`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`VERANCE CORP.,
`Petitioner,
`
`v.
`MZ AUDIO SCIENCES, LLC,
`Patent Owner.
`____________
`
`IPR2022-01544
`Patent 7,289,961 B2
`____________
`
`
`PATENT OWNER’S RESPONSE
`UNDER 37 C.F.R. § 42.120
`
`
`
`
`

`

`TABLE OF CONTENTS
`
`INTRODUCTION ........................................................................................... 1
`I.
`LEVEL OF ORDINARY SKILL IN THE ART ............................................. 2
`II.
`III. TECHNOLOGICAL BACKGROUND .......................................................... 2
`A. General Principles and Properties of Sound Waves .............................. 2
`B.
`The ᾽961 Patent ..................................................................................... 6
`1.
`First Preferred Embodiment ........................................................ 8
`2.
`Second Preferred Embodiment ................................................... 9
`IV. THE ASSERTED PRIOR ART REFERENCES .......................................... 12
`A.
`Srinivasan (Ex. 1005) .......................................................................... 12
`B.
`Cabot (Ex. 1006) ................................................................................. 16
`C.
`Kudumakis (Ex. 1007) ......................................................................... 18
`D.
`Tilki (Ex. 1008) .................................................................................... 20
`E.
`Hobson (Ex. 1042) .............................................................................. 21
`CLAIM CONSTRUCTION .......................................................................... 23
`V.
`VI. THE PETITION FAILS TO ESTABLISH THAT ANY OF THE
`CHALLENGED CLAIMS ARE UNPATENTABLE .................................. 23
`A.
`Cabot Teaches Away From Using Fundamentals and Third
`Harmonics, Destroying Any Motivation to Combine Required
`by All Asserted Grounds ..................................................................... 23
`1.
`Cabot’s Experimental Data Contradicts Petitioner ................... 24
`2.
`Dr. Scordilis’ Testimony Undermines the Foundation of
`the Petition ................................................................................ 29
`Petitioner’s Tertiary References Would Dissuade a
`POSA ........................................................................................ 33
`
`3.
`
`i
`
`

`

`2.
`
`2.
`
`B.
`
`C.
`
`D.
`
`Petitioner Fails to Demonstrate a Reasonable Likelihood of
`Prevailing on Ground 1 ....................................................................... 36
`1.
`Srinivasan Operates Within a Specific Frequency Range
`that Is Incompatible with Petitioner’s Proposed
`Modifications ............................................................................ 36
`A POSA Would Not Have Been Motivated to Combine
`Srinivasan, Cabot, and Kudumakis ........................................... 41
`a)
`A POSA Reading Srinivasan Would Not Consider
`Cabot ............................................................................... 42
`Kudumakis Teaches Away from the Modification
`Proposed by Petitioner .................................................... 43
`Petitioner Fails to Explain How to Combine the
`Disparate Systems of the Purported
`Srinivasan/Cabot/Kudumakis Combination ................... 45
`Petitioner Fails to Demonstrate a Reasonable Likelihood of
`Prevailing on Ground 2 ....................................................................... 47
`1.
`A POSA Would Not Have Been Motivated to Combine
`Srinivasan, Cabot, Kudumakis, and Hobson ............................ 47
`Hobson is Not Analogous Art ................................................... 47
`a)
`Legal Standard for Analogous Art ................................. 48
`b)
`Hobson Does Not Qualify as Analogous Art Under
`the “Same Field of Endeavor” Test ................................ 50
`Hobson Does Not Qualify as Analogous Art Under
`the “Reasonably Pertinent to the Particular
`Problem with Which the Inventor Is Involved”
`Test.................................................................................. 55
`Petitioner Fails to Demonstrate a Reasonable Likelihood of
`Prevailing on Ground 3 ....................................................................... 57
`1.
`A POSA Would Not Have Been Motivated to Combine
`Kudumakis, Cabot, and Tilki ..................................................... 57
`
`b)
`
`c)
`
`c)
`
`ii
`
`

`

`3.
`
`4.
`
`2.
`
`Kudumakis and Cabot Do Not Motivate the Use of
`Fundamentals and Third Harmonics for Encoding ................... 63
`A POSA Would Not Have Modified Tilki to Decrease Its
`Data Rate ................................................................................... 65
`A POSA Would Not Have a Reasonable Expectation of
`Success in Modifying Tilki to Use Fundamentals and
`Third Harmonics as Reference and Signal Bins ....................... 66
`VII. CONCLUSION .............................................................................................. 67
`
`
`
`
`
`
`iii
`
`

`

`TABLE OF AUTHORITIES
`
`Cases
`
`Adidas AG, v, Nike, Inc.,
`IPR2016-00922, Paper 31 (PTAB Feb. 19, 2019) ............................................... 39
`Agamatrix, Inc. v. Dexcom, Inc.,
`IPR2018-01718, Paper 10 (PTAB Mar. 13, 2019) .............................................. 29
`Apple Inc. et al. v. Arigna Technology, Ltd.,
`IPR2022-0139, Paper 9 (PTAB Dec. 2, 20222) .................................................. 47
`Arctic Cat Inc. v. Bombardier Rec. Prods. Inc.,
`876 F.3d 1350 (Fed. Cir. 2017) ............................................................................ 28
`Axalta Coating Systems, LLC v. PPG Industries Ohio, Inc.,
`IPR2022-00676, Paper 10 (PTAB Sept. 12, 2022) .............................................. 28
`In re Bigio,
`381 F.3d 1320 (Fed. Cir. 2004) ............................................................... 48, 49, 56
`
`In re Clay,
`966 F.2d 656 (Fed. Cir. 1992) ....................................................................... 48, 49
`In re Deminski,
`796 F.2d 436 (Fed. Cir. 1986) .............................................................................. 51
`In re Fulton,
`391 F.3d 1195 (Fed. Cir. 2004) ............................................................................ 28
`In re Gurley,
`27 F.3d 551 (Fed. Cir. 1994) ......................................................................... 27, 43
`In re Kahn,
`441 F.3d 977 (Fed. Cir. 2006) .............................................................................. 43
`In re Klein,
`647 F.3d 1343 (Fed. Cir. 2011) ..................................................................... 48, 50
`In re Magnum Oil Tools Int’l, Ltd.,
`829 F.3d 1364 (Fed. Cir. 2016) ............................................................................ 41
`
`iv
`
`

`

`In re NuVasive, Inc.,
`842 F.3d 1376 (Fed. Cir. 2016) ............................................................................ 41
`In re Oetiker,
`977 F.2d 1443 (Fed. Cir. 1992) ............................................................................ 50
`
`In re Sponnoble,
`405 F.2d 578 (1969) ............................................................................................. 40
`In re Wood,
`599 F.2d 1032 (CCPA 1979) ............................................................................... 50
`
`Innovention Toys, LLC v. MGA Entm’t, Inc.,
`637 F.3d 1314 (Fed. Cir. 2011) ..................................................................... 48, 49
`Intel Corp. v. XMTT, Inc.,
`IPR2020-00144, Paper 12 (PTAB May 20, 2020) ............................................... 29
`ipDataTel, LLC v. ICN Acquisition, LLC,
`IPR2018-01822, Paper 19 (PTAB Apr. 22, 2019) ............................................... 41
`Johns Manville Corp. v. Knauf Insulation, Inc.,
`IPR2018-00827, Paper 9 (PTAB Oct. 16, 2018) ................................................. 48
`Kinetic Techs., Inc. v. Skyworks Solutions, Inc.,
`IPR2014-00529, Paper 8 (PTAB Sept. 23, 2014) ................................................ 30
`KSR Int'l Co. v. Teleflex Inc.,
`550 U.S. 398 (2007) ............................................................................................. 47
`Marvell Semiconductors, Inc. v. Intellectual Ventures I LLC,
`IPR2014-00547, Paper 17 (PTAB Dec. 3, 2014) ................................................ 44
`McGinley v. Franklin Sports, Inc.,
`262 F.3d 1339 (Fed. Cir. 2001) ..................................................................... 39, 40
`Medichem, S.A. v. Rolabo, S.L.,
`437 F.3d 1157 (Fed. Cir. 2006) ............................................................................ 43
`
`Metalcraft of Mayville, Inc. v. The Toro Co.,
`848 F.3d 1358 (Fed. Cir. 2017) ............................................................................ 46
`
`v
`
`

`

`Ormco Corp. v. Align Tech., Inc.,
`463 F.3d 1299 (Fed. Cir. 2006) ............................................................................ 43
`Par Pharmaceutical, Inc. v. Jazz Pharmaceuticals Ireland Ltd., et al.,
`IPR2016-00002, Paper 12 (PTAB Apr. 12, 2016) ............................................... 28
`Plas-Pak Industries, Inc. v. Sulzer Mixpac AG,
`600 F. App’x 755 (Fed. Cir. 2015) ...................................................................... 39
`Schott Gemtron Corp., v. SSW Holding Co., Inc.,
`IPR2014-00367, Paper 62 (PTAB May 26, 2015) ............................................... 56
`SDS USA, Inc. v. Ken Specialties, Inc.,
`No. 99-133, 2002 WL 31055997 (D.N.J. Aug. 28, 2002) ................................... 52
`Unirac, Inc. v. EcoFasten Solar, LLC,
`IPR2021-00532, Paper 7 (PTAB July 22, 2021) ................................................. 29
`Statutes
`35 U.S.C. § 103 ........................................................................................................ 56
`Rules
`37 C.F.R. § 42.65(a) ................................................................................................. 29
`
`
`
`
`
`
`vi
`
`

`

`EXHIBIT LIST
`
`Exhibit No.
`
`Brief Description
`
`2001
`
`2002
`
`2003
`
`2004
`
`2005
`
`2006
`
`2007
`
`2008
`
`2009
`
`2010
`
`2011
`
`Excerpts from John Backus, The Acoustical Foundations of
`Music (2nd ed. 1977).
`
`Excerpts from Harry F. Olson, Music, Physics and
`Engineering (2nd ed. 1967).
`
`Excerpts from McGraw-Hill Dictionary of Scientific and
`Technical Terms (6th ed. 2003).
`
`Excerpts from Arthur H. Benade, Fundamentals of Musical
`Acoustics (2nd ed. 1976).
`
`Excerpts from Harvey E. White, Physics and Music: The
`Science of Musical Sound (1980).
`
`Excerpts from Random House Webster’s Unabridged
`Dictionary (2nd ed. 2001).
`
`Excerpts from Glen M. Ballou, Handbook for Sound
`Engineers (3rd ed. 2002).
`
`U.S. Patent No. 6,995,521.
`
`Microsoft Word comparison of the specification text of Ex.
`1005 to the specification text of U.S. Patent No. 6,504,870.
`
`Shah Mahdi Hassan, Breaking down confusions over Fast
`Fourier Transform (FFT), Medium (Apr. 15, 2020),
`https://medium.com/analytics-vidhya/breaking-down-
`confusions-over-fast-fourier-transform-fft-1561a029b1ab
`(last visited July 31, 2023).
`
`Deposition transcript of Dr. Michael Scordilis dated July 28,
`2023.
`
`
`
`
`
`vii
`
`

`

`I.
`
`INTRODUCTION1
`Petitioner submitted a Petition seeking Inter Partes Review of U.S. Patent
`
`No. 7,289,961 (“the ’961 patent” (Ex. 1001)), challenging claims 1-10 (the
`
`“Challenged Claims”). MZ Audio Sciences, LLC (“Patent Owner”) requests that
`
`the Board deny the Petition because Petitioner fails to meet its burden of showing
`
`unpatentability of any of the Challenged Claims.
`
`The ’961 patent is directed to methods of embedding and extracting hidden
`
`data in audio signals by changing the phases of fundamental tones and related
`
`overtones. The techniques taught permit embedding data in an auditorily
`
`undetectable manner, while at the same time making the data robust against both
`
`blind signal processing attacks and loss due to digital to analog conversion
`
`processing.
`
`In challenging all claims of the ’961 patent, Petitioner fails to demonstrate a
`
`motivation to combine the cited references, which represent disparate methods and
`
`systems that are not readily compatible or combinable. Furthermore, several of
`
`Petitioner’s references teach away from the techniques of the ’961 patent, thus
`
`discouraging a person skilled in the art (“POSA”) from implementing the
`
`combinations proposed by Petitioner. Further, Petitioner’s expert, Dr. Scordilis,
`
`
`1 All emphasis added by Patent Owner unless indicated otherwise.
`
`1
`
`

`

`confirmed that none of the prior art references relied upon in the Petition have a
`
`watermarking algorithm that modifies fundamental and harmonic frequencies (Ex.
`
`2011, 90:4-20), which is what the challenged claims teach. These deficiencies
`
`preclude a finding that Petitioner has met its burden to demonstrate a reasonable
`
`likelihood that it would prevail in showing unpatentability of any of the
`
`Challenged Claims.
`
`II. LEVEL OF ORDINARY SKILL IN THE ART
`For purposes of this Response only, Patent Owner does not dispute
`
`Petitioner’s definition of a POSA. (Paper 7 (“Petition” or “Pet.”), 10.)
`
`III. TECHNOLOGICAL BACKGROUND
`A. General Principles and Properties of Sound Waves
`The ᾽961 patent concerns the manipulation of audio signals generated by, for
`
`example, recordings of sound waves, which are a type of energy released by
`
`vibrating objects. In the time domain, sound waves are represented as waveforms
`
`(e.g., sine waves, see below), and have several characteristics. Amplitude,
`
`measured in decibels (dB), indicates the sound wave’s relative strength or intensity
`
`(loudness). (Ex. 2001, 922; Ex. 2002, 12-13.) The period is the time required for
`
`the sound wave to complete one oscillation or wave cycle. Frequency, measured in
`
`
`2 Consistent with the Petition, Patent Owner cites to the original page numbers in
`
`its exhibits.
`
`2
`
`

`

`Hertz (Hz), indicates the number of oscillations the sound wave makes in one
`
`second, which the ear perceives as pitch. (Ex. 2002, 10.) Low-frequency sounds
`
`produce fewer oscillations per second than high-frequency sounds. Period and
`
`frequency are inverses (e.g., frequency of 1,000 Hz = period of 1/1000 second).
`
`Phase indicates “the fractional part of a period through which the time
`
`variable of a periodic quantity (alternating electric current, vibration) has moved,
`
`as measured at any point in time from an arbitrary time origin; usually expressed in
`
`terms of angular measure, with one period being equal to 360o or 2π radians.” (Ex.
`
`2003, 1572 ([PHYS] definition of “phase”). For example, a phase of 90° is 1/4 of a
`
`wave cycle, a phase of 180° is 1/2 of a wave cycle, and so forth. Phase describes
`
`the time relationship between different waveforms. If two waveforms begin their
`
`wave cycle at 0° simultaneously, they are “in phase” with one another. If, however,
`
`the second waveform starts its wave cycle when the first waveform is already 1/4
`
`of the way through its wave cycle, the two waveforms are 90° “out of phase”
`
`because 90° is 1/4 of a wave cycle. (Ex. 2001, 113 (discussing the “relative phase”
`
`of two waveforms).
`
`3
`
`

`

`(Ex. 2007, 11.)
`
`The above graph represents a pure-pitch sine wave. General experience
`
`teaches that multiple notes and pitches are present in everyday sounds and audio
`
`signals. For example, music is a collection of multiple instruments/voices
`
`playing/singing different notes concurrently, which raises the question: what does
`
`the sound wave look like when multiple pitches/sounds occur simultaneously?
`
`Individual components of a sound are combined and represented as a composite or
`
`complex waveform. (Ex. 2002, 207-208.) As shown below, adding the top and
`
`middle pure sine waves yields the complex composite bottom waveform:
`
`(Ex. 2004, 61; see also Ex. 2005, 84-86 (showing examples of composite
`
`waveforms); Ex. 2002, 208 (showing examples).) The more components a given
`
`sound includes, the more complex the composite waveform. Thus, much like white
`
`light can be broken down into component colors by refraction (i.e., a rainbow),
`
`complex sounds (such as music) can be broken down into component frequencies
`
`along with their specific amplitudes and phases (often by applying mathematical
`
`4
`
`

`

`operations such as Fourier analysis). (Ex. 2003, 448 (definition of “complex
`
`wave”).); Ex. 2001, 62 (describing the “famous theorem” of Fourier).)
`
`Timbre is the “color” of an instrument or voice. (Ex. 2003, 2152 (definition
`
`of “timbre”).) Most sounds consist of more than one frequency. Timbre depends on
`
`the sound’s waveform, which is determined by the presence of additional
`
`frequencies beyond the pure or fundamental tone (i.e., the tone with the lowest
`
`frequency in a series of related tones (see Ex. 2005, 79; Ex. 2002, 201)), such as
`
`overtones (generally, frequencies above the fundamental (see Ex. 2004, 63)) and
`
`harmonics (integer multiples of a fundamental (see Ex. 2003, 959 (definition of
`
`“harmonic”); Ex. 2004, 63)), as well as those additional components’ frequencies
`
`and relative intensities of those additional components. The presence of additional
`
`frequency components allows one to distinguish between sounds of the same pitch
`
`and loudness produced by two different instruments or voices. (Ex. 2002, 202.)
`
`Timbre is why a note played on a flute sounds distinct from the same note played
`
`on a violin. The example below of the note “A” played on a violin is not a pure
`
`sine wave because it is comprised of several parts, including a fundamental tone at
`
`5
`
`

`

`440 Hz, and overtones and harmonics of various frequencies and intensities
`
`produced by the vibrating string:
`
`(Ex. 2002, 255.)
`
`B.
`The ᾽961 Patent
`The ᾽961 patent relates to a method and system for “insertion of hidden data
`
`into audio signals and retrieval of such data from audio signals and is more
`
`particularly directed to such a system and method using a phase encoding scheme.”
`
`(Ex. 1001, 1:20-24.) The ’961 patent teaches changing a tone’s phase to embed
`
`hidden data into audio signals, and a method and system of extracting the
`
`embedded data by identifying the phase change. The specification explains:
`
`A watermark is data that is embedded in a media or document file that
`serves to identify the integrity, the origin or the intended recipient of
`the host data file. One attribute of watermarks is that they may be
`visible or invisible. A watermark also may be robust, fragile or semi-
`fragile. The data capacity of a watermark is a further attribute. Trade-
`offs among these three properties are possible and each type of
`watermark has its specific use. For example, robust watermarks are
`
`6
`
`

`

`useful for establishing ownership of data, whereas fragile watermarks
`are useful for verifying the authenticity of data.
`
`(Ex. 1001, 1:31-41.)
`
`Figure 1 is a “conceptual diagram illustrating the attributes of various data
`
`embedding techniques” (Ex. 1001, 4:52-53):
`
`Historical attempts to develop robust audio watermarking schemes have
`
`failed. (See Ex. 1001, 2:34-3:53 (discussing failed attempts by others).) The
`
`inventors recognized that “[n]aturally occurring audio signals such as music or
`
`voice contain a fundamental frequency and a spectrum of overtones with well-
`
`defined relative phases. When the phases of the overtones are modulated to create
`
`a composite waveform different from the original, the difference will not be easily
`
`detected. Thus, the manipulation of the phases of the harmonics in an overtone
`
`spectrum of voice or music may be exploited as a channel for the transmission of
`
`hidden data.” (Ex. 1001, 4:14-21.) The ᾽961 patent teaches a novel invention that
`
`7
`
`

`

`overcomes the disadvantages in the prior art, is “undetectable and robust to blind
`
`signal processing attacks,” is “uniquely robust to digital to analog conversion
`
`processing,” and “can be used to watermark movies by applying the watermark to
`
`the audio channel in such a way as to resist detection or tampering.” (Ex. 1001,
`
`4:32-37.) As discussed below, the ᾽961 patent describes two embodiments.
`
`1.
`First Preferred Embodiment
`The first preferred embodiment shows frequency components on the y-axis
`
`and time frames on the x-axis (denoted by vertical lines). In each time frame, ϕ0
`
`and ϕ1 denote pairs of frequency component partials.
`
`“[D]uring each time frame one selects a pair (or more) of frequency
`
`components of the spectrum and re-assigns their relative phases. The choice of
`
`spectral components and the selected phase shift can be chosen according to a
`
`pseudo-random sequence known only to the sender and receiver. To decode, one
`
`8
`
`

`

`must compute the phase of the spectrum and correlate it with the known
`
`pseudorandom carrier sequence.” (Ex. 1001, 5:22-30.)
`
`2.
`Second Preferred Embodiment
`The second preferred embodiment (called the Relative Phase Quantization
`
`Encoding Scheme or Quantization Index Modulation (QIM) Scheme), involves
`
`“comput[ing] the spectrum of a frame of audio data [i.e., breaking the audio signal
`
`into timeframes], then select[ing] an apparent fundamental tone and its series of
`
`overtones [within the timeframe]” with the suggestion that “it is convenient to
`
`select the strongest frequency component in the spectrum.” (Ex. 1001, 5:43-47.) As
`
`discussed above, a fundamental tone is the tone with the lowest frequency in a
`
`series of related tones. (Ex. 2005, 79; Ex. 2002, 201.) Overtones are components of
`
`a complex tone having higher pitch than the fundamental, while harmonics are
`
`integer multiples of the fundamental. (Ex. 2003, 959; Ex. 2004, 63.)
`
`After selecting the apparent fundamental tone, “two of the overtones in the
`
`selected series are ‘relative phase quantized’ according to one of two quantization
`
`scales.” (Ex. 1001, 5:47-49.) Quantization is the “restriction of a variable to a
`
`discrete number of possible values” or “restrict[ing] (a variable quantity) to
`
`discrete values rather than a continuous set of values.” (Ex. 2003, 1716 (definition
`
`of “quantization”); Ex. 2006, 1579 (definition of “quantize”).) The specification
`
`explains that, e.g., “[t]he choice of quantization levels indicates a ‘1’ or ‘0’ datum.
`
`9
`
`

`

`The relative phase-quantized spectrum is then inversely transformed to convert
`
`back to the time domain.” (Ex. 1001, 5:50-54.) In other words, the “relative
`
`phases” of one overtone to another are calculated, and a discrete value is assigned
`
`depending on the magnitude of the difference. Figure 4 depicts this process:
`
`The specification describes this embodiment as a four-step process:
`
` Step 1: segment audio signal into frames.
`
` Step 2: compute spectrum of each frame and calculate the phase of
`
`each frequency therein.
`
` Step 3: quantize the relative phases of two overtones in the selected
`
`frame to embed data, wherein the number of quantization levels is
`
`variable.
`
` Step 4: inverse transform the phase-quantized spectrum to convert
`
`back to the time representation of the signal by applying an inverse
`
`fast Fourier transform.
`
`(Ex. 1001, 5:55-6:21.)
`
`Claim 1 is a method of embedding data using the specification’s teachings.
`
`(Ex. 1001, 9:10-25.) An audio signal is divided into a plurality of timeframes,
`
`within each are a plurality of frequency components (indicated by the various
`
`waves in Figs. 1(a) and 1(b), below), and at least two of the frequency components
`
`10
`
`

`

`within the timeframe are selected. Next, the phase of at least one of the selected
`
`frequency components is altered. The following is a visualization of that process:
`
`
`
`
`
`Fig. 1(a) – Time frame with
`
`Fig. 1(b) – Time frame with
`
`fundamental tone (green box) and
`
`fundamental tone (green box) and
`
`overtones, including selected
`
`selected overtone shifted 180
`
`overtone (orange box).
`
`degrees out of phase (orange box).
`
`In this example, the fundamental tone (green) remains constant while the
`
`selected overtone (orange) is shifted 180 degrees (or π rad).
`
`Composite waveforms represent various simultaneous frequency
`
`components. Interestingly, while shifting one frequency component relative to a
`
`fundamental might alter the composite waveform (to account for the shift), it does
`
`not impact the ultimate sound perceived by the human ear. (Ex. 1001, 4:14-21.)
`
`Thus, using the process taught in the ᾽961 patent, phase shifts in at least one
`
`11
`
`

`

`overtone relative to a fundamental or any overtone in the selected harmonically
`
`related series within the selected time frame enable data encoding while remaining
`
`undetectable to the human ear.
`
`IV. THE ASSERTED PRIOR ART REFERENCES
`A.
`Srinivasan (Ex. 1005)
`Srinivasan discloses a method in which the amplitude or phase of a pair of
`
`spectral components of an audio signal are manipulated to encode data. The
`
`manipulations are made to randomly selected pairs of spectral frequencies (referred
`
`to as code frequencies f1 and f0) lying between 4.8 kHz to 6 kHz (i.e., 4800-6000
`
`Hz). This narrow band of high frequencies is employed specifically to avoid
`
`perceptibility of the modifications. (Ex. 1005, 7:64-67 (“The code frequencies fi
`
`used for coding a block may be chosen from the Fourier Transform ℑ{v(t)} at a
`
`step 46 in the 4.8 kHz to 6 kHz range in order to exploit the higher auditory
`
`12
`
`

`

`threshold in this band.”).) Figure 2 from Srinivasan depicts the steps performed
`
`by an encoder:
`
`Srinivasan describes two methods to randomize the selection of the specific
`
`code frequencies f1 and f0 that are manipulated at step 46. (Ex. 1005, 8:3-5.) In the
`
`first method, referred to as “Direct Sequence,” a pre-selected sequence of random
`
`numbers specifies which frequency components are manipulated using a “hopping
`
`algorithm employing a hop sequence Hs and a shift index Ishift.” (Ex. 1005, 8:10-
`
`11.) Srinivasan explains:
`
`if Ns bits are grouped together to form a pseudo-noise sequence, Hs is
`an ordered sequence of Ns numbers representing the frequency
`deviation relative to a predetermined reference index I5kꞏ For the case
`where Ns =7, a hop sequence Hs={2, 5, 1, 4, 3, 2, 5} and a shift index
`Ishift=5 could be used. In general, the indices for the Ns bits resulting
`from a hop sequence may be given by the following equations:
`
`13
`
`

`

`I1=I5K+Hs-Ishift
`
`(2)
`
` and
`
`(3)
`IOI5K+Hs+Ishift
`One possible choice for the reference frequency f5k is five kHz,
`corresponding to a predetermined reference index I5k=53. This value of
`f5k is chosen because it is above the average maximum sensitivity
`frequency of the human ear. When encoding a first block of the audio
`signal, I1 and IO for the first block are determined from equations (2)
`and (3) using a first of the hop sequence numbers; when encoding a
`second block of the audio signal, I1 and IO for the second block are
`determined from equations (2) and (3) using a second of the hop
`sequence numbers; and so on. For the fifth bit in the sequence
`{2,5,1,4,3,2,5}, for example, the hop sequence value is three and, using
`equations (2) and (3), produces an index I1=51 and an index IO=61 in
`the case where Ishift=5.
`(Ex. 1005, 8:12-38.)
`In the second method, the frequency index of the strongest low-frequency
`
`peak in the spectrum between 0-2 kHz, is included in a formula to specify which
`
`high-frequency components are modified. Specifically, Srinivasan explains:
`
`Another way of selecting the code frequencies at the step 46 is to
`determine a frequency index Imax at which the spectral power of the
`audio signal, as determined as the step 44, is a maximum in the low
`frequency band extending from zero Hz to two kHz. In other words,
`Imax is the index corresponding to the frequency having maximum
`power in the range of 0-2 kHz. It is useful to perform this calculation
`starting at index 1, because index 0 represents the “local” DC
`
`14
`
`

`

`component and may be modified by high pass filters used in
`compression. The code frequency indices I1 and IO are chosen relative
`to the frequency index Imax so that they lie in a higher frequency band
`at which the human ear is relatively less sensitive. Again, one possible
`choice for the reference frequency f5k is five kHz corresponding to a
`reference index I5k=53 such that I1 and IO are given by the following
`equations:
`
`I1=I5K+Imax-Ishift
`
`(5)
`
` and
`
`IO=I5K+ Imax + Ishift (6)
`where Ishift is a shift index, and where Imax varies according to the
`spectral power of the audio signal. An important observation here is
`that a different set of code frequency indices I1 and IO from input block
`to input block is selected for spectral modulation depending on the
`frequency index Imax of the corresponding input block. In this case, a
`code bit is coded as a single bit: however, the frequencies that are used
`to encode each bit hop from block to block.
`
`(Ex. 1005, 8:50-9:12.)
`
`In both methods, the spectral modifications are made to frequencies in the
`
`range 4.8 kHz to 6 kHz, and the fundamental tone and its overtones (including
`
`harmonics, which are overtones with frequencies that are integer multiples of the
`
`fundamental’s frequency) are not modified. (See id.)
`
`15
`
`

`

`B.
`Cabot (Ex. 1006)
`Cabot consists of a summary of the experimental results of a refined
`
`procedure developed to evaluate the “audibility of phase shifts in harmonically
`
`related tones” which had been “a topic of discussion for many years.” (Ex. 1006,
`
`568.) Prior “crude experiments” had been conducted by other researchers to “show
`
`that phase shifts were not audible.” (Id.) Cabot observed:
`
`The testing methods used by previous experimenters do not appear to
`give the listener the most convenient method of comparison. We felt it
`would be more informative to allow the listener complete control over
`all timing of signal presentations. In our approach the listener controlled
`the audition intervals of both the reference and comparison signals.
`Careful note was taken of the length and number of audition intervals
`chosen by listeners, and some interesting trends were noted.
`
`(Id.)
`
`Cabot’s “experiment shows phase shifts of harmonic complexes to be
`
`detectable, but judging from the difficulty experienced by the subjects, the effect
`
`appears to be small.” (Id., 570.) More specifically, the experimental evidence
`
`showed that making relative phase shifts greater than 30 degrees in a 2-component
`
`composite signal consisting of a fundamental and third harmonic was audible at
`
`least 75% of the time. (Id. (“it was found that a 30° shift was still fairly well
`
`recognized.”).) Below is a table of test data and a graph summarizing Cabot’s
`
`experimental results:
`
`16
`
`

`

`Table Il. Test data.
`
`Phase
`Shift 6
`(degrees)
`
`Dectection
`Rate Py
`(%)
`
`Alarm
`Rate Py,
`(%)
`
`Corrected
`Rate P
`(%)
`
`23
`33
`
`Experiment 2, 5 Subjects
`0
`55
`7.5
`53
`Ba
`80
`22.5
`
`(Ex. 1006, Table II.)
`
`Experiment 1, 15 Subjects
`0
`23
`30
`83
`60
`92
`90
`85
`120
`92
`
`difference.
`
`75
`
`#15
`
`225
`
`30
`
`60
`PHASE DEGREES
`
`Fig. 3. Computed values of the probability of a perceived
`
`(Ex. 1006, Fig. 3.)
`(Ex. 1006, Fig. 3.)
`
`17
`17
`
`

`

`C. Kudumakis (Ex. 1007)
`Kudumakis describes the use of notch frequencies, which involves filtering
`
`out or removing part of the original audio and replacing and inserting data (a code)
`
`in the removed portions:
`
`Two notches are inserted in the audio band to provide frequencies at
`which the code may be inserted. The code signal is inserted as a series
`of pulses at the center frequencies of the notches, and insertion is
`initiated when the program content provides sufficient masking
`conditions for the code to be inserted inaudibly.
`(Ex. 1007, 1:17-22.)
`
`In Kudumakis, the method described makes modifications to the host signal
`
`nearby, but not directly to, the strongest frequency components of the host signal,
`
`such as its fundamental and harmonics:
`
`The codes are more perceptible if the notch frequencies coincide with
`the main frequency component of the signal. On the other hand, they
`have to be placed in a part of the spectrum with sufficient energy so that
`frequent [sic] masking conditions can be met.
`
`(Ex. 1007, 3:4-8.) The Kudumakis method relies on the phenomenon of auditory
`
`masking. More specifically, there is a band of frequencies near a strong peak in a
`
`spectrum that is “masked” by the presence of a strong signal. The threshold for
`
`hearing spectral components near a strong peak goes up, i.e., one cannot hear weak
`
`frequency components near a strong peak. This auditory phenomenon allows one
`
`18
`
`

`

`to alter the ampli