,543. 35 U.S.C.A. § 146 (1954). The interference was declared in 1955 between Boothroyd and Creamer, who assigned their invention to Philco, and RCA's assignor, Ballard. Ballard's application had been filed on September 24, 1949; Boothroyd and *26 Creamer's application had been filed on January 21, 1950. Accordingly, Boothroyd and Creamer were denominated the junior parties, faced with the burden of proving conception prior to Ballard.^[1]

In order for Philco to succeed before the Board of Patent Interferences, it had to prove that its assignors were the first to conceive the invention, and that they were diligent from a time just prior to the date of Ballard's conception until they constructively reduced their invention to practice by filing their application for letters patent on January 21, 1950. Kruger v. Resnick, 197 F.2d 348, 39 CCPA 994 (1952).^[2] At the close of testimony, the Board of Interferences determined that neither party had proven a conception date which antedated their respective application filing dates. Accordingly, the Board held that Ballard was entitled to priority.^[3]

In this § 146 action Philco must essentially prove the same thing that it conceived prior to Ballard, and that it exercised diligence from a time prior to Ballard's conception to its own filing date. Although the § 146 action is frequently referred to as a "trial de novo",^[4] which it is in the sense that the parties are free to introduce new evidence and are entitled to an independent evaluation of their respective proofs, the plaintiff in such an action bears a difficult burden of proof. Before a District Court will upset a determination of the Board of Patent Interferences it must be convinced that the Board's decision was clearly erroneous. Absent such a conviction the Board's expert conclusion

 

"must be accepted as controlling upon that question of fact in any subsequent suit between the same parties, unless the contrary is established by testimony which in character and amount carries thorough conviction." Morgan v. Daniels, 153 U.S. 120, 125, 14 S. Ct. 772, 773, 38 L. Ed. 657 (1894).

With this standard of proof in mind, the Court turns first to a description of the invention in issue, and then to a review of the conception proofs of the parties.

 
THE INVENTION

The invention is called a "dot interlaced, time division multiplexed color television *27 system." The purpose of the invention is to permit the transmission of greater detail in a given bandwidth.

At the time color television was being developed, black and white, or monochrome television had been commercially available for several years. The American consumer had made a substantial investment in monochrome television receivers. Therefor, it was eminently desirable that any proposed system of color transmission and reception be "compatible" with existing black and white facilities. In other words, that color programming be receivable on existing monochrome receivers. The existing Federal Communications Commission standards had set up bandwidth limitations for television transmitting stations. A transmitter was required to emit its programming within a 6 megacycle range; for reasons which need not be explored here, that meant that the video portion of the programming could occupy no more than a 4 megacycle band. Early color television proposals, notably the simultaneous system which transmitted the three component colors of color television at the same time could not adapt themselves to the 4 megacycle bandwidth necessary for compatibility of operation with monochrome television.^[5] The solution to this dilemma was provided by the techniques of time division multiplexing. Time division multiplexing is a means of combining several inputs into one output by creating an output which is comprised of snatches of the respective inputs arranged sequentially in time. In other words, if there are three inputs, A, B and C, a device called a "sampler" selects a small portion of A, and transmits that, followed by a small portion of B, followed by a small portion of C, and then again by a small portion of A. Hence, the output of the sampler is a series of snatches or samples of the respective inputs. To reconstruct the input waves at the receiver, the composite wave is desampled, and that portion of it which corresponds to input A is routed to the appropriate part of the receiver, and similarly for those portions corresponding to inputs B and C.^[6] This sampling is done at a rate which is rapid enough so that the viewer is unaware of the discontinuity in the respective inputs.^[7] The time division multiplexing technique enables the simultaneous transmission of a series of inputs over a bandwidth no larger than that of the largest input. Dot interlace is an improvement or modification of this time division multiplex system which enables the transmitter to convey a significantly greater amount of information in a given bandwidth. In the time division multiplex system described above, the transmitted video information is reproduced on the receiver screen as a series of minute dots. The interlace principle involves the transmission of only a portion of the total picture information in one time interval; the remainder of the picture information is transmitted in the ensuing time interval. When the remaining portion of the picture information is transmitted, the dots are reproduced on the receiver's screen between the dots of the previous portion. The human eye retains its perception of the previous grouping of dots, and integrates the second group of dots into the first.^[8]

In more detail, the system operates as follows. In the color television transmitter, a color camera "scans" the image to be reproduced. That is, the camera is moved from left to right across the image, *28 then turned off, retraced to the left-hand margin, and dropped a line; then it is turned on and scans across from left to right again.^[9] Actually, the camera is dropped two lines. It scans first line one, then line 3, then line 5 and so forth until all odd numbers of the 525 lines which comprise the raster have been scanned. Next, the camera is redirected to the top of the image, and it scans the even numbered lines.^[10] The odd numbered lines are collectively referred to as a field, as are the even numbered lines. The odd and even numbered lines together are called a frame, and for the camera to "scan" one frame requires 1/30 of a second.

The color camera is really three cameras in one red, green and blue and it emits simultaneously three trains of video information. These three strands of information are fed into a circuit called a "sampler" which selects a small piece of color information from the red input, then from the green, then from the blue, interweaving them into one composite wave form.^[11] This "sampled color information" is then led to a "gate circuit" where it is coupled with the information necessary to synchronize the receiver's desampling circuitry with the sampling circuitry of the transmitter. Before transmission the "coupled" signal is fed through a filter to restrict the bandwidth of the signal.^[12]

The information necessary to synchronize the receiver is of two sorts, that which controls the behavior of the receiver's electron guns, and that which governs the speed at which the desampling takes place. The television receiver receives or detects the transmitted signal, and breaks it into its component parts by means of highly sensitive filters. The "sync separator" breaks out the information which controls the scanning of the electron guns;^[13] the carrier wave filter culls out the sampling frequency;^[14] and, the remaining video portion is conveyed to the desampling circuitry. The desampler breaks up the composite video wave into a train of red pulses which are fed to the red electron gun, a train of blue pulses which are fed to the blue gun, and a train of green pulses which are fed to the green gun. The strength of the pulses determines the strength of the beam of electrons the respective guns will shoot toward the receiver's screen. The screen has been coated with phosphorescent material which will glow red, green and blue. By means of a "mask" the respective electron guns are aimed at those dots which correspond to the color of the information being fed into the electron guns.^[15]

When the image at the receiver is dot interlaced, the red electron gun, for example, as it scans line one of the raster will excite every other red dot on the screen. One-thirtieth of a second later, when the red beam is next scanning line one of the raster it will excite those dots which were not excited on the previous scanning of that line.^[16] In order to accomplish this it is essential that the transmitter sample a different portion of line one of the raster on successive *29 scannings. This is accomplished by reversing the phase of the wave which powers the sampler every 1/30th of a second so that on successive scannings of a given line, the sampler will be 180 degrees out of phase with the previous scanning of that line. Broadly speaking, there are two ways of accomplishing this phase reversal. First, an electronic switch may be included which reverses the carrier wave's phase every 1/30th of a second;^[17] or, second, a frequency for sampling may be chosen which is an un-even multiple of the horizontal scanning frequency.^[18] The "switch" method is the more easily explained. The carrier wave is a sine wave, and for purposes of simplicity, the sampler samples the red information at a point in time which corresponds to a particular place on that sine wave say 120 degrees.^[19] If the polarity of the carrier wave is reversed, then the 120 degree point on the carrier wave will occur at a different place along the horizontal line. And, if the receiver is synchronized, the red dot will be reproduced at a different location on the receiver screen. If the polarity is reversed every 1/30th of a second then the 120 degree point will occur midway between the points of its occurrence on the previous scanning of that particular line.^[20] As can be seen from the above description, the dot interlaced system transmits a completely interlaced color picture every four fields, or every 1/15th of a second. With this elementary explanation of the invention at issue, the Court turns to a review of Ballard's conception evidence.

 
BALLARD'S (RCA'S) CONCEPTION EVIDENCE

RCA contends, as it did before the Board of Patent Interferences, that Ballard conceived the invention in the Spring of 1947.^[21] The principal support for this position comes from two documents, Ballard's laboratory notebook, and a patent disclosure sheet, corroborated by the recollection of Schroeder and others.^[22]

Ballard's patent disclosure sheet, dated April 30, 1947, contains the following brief description of his invention:

 

"A switching [sampling] rate is chosen such that the spot of a given color is switched on at intervals corresponding to approximately twice the spot or element width. This frequency is approximately that maximum frequency which would fully modulate the spot making one picture element dark, the next light, the next dark, etc. On successive frames the switching [sampling] is of opposite phase, so that the `on' elements are meshed, thus interlacing successive pictures along the line as well as from line to line."^[23]

This description is a remarkably precise explanation of the operation of a dot interlaced system. It describes exactly how the interlace occurs on the receiver's screen. But, the description is a description of the result Ballard desired to accomplish, and not of the means for accomplishing it. A long line of cases holds that such a description will not suffice for the complete conception required by the patent laws. Land v. Dreyer, *30 155 F.2d 383, 387, 33 CCPA 1108 (1946); Townsend v. Smith, 36 F.2d 292, 295 (Cust. & Pat.App.1929).

To supplement the meager patent disclosure sheet description RCA relies on excerpted portions of Ballard's 1947 laboratory notebook. The 1947 notes are remarkably complete. Page one of those notes indicates the author's awareness of the undesirable effects which would attend continuous transmission of the carrier wave frequency. Ballard suggests as an alternative the use of a start-stop oscillator in the receiver.^[24] But then on page three of the excerpted notes, Ballard seems to suggest that either the continuous transmission or the start-stop oscillator may be used. Page 13 of the 1947 notes contains a primitive block diagram which roughs out the system. However, none of the synchronizing circuitry necessary to maintain synchronism between the receiver and the transmitter is shown.^[25] The overall contents of the notes indicates that Ballard was more concerned with the intricacies of the time division multiplexing aspects of the system, than with the interlace aspects. Several pages are given over to circuit diagrams of exotic variations of the sampling mechanism three phase oscillators and polyphase oscillators and the like.^[26] The overall impression gained from the 1947 notes is that the conception was germinating, but had not fully developed. At the very least, Ballard's 1947 notes do not demonstrate an idea which is sufficiently well-developed that all that remains to be done is to have a mechanic, skilled in the art, reduce it to practice. To paraphrase Deller: the drawings in Ballard's 1947 laboratory notebook, together with the explanatory text do not "illustrate an embodiment of the invention clearly and completely and free from ambiguity or doubt." 1 Deller, Walker on Patents § 75 (2d ed. 1964).

Also offered to supplement the 1947 notes and patent disclosure sheet were the testimonies of others in the Kell television group at RCA Bedford and Sziklai and the patent attorney who was in charge of Ballard's initial disclosure, Darke.^[27] Both Bedford and Sziklai testified before the Board of Patent Interferences that Ballard's disclosures to them in 1947 were of the complete system.^[28] But the Court is unwilling to accept *31 their testimony at face value, distorted as it must be by the passage of time, and uncorroborated by the contemporaneous notes. The systems described by Sziklai and Bedford were much more detailed than the system sketched in Ballard's 1947 notes, which creates the inference that these witnesses were unable to segregate the 1947 disclosures from the later developments in the color television art.^[29] Similarly, Darke testified that Ballard had made available to him sufficient information in the Spring and Summer of 1947 to enable the construction of the system taught by Figures 1 and 2 of the Ballard patent.^[30] But, here again, the contemporaneous documentation lends little encouragement to the accuracy of Darke's flat assertion. Further, aside from some preparatory work searching the Patent Office files, no real work was done on the Ballard application until the early Summer of 1949, almost two full years after Darke first received the initial patent disclosure sheet.^[31]

Ballard's 1948 notes, introduced by the plaintiff Philco, indicate that some progress had been made toward finalizing the conception, but the invention was still incomplete.^[32] The 1948 notes indicate that some tests had been made to determine the effects of a high frequency for switching.^[33] It had been feared that high switching or sampling frequencies might distort the video image by creating the condition known as "color cross-talk". But, in his May 1948 progress report Ballard was only able to point to the need for "[a] keying circuit is needed to tie in the high frequency oscillator with the horizontal deflection." And, the tests with regard to high frequency sampling had been found to be inconclusive.^[34]

Also important is the fact that Ballard's notes reveal the difficulty their author was having in "selling his system". Schroeder testified that Ballard was a poor salesman, and the events from 1947 to 1948 certainly validate Schroeder's opinion.^[35] As Ballard himself noted in his 1948 notes:

 

"Very little encouragement has been given this method of producing color television. Of late, the quantizing idea of doubling up information has been pushed."^[36]

*32 Evidently, RCA had not yet reached a decision as to which of the competing ideas for color television should be selected. RCA's indecision supports the Court's conclusion that the Ballard system had not been completely formulated or its circuitry completely developed as of 1948.^[37]

This is not to say that Ballard was not making progress on the conception of the results attainable with his system. His 1948 dot patterns reveal his understanding of the qualitative differences in the dot patterns resulting from the phase reversal as opposed to the magic frequency approach to dot interlace. Ballard had discussed his system with Cherry, and the latter had concluded that maximum resolution could be attained within the 4½ megacycle bandwidth restrictions imposed by the goal of compatibility with existing monochrome television.^[38] Still, Ballard had not fully developed the circuitry, notably the gating circuitry needed to employ the "burst" method of receiver synchronization.

Ballard's conception begins to jell in 1949. His 1949 notes demonstrate a complete block diagram of his system, including the "tie-in" between the sync generator and the carrier wave oscillator which he describes as the "keying oscillator and low frequency switch."^[39] His diagram dated July 22, 1949 is sufficiently akin to the diagrams of the Ballard patent for this Court to hold that by the date of the diagram Ballard had conceived his invention.^[40] The plaintiff makes much of the fact that the diagram does not describe the precise interlocking means to maintain a stable phase relationship between the carrier wave generator and the horizontal line scanning frequency, but the Court finds little merit to this element of the plaintiff's case.

First, to whatever extent Ballard's notes fail to disclose this tie-in or "interlocking" as Dr. Friend called it, the same deficiency is present in the Ballard patent itself.^[41] Since the plaintiff did not attack the sufficiency of the Ballard patent application on the ground of non-operativeness by a Rule 232 motion *33 to dissolve for inoperativeness of the applicant's disclosure, the plaintiff is estopped from attacking Ballard's conception proofs for the same deficiency.^[42] Numerous cases support this doctrine of estoppel which is bottomed upon the sound policy that questions of inoperativeness should in the first instance be tried before the Patent Office which has the expertise to lay the technical ground-work without which judicial review is likely to be uninformed. See Garand v. Pederson, 76 F.2d 407 (Cust. & Pat. App., 1935); cf., Radio Corp. of America v. Philco Corporation, 201 F. Supp. 135, 143 (E.D.Pa.1961).

Second, although the precise circuitry required to effect the tie-in was not detailed in either the Ballard notes or application, all the evidence compels the conclusion that Ballard was well-aware of the need for such a "tie-in".^[43] As Bedford testified before the Board of Patent Interferences, the means for providing the "tie-in" were well-known to those skilled in the television art.^[44] Ballard had several alternatives from which he could choose. He could have used a frequency multiplication system which is strongly suggested by the remark that the magic frequency approach requires a sampling frequency which is (n + ½) times the line frequency. Or, he could have used start-stop oscillators keyed on at the beginning of each line by a pulse of the horizontal scanning frequency.^[45] Or, yet again, he could have used crystal oscillators of a high degree of phase stability to maintain a precise phase relationship even without a direct "tie-in".^[46] The means to effect the interlock existed; they were known to those skilled in the field. Ballard appreciated that a constant phase relationship was necessary to avoid crawl. Therefore, the Court holds that Ballard's conception embodies by implication the means necessary to maintain a correct phase relationship between sampling and scanning.^[47]

*34 Third, Ballard himself recognized that his system was a modification upon, and improvement of Evans' system for time division multiplexed color television.^[48] The Ballard invention actually related to the production of dot interlace, not to the time division multiplexing of video signals.^[49] As an avowed improvement upon another's system, Ballard's system is entitled to all the refinements disclosed in that other system; he need not disclose those refinements in his disclosure. As Deller notes:

 

"[W]hen an improved machine or an improvement upon a prior machine is involved, it is only necessary for the inventor to illustrate his improvement or improved features. * * * What is common and well known to those skilled in the art is as if it were written out in the description and delineated in the drawings." Citing Loom Co. v. Higgins, 105 U.S. 580, 26 L. Ed. 1177 (1882). Deller, Walker on Patents § 75 (2d ed. 1964).

The Evans patent makes a full disclosure of "interlocking" circuitry. Evans teaches the use of pulsed oscillators in both the receiver and the transmitter, which are keyed on by pulses from the horizontal sync pulse generator.^[50] As Evans says:

 

"It is essential that the oscillations [of the carrier wave generator] always start in the same phase as the horizontal synchronizing pulse and it is desirable that the starting transient be of as short duration as possible. In addition, it should be possible to stop the oscillations quickly in order that the circuit may be ready for another timing cycle.

 

"* * * Resonant circuit oscillators and in particular, the Hartley circuit, are better adapted to this type of operation, provided certain precautions are taken to time constants involved in the biasing circuit." Evans Patent No. 2,810,781; PX 21, col. 3, at lines 59-75.

Support for the Court's conclusion that Ballard had conceived his invention by July 22, 1949 may be found in a series of tests conducted in the RCA laboratories in the Summer of 1949.^[51] While the Court is unwilling to hold that these tests constituted an actual reduction to practice, as urged by counsel for RCA, they do show that RCA was fully acquainted with the equipment necessary to render Ballard's invention operative, and hence support the conclusion that *35 Ballard's notes and diagrams of 1947 through 1949 must have fallen on fertile soil. Ballard's descriptions were generic; they failed to specify what sampling frequency should be used, and what interlace patterns would be the most pleasing. In a series of tests from July 21 through August 10, RCA sought the answers to these questions. Theoretically, since the higher the sampling frequency the greater the number of individual dots comprising the raster, maximum resolution would be obtained at the highest sampling frequency consistent with bandwidth limitations. But, there was some question in everyone's mind about the possibility of deleterious effects, notably crosstalk, should excessively high sampling frequencies be resorted to. Accordingly, on July 21, 1949 a series of tests of color picture reproduction were made utilizing a sampling oscillator with a variable frequency control. Satisfactory patterns were achieved at sampling frequencies as high as four megacycles. From July 22d to August 10th various interlace patterns were analyzed for their subjective appeal to the viewer. A unit called an interlace chassis, which had been specially constructed by the RCA Communications Laboratory at Rocky Point, enabled experimentation with several varieties of interlace. Finally, on August 10th, a dot interlaced picture was transmitted over a radio link from transmitter to receiver. As the Court has already stated, plaintiff's counsel is probably quite right in arguing that these tests were not an actual reduction to practice of the Ballard system; in any case the evidence is far too sketchy to permit the Court to hold with confidence that there was such a reduction. Such evidence as there is about the circuitry of the interlace chassis indicates that synchronism between transmitter and receiver during these tests was achieved by the start-stop oscillator method in which the oscillator was keyed on by Schroeder's "jiggled back edge".^[52] The tests demonstrate a high degree of sophistication in color television technology at RCA in the Summer of 1949, and compel the conclusion that the incidental interstices in Ballard's notes and diagrams would be rectified by skilled engineers in the ordinary course of business.

 
BOOTHROYD AND CREAMER'S (PHILCO'S) CONCEPTION EVIDENCE

Turning now to the plaintiff's conception evidence, the disparity is striking. Philco contends that its inventors conceived the invention in the Spring of *36 1949.^[53] Before the Board of Patent Interferences, in addition to the testimony of the inventors and their colleagues, Philco relied primarily upon two exhibits to support the contention that conception occurred in the Spring of 1949.

The first document, a paper entitled "A Time Division Multiplexing System",^[54] was prepared for delivery at the AIEE winter meeting in 1949. This paper teaches a time division multiplexing system which enables the simultaneous transmission of 30 audio inputs. These inputs are sampled at an 8 kilocycle rate,^[55] which is higher than the highest frequencies normally generated by human speech.^[56] The resultant pulse train will consist of a snatch of audio information from channel one followed by a snatch of information from channel two, until all thirty channels have been sampled, then a second snatch of information from channel one will be transmitted. Each individual channel will be sampled 8,000 times every second. At the receiver the pulse train is broken up, and the individual pulses are routed to individual listening stations. In order that the receiver may discriminate between the thirty incoming pulses, and send them to the proper listening stations, a 3,900 cycle "pilot tone" is added to one of the input channels.^[57] Basically, the technology involved here is quite similar to the technology involved in multiplexing color television signals. The only salient difference is in the rate of sampling.

The second exhibit, is a paper based upon a speech by Boothroyd, given at the March 1949 meeting of the IRE.^[58] The speech deals with a series of problems associated with the transmission of information through facilities which are restricted by bandwidth limitations. The speech contains no discussion of color television or interlacing, but concerns itself with the time division multiplexing of monochrome television. The paper teaches the use of a sampling frequency of 8 megacycles,^[59] and demonstrates a sufficient familiarity with time division multiplexing at the frequencies requisite for television transmission for the Court to conclude that Boothroyd and Creamer were expert in that field. But, time division multiplexing was not new in 1949; RCA had been using equipment similar to that recommended in the AIEE paper for some time.^[60] Aside from demonstrating the necessary background familiarity, these two papers do not advance Philco's conception case for the invention here at issue.

The testimonies of Smith^[61] and Bradley^[62] cannot be relied upon to meet the burden Philco must shoulder to prove conception. Smith, Philco's Vice President in charge of Research, testified that Bradley, Boothroyd and Creamer's superior, had come to him in early April *37 of 1949 with news of Boothroyd and Creamer's invention of dot interlace:

 

"Shortly after the March I.R.E., that's the general meeting in March of 1949, Mr. Bradley came to me and told me that Mr. Bradley I should identify was the Director of Research for the company and Mr. Boothroyd and Mr. Creamer were in one section which reported directly to him he came to me and told me that Boothroyd and Creamer had come up with an idea for increasing the amount of information which could be transmitted through a conventional television channel by a method which they called `Dot interlace', and, secondly, that this method would make it possible to transmit color information, in which we were very much interested, by the same technology and by using the time multiplex methods for getting the three red, green and blue channels transmitted.

 

"This happened either in late March or early April of 1949." 1 Boothroyd Record at 64-65.

In his subsequent testimony Smith proceeded to detail much the same system as that recalled by Bradley, but since the latter's position in the corporate hierarchy placed him closer to the actual inventors, Bradley's recollection is more detailed.^[63] The system allegedly disclosed to Bradley is a time division multiplexed, dot interlaced color television system. The system recounted by Bradley utilizes a 2.7 megacycle carrier wave generator.^[64] The output of the carrier wave generator is modulated by a 0°-180° switch so that it reverses its phase every 1/30th of a second.^[65] The 2.7 megacycle wave, with the periodic phase reversal, is then led into three blocks which comprise a "delay line"; one block delays the signal by 120°, a second block delays it by 240°, and a third block permits it to pass unimpeded.^[66] The outputs from the respective delay blocks are then used to power the gate circuits which sample the inputs derived from the three color cameras. Under Bradley's recollection the carrier wave frequency of 2.7 megacycles would be divided by two and then multiplied by three to 4 megacycles before being introduced into the video signal as a continuous element thereof for the purpose of coordinating the sampling at the receiver.^[67] It was theorized that the "translation" of the carrier wave frequency to the uppermost edge of the filtered video signal would avoid contamination of the sampling frequency by the video and would avoid the possibility of the carrier wave's being visible on the receiver screen. At the receiver the translated carrier wave frequency would be picked off by a highly sensitive filter,^[68] and used to lock in an oscillator running at one-third the translated frequency.^[69] The output of this oscillator would then be multiplied by two to recreate the 2.7 megacycle sampling frequency.^[70] The 2.7 megacycle frequency would then be led through phase shifting networks peculiar to the colors and used to desample the picture information.^[71]

Smith and Bradley's recollections^[72] are insufficient to establish conception *38 for two reasons. First, the Court is not convinced that Boothroyd and Creamer's conception in April of 1949 was as finely honed as Bradley's recollection would suggest. The passage of time, and the hurdles to accurate recollection imposed by the intervention of numerous important television technology developments cannot help but color the testimony of even the most precise witness.^[73] Although a contemporary Research Status Report bears out Bradley's recollection that Boothroyd and Creamer first suggested dot interlace in April of 1949,^[74] the contemporary documentary evidence is at variance with the systems recollected by Messrs. Smith and Bradley, suggesting that these witnesses may have inadvertently borrowed some of the later conception developments at Philco, and recalled that these later developments occurred at an earlier point in time. The Research Status Report's description of the invention is generic, and relates to results hoped to be attained from the system rather than the means for implementing the system. And, the Patent Disclosure Sheet of Boothroyd and Creamer,^[75] although not being relied upon by the plaintiff to prove conception,^[76] discloses a completely different approach to achieving dot interlaced color television. The Patent Disclosure Sheet, which bears the date March 23, 1949, teaches a system of color television which does not comport with the disclosures recounted by Smith and Bradley. The system disclosed in the Patent Disclosure is what Schroeder called a "simplex" system.^[77] That is, the outputs of each of the three color cameras would be individually interlaced and time division multiplexed at a 4.5 megacycle rate, then the resultant trains could be time division multiplexed into a 13.5 megacycle bandwidth. Even the two suggested "manipulations" involve systems which differ markedly from those recalled by Smith and Bradley,^[78] requiring as they do bandwidths of 6 and 8 megacycles respectively. If Boothroyd and Creamer were as far along toward conception as they *39 would have the Court believe in March and April of 1949, it is unlikely that they would have filed an invention disclosure for a system such as that taught in the March 23d Disclosure.

Second, even if Smith and Bradley's recollections of the stage of Boothroyd and Creamer's conception as of March-April 1949 be taken at face value, the system described by those witnesses is inoperative. That system calls for a "continuous carrier wave" method of synchronizing the receiver with the transmitter. For reasons more fully set forth below, the continuous carrier wave synchronization method is completely inoperative.^[79] Boothroyd and Creamer themselves were later to realize this, and their system was modified, before their patent application was filed, to provide for the burst method of synchronization.

The major piece of new evidence on conception proferred to this Court which was not before the Board of Patent Interferences is an article which appeared in the December 1949 and January 1950 issues of Electronics.^[80] Preliminary drafts of this paper were written on September 15th and September 21st.^[81] The Ballard application, it will be recalled, was filed September 24th. Since the Court has already held that the RCA conception occurred in July of 1949, the fact that these drafts preceded the Ballard application is irrelevant. But, the plaintiff's argument based on these drafts goes farther, the plaintiff argues that the Electronics drafts demonstrate that the AIEE and IRE Papers^[82] did indeed contain all the theory requisite to an understanding of dot interlaced color television, and that the Electronics drafts reveal that Philco had developed the system well before September, at the time of the disclosures to Smith and Bradley. Although the ultimate draft and articles call for the "burst" method of receiver synchronization,^[83] the early drafts do not; they call for the continuous transmission of a low level carrier wave.

The burst method uses a gate circuit to transmit a portion of the carrier wave during the horizontal blanking interval when the video information is not being transmitted. Naturally, there is not any danger of the carrier wave contaminating the video or vice versa. When the continuous carrier wave method is used, the carrier wave frequency is intermixed with the video information, and sent along with it. Perceiving that the carrier wave might contaminate the video waves, Boothroyd and Creamer, in their alleged disclosures to Smith^[84] and Bradley, and again in the early drafts of the Electronics article,^[85] advocated either the division of the carrier wave frequency by two followed by a multiplication by three to translate the carrier wave to the edge of the video at 4.02 megacycles or a reduction in the strength of the carrier wave by 30 decibels. Although Boothroyd and Creamer were sensitive to the possibility of the carrier wave's contaminating the video signals,^[86] they *40 failed to consider the converse possibility that the video signals, especially after the carrier signal is reduced in strength, might contaminate the carrier wave.^[87] Schroeder's testimony demonstrates,^[88] and Dr. Friend was forced to admit,^[89] that the video would swamp out the carrier wave, throwing the colors at the receiver into whatever color happened to be the reference phase of the carrier wave. Thus, to offer a dramatic example, with a red reference phase a green golf course and blue sky would quickly shift on the receiver's screen to a red image as the video signals drowned out the carrier wave, and took control of it.^[90] Even without the contamination phenomenon just recounted, the Boothroyd and Creamer proposal^[91] to translate the carrier wave to the edge of the video band, involving as it must, the use of multipliers and dividers, would have the tendency to introduce phase distortions into the carrier wave which would be deleterious for accurate color reproduction.^[92] Finally, it is interesting to note as a footnote to the Court's earlier caveat about the unreliability of oral testimony when the events transpired long ago, that although Boothroyd and Creamer allegedly discussed the necessity for frequency translation with Smith and Bradley in early April of 1949, the September 15th and September 21st drafts of the Electronics paper do not discuss this proposal.

*41 Dr. Friend's suggested remedies for the defects inherent in continuous carrier wave synchronization are in themselves far removed from the Boothroyd and Creamer disclosure. Dr. Friend suggested that the 2.68 megacycle carrier wave be directly injected into the transmitted signal at an amplitude below the "black level" of the video signal, and that the receiver incorporate both a frequency selective and an amplitude selective filter to screen out this low level carrier wave.^[93] Since the amplitude selective filter would only be responsive to 2.68 megacycle frequencies with an amplitude below the "black level", the only portion of the continuous carrier wave which would be filtered out would be that portion transmitted during the horizontal and vertical blanking intervals. The segments of the carrier wave transmitted during the blanking intervals could not possibly be contaminated by the non-existent video information. While the Court, somewhat guardedly, concludes that Dr. Friend's suggestion might be operable under some conditions, and might remedy the defects in Boothroyd and Creamer's proposals, the Court feels that the suggestions of Dr. Friend are too far afield from the Boothroyd and Creamer disclosures to be within the ambit of those disclosures.^[94] Dr. Friend's suggested modification would be a change of sufficient importance that it might be denominated an "invention" in its own right.^[95] In any case Dr. Friend's suggestion is something more than the brainchild of a mere "mechanic".^[96] And, the accepted standard for completeness of conception is that nothing remains to be done, except for a mechanic, skilled in the art, to assemble the system.^[97] As of September 24, 1949, the date of Ballard's application, the Boothroyd and Creamer system was inoperative; it did not meet the standard requisite for a complete conception. Therefore, the Boothroyd and Creamer effort to overcome the priority accorded Ballard by his filing date must fail.

Support for this conclusion may be found in the diligence evidence offered by the respective parties with respect to Philco's activities during the Summer and Fall of 1949, following the alleged conception date. The Court does not propose an exhaustive review of the diligence issue. The question of diligence is superfluous since Boothroyd and Creamer have failed to upset the September 24, 1949 application priority date of Ballard, much less the July 22, 1949 conception date which this Court has found the evidence to warrant. But, the diligence evidence does provide an interesting backdrop, and when the Philco allegations of priority are viewed against *42 this backdrop they become even more unconvincing.

According to Philco's theory of this case, Boothroyd and Creamer conceived the system in the Spring of 1949, and polished and refined it until the date of their patent application. In fact, the evidence indicated that Philco was interested in Boothroyd and Creamer's suggestions only as one of a number of systems. The Philco position in color television in the Summer and Fall of 1949 may be analogized to RCA's position in the color television field from 1947 to the Spring of 1949 the Ballard system, incomplete and untried had not yet gained currency in the Kell group, and other methods for conserving bandwidth were being pushed, notably quantizing. Similarly, at Philco the Boothroyd and Creamer system was far from preeminent, and other systems were vieing for the honor of becoming Philco's entrant in the color television market.^[98] Prominent among these competing systems was one proposed by Color Television Incorporated (CTI). The CTI system was the line sequential system which had been tried and discarded by RCA; it advocated transmitting one entire line in red, followed by a line of blue, followed by a line of green. The CTI system was so highly thought of at Philco that two of the four members of the color television group Moore and Creamer were sent to San Francisco to assist CTI in preparing and perfecting the line sequential system.^[99] The resulting six week absence of two of the four men in Philco's color television section provides convincing support for the Court's conclusion that conception of the Boothroyd and Creamer invention did not crystallize until some time in the late Fall of 1949. In fact, Smith, Philco's Vice President in charge of research and engineering, indicates that Philco was preoccupied with the CTI and CBS systems throughout the Fall of 1949.^[100]

In conclusion, and by way of recapitulation, the Court finds that Ballard conceived his system sometime before July 22, 1949. The Boothroyd and Creamer conception, on the other hand, occurred at the earliest on October 5, 1949,^[101] and most likely did not occur until late in the Fall of 1949. Therefore, Ballard was the first to conceive the invention, and the first to constructively reduce it to practice by filing an application for a patent on September 24, 1949 before Boothroyd and Creamer had completely conceived their invention.

The Court wishes to emphasize that no decision has been essayed with respect to Philco's alleged diligence from the period commencing in the Spring of 1949 *43 and terminating with the patent application filed by Boothroyd and Creamer on January 21, 1950, nor as previously stated have the tests undertaken at RCA during the Summer of 1949 been found to be an actual reduction to practice of the Ballard invention. Evidence relating to both these events has been used solely for the purpose of supporting and illuminating the Court's conclusions with respect to the conception issues.

In accordance with the foregoing opinion, which shall constitute the Court's Findings of Fact and Conclusions of Law pursuant to Rule 52(a) of the Federal Rules of Civil Procedure, judgment shall be entered against the plaintiff, Philco, and in favor of the defendant. RCA.

Let an appropriate order, in conformity herewith, be submitted.

NOTES

[1] An interference is a Patent Office proceeding which is designed to determine priority of invention between two or more parties claiming the same subject matter. After the Examiner determines that the application contains patentable subject matter, he is required to make an interference search; if an interfering application or patent is found an interference is declared. 35 U.S.C.A. § 135 (1954). Rule 201 Patent Office, 37 C.F.R. § 1.201 (1949). Before an interference is declared the junior party, the party with the latest application date, is required to file a preliminary correspondence setting forth the earliest act or event which will establish his conception. Rule 202, 37 C.F.R. § 1.202 (1949). If the junior applicant's alleged date of conception precedes the filing date of the senior application an interference will be declared and preliminary statements will be required under Rules 209 and 215, 37 C.F.R. §§ 1.209, 1.215 (1949). The preliminary statements serve to join the conception issues between the junior and senior parties. After the preliminary statements have been filed, the case proceeds to a hearing before the Board of Interferences where the junior party has the burden of proving a conception date in advance of the filing date of the senior party by a preponderance of the evidence. Gladrow v. Weisz, 354 F.2d 464 (5th Cir. 1965).

[2] "Diligence" for purposes of § 135 and § 146 means that the inventor must continuously direct his efforts toward a reduction to practice of his invention from a time just prior to his opponent's entry into the field until the invention is reduced to practice. Radio Corp. of Amerca v. Philco Corp., 201 F. Supp. 135, 151 (E.D.Pa.1961).

[3] "To recapitulate, applying the well settled standards of conception, neither party has established conception prior to his filing date, and RCA must prevail on the Ballard patent." Boothroyd and Creamer v. Ballard, Patent Interference No. 87,543 at page 13.

[4] "This [§ 146] suit is a trial de novo of the interference issue, not an appeal from the Board's decision." Union Carbide Corp. v. Traver Investments, Inc., 238 F. Supp. 540, 542 (S.D.Ill.1965).

[5] Tr. 660-65.

[6] Tr. 119.

[7] In other words, high sampling frequencies permit the reconstruction of picture information on the television screen which, although it consists of a number of small dots, is not perceived as discontinuous by the eye at normal viewing distances.

[8] This phenomenon is known as retentivity of the eye; it is what accounts for the illusion of motion when the separate frames of a motion picture are flashed on the screen at a rate which is sufficiently rapid that the eye merges the separate still images into a continuous "moving" image.

[9] The camera is controlled by deflection circuitry. The deflection circuits consist of electromagnets which, when activated by an appropriate frequency from the synchronizing generator "deflect" the camera across the image whose reproduction is desired. Tr. 105-06.

[10] This scanning of first the odd numbered lines followed by the even numbered lines is called "line interlace"; another refinement which increases resolution, line interlace, was well-known in the color television art by 1947-49. Tr. 113-14.

[11] Tr. 135.

[12] Tr. 137, 140.

[13] Tr. 150.

[14] Ibid.

[15] The Ballard patent refers to the image reproduction portion of the system as a "standard kinescope", United States Letters Patent No. 2,678,348, col. 7, lines 2-3, hereinafter Ballard Patent. But, the Court has thought it more enlightening to give a more elaborate account of the "kinescope" which it gained from a very helpful meeting at a Pretrial Conference with counsel and Mr. John Wentworth on March 6, 1967.

[16] Tr. 121, 211-13.

[17] Boothroyd, "Dot Systems of Color Television", Electronics, p. 97 (January 1950); PX 2-2 at 97.

[18] Tr. 696-99.

[19] PX 13; Tr. 294.

Actually, the output of the sampling oscillator is led into a "phase splitting network" which splits the sampling frequency into three phases one phase remains unaltered, one is advanced 120°, one is retarded 120° and these three separate phases are used to power the keying circuits which actually sample the color information. Ballard Patent, col. 4, lines 34-75, col. 5, lines 1-5.

[20] This is because 1/30th second is the time required to transmit one frame of picture information.

[21] Preliminary Statement of Randall C. Ballard, 1 Record on Behalf of Randall C. Ballard 7-8 (hereinafter Ballard Record).

[22] PX 1-A "RCA Patent Disclosure Sheet."

PX 1-B Excerpts Ballard's 1947 Notebook "Color Television System Proposal."

[23] PX 1-A pp. 1-2.

[24] "A control frequency determining the rate of switching is generated at the transmitter. This is transmitted to the receiver where, by phase, or time control, the sequence is maintained in synchronism [in phase] with the transmitter. * * * The chief difficulty would be maintenance of phase synchronism [between transmitter and receiver]. * * * The switching may be accomplished by a multivibrator operating on a pulse [which unfortunately would be integrated by bandwidth limitations making precise switching difficult]. However it would not be as susceptible to undesirable phase shifting."

[25] Specifically, the diagram on page 13 of Ballard's 1947 notes does not show the gate circuitry necessary for the use of the "burst" system of receiver synchronization, used in the Ballard application.

When questioned about the absence of circuit diagrams Ballard replied that he did not feel such diagrams were necessary because the "idea or invention" did not present any "unforeseen difficulties". 1 Ballard Record, p. 22. The Court is not as convinced as Mr. Ballard was that his system was free from difficulties. Certainly the fact that the transmitter-receiver synchronism was not completely thought out when coupled with the fact that the "continuous carrier wave" approach which Ballard was suggesting in 1947 would suggest that the Ballard system had some conceptual interstices which a circuit diagram might have resolved.

[26] PX 1-B, pp. 6, 11, 16, 17, 18, 19, 20, 21, 22.

[27] Ballard Record pp. 239-324 (Darke); pp. 325-456 (Bedford); pp. 457-85 (Sziklai).

[28] PX 1 L & M are Bedford's diagrams which recollect Ballard's disclosures to him in 1947 through laboratory notebooks and other conversations. As Bedford testified:

"Mr. Ballard and I were very close together in this work, and we have both been very much interested in interlace scanning, had been for a long time.

* * * * *

"To highlight that relation between us I might say that we really talked the same language in this particular field, and I mean `really'. We understood each other very fully.

"So I can very well sketch a diagram which would represent the system as he intended it to be." 2 Ballard Record, p. 346.

Mr. Sziklai's testimony is concerned with the various intricate circuits which Ballard proposed in his 1947 notes.

George H. Brown, also a member of the Kell group RCA's Research Division who was assigned by RCA Vice President Engstrom in 1947 to take a closer interest in the activities of the television group, recalled that Ballard had made a complete disclosure of the system to him in March-April 1947, 2 Ballard Record pp. 495-97.

[29] Bedford's testimony indicates that in 1947 the means for maintaining transmitter-receiver synchronism had not been settled on. Bedford, at several places in his testimony, indicates that RCA and Ballard were evaluating the "continuous carrier wave" approach which Schroeder testified at trial was inoperative. 2 Ballard Record pp. 332, 333, 336-38, 354. See discussion at nn. 80-96, infra.

[30] 1 Ballard Record pp. 265, 270-71.

[31] 1 Ballard Record pp. 241-42.

[32] PX 20. The efficacy of Ballard's proposal for meeting the compatibility problem was becoming apparent to those at RCA. Ballard's 1948 notes reveal that he and Cherry had sought to evaluate the parameters of his system, and Cherry had concluded that maximum resolution was obtainable with a 3 megacycle switching rate and that transmission could be limited to the 4½ megacycle bandwidth imposed by compatibility restrictions.

[33] Id. at p. 10.

[34] PX 25 at p. 2; Tr. 419.

[35] Tr. 840-41.

[36] PX 20 at p. 1.

[37] Tr. 837-43.

[38] See n. 32 supra.

[39] DX DDD p. 2.

[40] Compare DX DDD p. 2 with PX 14 and PX 15.

DX DDD shows a color camera (PX 14 # 2) with three video outputs (PX 14 #s 4, 6, 8) leading to an electronic switch (PX 14 #s 64, 66, 68 and the delay circuitry represented by PX 14 #s 50, 52 & 54, 58 & 60). From the electronic switch the "interlaced sequential video" is transported to a modulator and sync mixer (PX 14 # 92). The camera is controlled by the sync generator (PX 14 # 94) which also controls the keying oscillator and low frequency switch (PX 14 #s 24, 46). The low frequency switch emits two waveforms, 180° out of phase (PX 14 #s 42, 44) which reverse the polarity of the sampling frequency. The sync generator controls the receiver's sampling and deflection circuitry by being interjected into the television signal on the "back porch" (PX 14 # 134). The signal is ultimately sent by a television transmitter (PX 14 # 156).

The 1949 notes describe a Television Receiver (PX 15 # 160) without the sync separation circuitry. A burst of the basic keying frequency is applied to the grid of a tube parallel to the receiver's sampling oscillator tube (PX 15 #s 190, 191, 192). The receiver's sampling oscillator is thus "slaved" to the transmitter's sampling oscillator (PX 15 # 174). The "low frequency switch" is not needed at the receiver because the use of a "start-stop" sampling oscillator makes the switch superfluous. The output of the keying oscillator is used to drive an electronic switch (PX 15 #s 162, 164, 166 and delay line circuitry PX 15 # 176). The electronic switch sends a green video to the green picture tube, red video to the red tube, and blue video to the blue tube (PX 15 #s 168, 170, 172).

[41] Although Ballard's patent teaches a tie-in between the sync generator and the multivibrator used for reversing the phase of the sampling oscillator so that the phase reversal will occur during the vertical blanking interval, there is no such interlocking between the sync generator and the sampling oscillator.

[42] 37 C.F.R. § 1.232 (1949).

[43] Ballard's earliest notes bespeak their author's awareness of the problem. "In order to avoid crawling effects, it is essential to maintain a fixed relation between scanning frequency and switching frequency." Ballard's 1947 Laboratory Notebook, PX 1-B, p. 2.

[44] 2 Ballard Record pp. 432-33.

[45] "The phasing of the switching may be accomplished by a local oscillator keyed on in the correct phase at the beginning of each line. * * *" PX 1-B p. 3.

[46] Dr. Friend testified that if the sampling oscillator were not tied to the deflection circuitry an extremely stable oscillator would be required to avoid crawl, and that such oscillators were unavailable in 1949. Tr. 343-44. On cross-examination Friend testified that oscillator stability of one-half part per million would be the starting point for crawl-free interlace patterns of the type proposed by Ballard, and that such oscillators were not available in 1949. Tr. 510-12. When confronted with an article from the May 1940 issue of the proceedings of the IRE by Mason entitled "A New Quartz-Crystal Plate, Designated the GT", DX EE, which taught an oscillator with a stability of 1 part in 10,000,000 over a 30° C temperature range, and a second article by Meacham entitled "The Bridge-Stabilized Oscillator", DX FF, which described an oscillator stability of 2 parts in 100,000,000, Dr. Friend was forced to modify his earlier testimony that such oscillator stability was unavailable in 1947-49. Tr. 517. In fact, such oscillators, i.e., crystals, were used in the Loran (Long Range Aid to Navigation) systems of that time. Tr. 517-17A. Further, the current television systems use a high frequency crystal oscillator for sampling and divide down to get the deflection signals. Tr. 430.

[47] Schroeder's testimony is perhaps the most incisive on this point:

"Q. [D]id Ballard propose such an interlock [between horizontal deflection and sampling] back in 1947?

"A. Yes, certainly.

* * * * *

"A. Yes, he certainly did propose interlocking. That was the whole point. Evans had proposed interlocking. He was interlacing what Evans had proposed, and of course he knew and all the way through, this interlocking is described, implied, and in every other way indicated." Tr. 688-89.

Nor could Schroeder be shaken on cross-examination:

"Q. Did you see a diagram which indicated the manner in which this was to be done [interlocking the sampling and the horizontal], a diagram which was drawn by Mr. Ballard?

"A. When he was describing it to me it was not necessary for him to do that.

* * * * *

"Q. Did Mr. Ballard describe to you how that [interlocking] was to be done?

"A. With my knowledge of how things like that should be done it was not necessary for him to do so." Tr. 933-34.

In communicating with Schroeder, who had built the distribution checker, a description of interlocking would have been banal. "[Schroeder] knew that whenever [someone] drew a pattern of dots that was supposedly stationary in the raster, this meant it [the sampling frequency] was tied to the deflection, tied to the horizontal." Tr. 689.

[48] PX 20, p. 1. Where Ballard refers to "the system proposed by Evans [and] modified to interlace relations by myself."

[49] The Ballard patent is entitled "Color Television Interlacing System". U. S. Letters Patent No. 2,678,348.

[50] See also Evans Patent Disclosure Sheet, DX GG.

[51] DX ZZ, AAA, BBB, CCC; Tr. 719-49. See particularly Schroeder's comparison of Ballard's block diagram of July 22, 1949, DX DDD, with the ultimate August 10, 1949 test, DX CCC at pages 747-49.

[52] The interlace chassis according to the limited evidence available to the Court employed a start-stop oscillator keyed on by the trailing edge of the horizontal sync pulse. By means of two potentiometers the width of the horizontal sync pulse could be varied in order to start the oscillator 90°, 120° or 180° out of phase every other line, every third line or every other field. This enabled the RCA engineers to evaluate a number of interlace proposals without having to separately construct the "keying circuitry" for each interlace pattern. Schroeder's Laboratory Notebook, DX SS p. 5; Tr. 719-21.

Although the synchronism between the transmitter sampling and the receiver sampling and the phase reversal were provided by Schroeder's jiggled back edge, the tests do fortify the maturity of Ballard's conception. Schroeder's invention was an improvement on Ballard's invention, an alternative and superior way of accomplishing Ballard's objective. Schroeder's invention in no way detracts from Ballard's. In fact, Schroeder concedes his intellectual debt to Ballard in the body of the patent:

"A system of dot or element interlace has been very well shown and described in the co-pending U.S. application to Randall Ballard, Serial No. 117,528, entitled System of Color Television, filed September 24, 1949." U. S. Letters Patent No. 2,678,351 col. 3, lines 73-75, col. 4, lines 1-2; PX 22, col. 3, lines 73-75, col. 4, lines 1-2.

Schroeder recognized that

"Various [other] arrangements have been proposed for controlling and changing the phase with each line by employment of special arrangements, including the transmission of sampling oscillator control frequency or alternate switching arrangements included in the receiving circuit." Id. col. 4, lines 9-15.

[53] Subparagraphs (a)-(c), Preliminary Statement Wilson Boothroyd and Edgar Creamer, 1 Record on Behalf of Wilson P. Boothroyd and Edgar M. Creamer, Jr., pp. 15-17 (hereinafter Boothroyd Record).

[54] PX 1-2. AIEE stands for American Institute of Electrical Engineers.

[55] The 8 kilocycle frequency is derived from a pulsed oscillator which is keyed on by the output of a 60 kilocycle carrier oscillator through frequency counting and doubling circuits much the same way that Boothroyd and Creamer ultimately proposed multiplication of the output of the sync generator to derive the sampling frequency. PX 1-2, p. 2.

[56] "Q. * * * What kind of channels were there involved in this AIEE paper, which is Plaintiff's Exhibit 1-2?

"A. These are, as I believe I indicated, speech channels, which I think in this instance represented audible frequencies up to about 3.3 kilocycles " Tr. 128.

[57] PX 1-2, p. 2.

[58] PX 1-3. Research Division Philco Corporation, "Report on the Transmission of Information through Band Limited Transmission Systems" (1949).

[59] PX 1-3, p. 8.

[60] Tr. 631-35.

[61] 1 Boothroyd Record 35-157.

[62] Id. at 158-352.

[63] Bradley was the Director of Research at Philco during the 1949 period; Boothroyd and Creamer reported directly to Bradley who in turn reported to Smith. Id. at 36, 64-65, 194.

[64] Id. at 196. The figures used are Bradley's; it is obvious that they are merely approximations. See n. 67 where Bradley testifies that the division of 2.7 megacycles by two followed by a multiplication by 3 would yield a 4 megacycle signal. Actually, the process recalled by Bradley would result in a 4.05 megacycle signal.

[65] Id. at 197-98.

[66] Id. at 196-97.

[67] Id. at 199.

[68] Ibid.

[69] Ibid.

[70] Ibid.

[71] Id. at 199-200.

[72] In addition to and in clarification of their respective testimonies Smith and Bradley each sought to schematically recreate the Boothroyd and Creamer disclosures. See PX 1-5 and PX 1-13 and 14, respectively.

[73] Both the Supreme Court and the Court of Patent Appeals have cautioned against the indiscriminate use of witnesses' recollections to plug gaps in the parties' proofs of conception.

"Granting the witnesses to be of the highest character, and never (sic) so conscientious in their desire to tell only the truth, the possibility of their being mistaken as to the exact device used, which, though bearing a general resemblance to the one patented, may differ from it in the very particular which makes it patentable, is such as to render oral testimony peculiarly untrustworthy; particularly so if the testimony be taken after the lapse of years from the time the alleged anticipating device was used." Deering v. Winona Harvester Works, 155 U.S. 286, at 301, 15 S. Ct. 118, at 123, 39 L. Ed. 153 (1894).

"Such a situation requires careful scrutiny of the evidence bearing in mind the tendency of witnesses, granting them to be of the highest character, to confuse knowledge acquired subsequent to an event with that acquired at the time the event occurred." Herrmann v. Otken, 201 F.2d 909, at 913, 40 C.C. P.A. 794 (1953).

[74] PX 1-6. "Status Report of the Research Division as of April 1, 1949."

[75] PX 1-7.

[76] See Plaintiff's Reply Brief at p. 7.

[77] Tr. 750-52.

[78] "However, now we can manipulate!

"1. If we maintain horizontal interlace we can have 60/15 frame rate and 345 dot horizontal resolution with a video bandwidth of 2 mc per color or 6 mc total. In black and white, this resolution is equivalent to present black and white resolution.

"2. We can modify our system as follows: double interlacing of all three colors at the transmitter (in order to simplify the transmitter scanner) followed by the transmission of the green color at an 8 mc dot rate and the red and blue at interleaved 4 mc dot rates for a second 8 mc carrier. Thus a total of 8 mc video bandwidth with a composite 16 mc carrier rate would yield good quality color. Red and blue colors would only have 2 mc resolution but tests have indicated this to be adequate provided the green color is of full resolution. Further the green flicker rate would be twice that of red or blue. This is important." PX 2-7 at pp. 5-6.

[79] See text at nn. 83-92, infra.

[80] PX 2-1 to PX 2-16. Originally these papers were numbered "Plaintiff's Boothroyd Deposition Exhibits 1-14". Since the numbers used when the papers were put in evidence before this Court are at variance with those used at the taking of Mr. Boothroyd's deposition, the parties have often incorrectly referred to a particular portion of PX 2. The Court, in its discussion, will use the evidentiary numbers given the papers by the Clerk of this Court.

[81] PX 2-8; PX 2-11. Almost invariably referred to as PX 2-7 and PX 2-9 respectively, by the parties.

[82] PX 1-2 and PX 1-3.

[83] Draft of October 5, 1949, PX 2-3 at pages 23-24; Boothroyd, "Dot Systems of Color Television", Electronics (January 1950) at pp. 97-98; PX 2-2 at pp. 97-98.

[84] 1 Boothroyd Record pp. 198-99.

[85] PX 2-8 at p. 6; PX 2-11 at p. 7.

[86] As Boothroyd's draft of September 15, 1949 notes:

"Such a fine grain, stationary, low level pattern is believed invisible under normal viewing conditions." PX 2-8 at p. 7.

[87] That is shift the phase of the carrier wave as the strong video impulses took control of the receiver's carrier wave oscillator, shifting the phase thereof.

[88] "Q. Is such a system [continuous carrier wave transmission] operative?

"A. Absolutely not.

* * * * *

"Q. Would you explain, please, the basis for your testimony that the system described in [PX 2-11] is inoperative?

"A. Well, it seems as though the writer of this paper did not understand what the output signal from a sampler would look like, and he therefore made a number of errors. He did not understand, for instance, that when there was a color a large area of color as part of the picture there would be a large amount of the sampling frequency as part of the signal. He therefore worries about and one of the things he says about the visibility of this three percent synchronizing signal and completely neglects the visibility of the actual signal itself, the dots in the actual signal itself; and because of this lack of understanding of what the signal looks like he worries about the interference that the continuous wave synchronizing signal will cause to the picture, but completely ignores the fact that the picture signal itself will completely swamp out, because it is many times bigger than completely swamp out the three percent synchronizing signal, and when he tries to use this very low level signal at the receiver it is just completely confused with and by the color signal itself." Tr. 753-54.

[89] See cross-examination of Dr. Friend. Tr. at 475.

[90] Ibid.

[91] This proposal was recounted by Bradley. 1 Boothroyd Record at 198-99.

[92] Schroeder testified that the transmission path at 4.02 megacycles, 4/3 the 2.68 megacycle sampling frequency, would be sufficiently different from the transmission path at the 2.68 megacycle frequency so that phase distortion would be introduced by the "vagaries of transmission" in addition to the phase distortion resulting from the frequency multiplication and division circuitry. Tr. 645.

Schroeder had himself tried an analogous arrangement. At one point RCA theorized that more depth of modulation could be introduced into the video portion of the television signal if the sync pulses were sent at a different channel. The system used to accomplish this "sync on sound" sent the horizontal and vertical synchronizing information on the sound channel. RCA found that the transmission path was so different at the sound channel that the edge of the picture at the receiver would wander all over the screen. Tr. 644-45. Based on this experience Schroeder concluded:

"[T]his business of trying to send the carrier at a different frequency, a sampling carrier at a different frequency from the sampling frequency. These are two carriers. They are 1-3/10ths megacycles apart, roughly, and my opinion is that in a practical case over the air, not in a closed circuit necessarily, but in a practical case over the air this would not work." Tr. 645.

[93] PX 27; Tr. 569-73.

[94] The Court's temerity stems from the glaring inconsistency between the testimonies of Drs. Friend and Schroeder. Dr. Friend opined that since the video signals do not cross the blanking level, a carrier wave with an amplitude below the blanking level would not be contaminated except by substantially saturated color information. Tr. 594-95. Schroeder on the other hand testified that the amplitude of the video signal would swing blacker than black anytime a 50% color saturation was being broadcast. When the color saturation reached 100% the signal would swing 33% of the peak signal in the direction blacker than black. Tr. 768. To make certain that the reduced amplitude carrier wave was free from contamination would require setting the black level at least 33% below the pedestal level where the carrier wave would be transmitted. This would entail a weakening of the television signal. Tr. 769.

[95] Tr. 772-73.

[96] As Schroeder testified:

"I don't believe it is suggested any place that you only modulate with 50 percent of the signal in order to avoid the signal from the picture going into the area of the blacker than black. So that is another invention it seems to me. Yes, if you make all these other inventions, it [the Boothroyd & Creamer system as taught in PX 28 & 11 and modified by Dr. Friend] might work poorly, but perhaps work." Ibid. (Emphasis added).

[97] Summers v. Vogel, 332 F.2d 810, 814 (Cust. & Pat.App., 1964), harking back to the classic Merganthaler v. Scudder, 11 App.D.C. 264 (1897), formulation.

[98] In a memorandum dated July 6, 1949 Smith described "multiple interlace", one of the elements involved in Boothroyd and Creamer's invention, as follows in connection with suggested refinements for the CTI System:

"They [CTI] propose to try a well-known expedient to overcome the low resolution. This [interlacing] was not demonstrated to us, nor has it worked in the past." DX EEE1 at p. 3.

DX EEE16 Smith, "Chronological History of Color Television at Philco", is persuasive evidence that Philco's Boothroyd and Creamer system was not fully thought out in the Summer of 1949.

"Thus, while in the summer of 1949 we were hopeful that it would be possible to make a completely compatible color signal on a 6 mc. channel, it was not until we had been able to carry out these tests in early '50 that we had become reasonably sure of that fact." DX EEE16 at p. 9. (Emphasis added).

Notice that in the Summer of 1949 Philco was "hopeful" and that even by early 1950 it was only "reasonably sure"; this language does not impress the Court that Philco had completely conceived the invention, as it maintains, in April of 1949.

Further, DX EEE2 indicates that Philco was evaluating the CBS System in addition to the CTI System.

[99] DX EEE4 demonstrates that Moore and Creamer's trip necessitated the discontinuance of some of Philco's "very important development work" inferentially work on the Boothroyd and Creamer system.

[100] DX EEE16 at p. 9.

[101] PX 2-3, Boothroyd's draft and revision of his Electronics article.