Apple Computer Inc. v. Burst.com, Inc.

Filing 157

Declaration of Nicholas A. Brown in Support of 156 Reply Memorandum filed byApple Inc.. (Attachments: # 1 Exhibit A part 1# 2 Exhibit A part 2# 3 Exhibit A part 3# 4 Exhibit A part 4# 5 Exhibit A part 5# 6 Exhibit B part 1# 7 Exhibit B part 2# 8 Exhibit B part 3# 9 Exhibit C# 10 Exhibit D# 11 Exhibit E# 12 Exhibit F# 13 Exhibit G# 14 Exhibit H# 15 Exhibit I# 16 Exhibit J# 17 Exhibit K# 18 Exhibit L# 19 Exhibit M# 20 Exhibit N# 21 Exhibit O)(Related document(s) 156 ) (Brown, Nicholas) (Filed on 9/6/2007) Modified on 9/18/2007 (gba, COURT STAFF).

1 1 1 1

I Y F O A B

12 1 1

w

1
Page 1 of 18

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007
9/4/07

Sheila Hememi

37 (Pages 142 to 145 )
Page 14 2 23956 1 124002 2 124008 3 124012 4 1240 :15 5 124016 6 124020 7 24021 8 124025 9 2402710 2403011 12403612 12403613 12404314 12405015 12405616 12410017 12410318 12410619 2410720 1241082 1 12411022 12411223 12411524 12411925

Page 12 : 4 2 :5 5 1 12 : 4 2 : 5 9 2
43 :01 3

144

but it's not akin to what we would consider to be the computational heart of either a - a personal computer or a -a higher-end scientific computation work station. Q. ut it -- it doesn't say that it's a microcontroller,
right?

24303 4 24307 5
24307 6

A. ll it tells us is that it is host computer so -Q . kay . And you can't say from that description whether or not it's capable of performing the video compression described, correct? A. rom - from what's described here, I do not know, really, w=hat the computational horsepower of the host computer is . Q . kay . Is it your -- okay . Never mind. Now, again, looking at Figure 1, I think your declaration says that the memory modules, 24, 26, 28, et cetera, are removable ; is that right? A. eah. And if I can just take a small -- I just want to -- I found a typo yesterday, and I would like to just clarify that . Q. Okay . A . was a little bit overeager in my use of the term "memory module."
Q. Which declaration are you referring to?

243 :11 7

43 :14 8 43 :15 9
12 :43 :171 0 12 :43 :201 1 2 :43 :221 2 ;12 :43 :261 3 12 :43 :281 4 (12 :43 :291 5 12 :43 :331 6 12 :43 :351 7 112 :43 :381 8 '12 :43 :431 9 112 :43 :462 0 12 :43 :472 1 ;12 :43 :472 2 ;12 :43 :502 3 ;12 :43 :552 4

A. This is the second one. Which paragraph?
Page 143

12 :4 3 : 5 8 2 5

I'm a little uncomfortable making a definitive statement as to what RAID does or doesn't do . Q . Okay . So, again, though, the question was : The memory modules are removable, you say, right, in the central data station? A . Sorry, I'm still back on RAID. Let me say, I'm definitely not sure what RAID does . Let's just leave it at that. Now, I'm sorry, if you could repeat your question . Q. Okay. But RAID uses that term "striping" ? A. I -- I don't remember. So, I was trying to sort out hat did I remember, what do I not remember . And what I remember about RAID is often having an error-correcting disk . Q. Where did you come with you the word "striping" in your declaration? A. This is a term I'm familiar with from my history as an electrical engineer. I don't think I drew on any - you know, at some point in my life, I learned and understood what striping data across multiple storage entities, for lack of a better word, meant . And -Q . And that's where you borrowed it ? A . -- that's where it came from, yes . Now, that - what did I - that may well have come from my interaction with RAID at some point in my career . I just don't remember its origin . Q . Okay. So, again, memory modules in the central dat a station are removable, right ?
Page 14 5

2 :41 :19 1 12 :4126 2 124128 3 124133 4 24142 5 124150 6 124155 7 124200 8` 124205 9 2420710 12420811 12421012 12421213 12421214 12421515 12421516 12421817 12422218 12422519 2422820 2423421 12423722 12424323 12424624 12425125

A . And, yeah, let me find that . So, I'll point out that 244 :01 Item 39 correctly states that Walter's central data station 244 :01 stores a single/program in each memory module and -- okay . 124404 So, in Item 45 on page 13, I've used the term "memory modules" 124406 124408 in Item 45 ; and what I should have said was "data cells ." So, 12 :44 :12 that's both -- in the first sentence of-- of Item 45 shoul d 12 :44 :16 read "rather uses multiple data cells to store a single program ." And at the very top of -- yeah, you see that, 12 :44 :19 too -12 :44 :22 9 Q. I do . 2 :44 :2810 A. - of-- of page 14, that's propagated through the -244 :3611 through that item. 12444012 Q. Okay. 12444013 A. So, I apologize . I just want to make that point 12444314 so-2444515 12 :44 :5116 Q. So, while we're there with the "striped," that 12445317 striping is sort of like you would do in a RAID array? Is that the wav you understand it? 12 :445618 A. Yeah, yeah . At some level - I mean, let me say, I'm 1245001 9 not -- you know, I'm familiar with RAID peripherally from a 12 :4s :o12 o long time ago ; but the general concept of rather than stacking 12450221 up the data sequentially in a single storage, we hav e 12 :45 :0522 essentially hopping from -- from device to device . Now, let :12 :45 :0823 me add the caveat that -- that I believe that's as done i n 12 :45 :1124 RAID ; but having not done anything with RAID since about 1994, 12 :4 5 : 11 2 5

A . Yes. Q . And -- I mean, at least that's what you say . I don't see that anywhere in the patent . Is that fair to say? A . Yes . And you - you clarified that . And, yes, let me follow that up . But my -- my reading of the patent with the extensive use of the term "preprogrammed," as we have discussed, says to me that these are -- these are removable modules. And let me just add that, also, given that Walter doesn't teach any mechanism by which the central data station would receive content, that I -- I took as support of the fact that these modules themselves could be removed and -- and programmed . Q. So, you -- from the word "preprogrammed" and from the fact that it doesn't describe how you load the information into the modules, you took it to mean that you received them by FedEX or something like that? A . That would be one option. They -- you know, courier, somebody could walk them . We don't know where the compression occurs. Q . Okay . A . I am very confident that the compression doesn't occur here . But other than saying it doesn't occur in th e central data station, I -- I don't know where it occurs . Q. Okay. A. So, FedEx may not be necessary .

Merrill Legal Solutions w ww .merrillcorp .co m -888-513-980 0

1 1 1

dI Q A

A I

' :1 ~1

1

1
Filed 09/06/2007
9/4/0'7 38 (Pages 146 to 149 )

Case 3:06-cv-00019-MHP

Document 157-9

Page 2 of 18

Sheila Hememi
Page 14 6
24517 1 124520 2 124523 3 124527 4 1245 :34 5 1245 :39 6 124545 7 124554 8 124557 9 2460110 2460711 12461012 12461413 12461714 12462015 12462516 12462917 1246 :3218 12463719 2464020 2464621 12465322 12465623 12 :46 :5924 12 :47 :052 5

Page 14 8

2 . Okay. Now, is there any reason why "preprogram" oesn't simply mean programming before you make the content i 12 2 available for download? 2 . In the context of the Walter patent, given that there 2 was no mechanism described to get the information on to the 2 central data station or no video jukebox or anything of the 2 sort, that is how I interpreted that "preprogrammed" in this 2 case, in the specification . The way it gets used implies that 2 these are removable . And, you know, I think that -- the way I looked at this was -- "preprogrammed" is -- it's quite a verb . Contrast it with "stored ." All right . From the standpoint, I think, of describing the operation of the Walter unit to somebody sitting on the bus next to me, I would say, "Well, see these things? They store the video ." But, in fact, Walter does not use that term . Walter specifically uses "preprogrammed" over and over to describe those modules . Q . Well, and he might have been using that to indicate that they're dedicated to a particular program, right ? A . Well, from the standpoint that he doesn't give us any mechanism to alter them using a central data station and he also tells us that a single module has a single program, I think that we could call that dedicated. Q . Okay . Well, he does, in fact, give us a way to -- to alter the memory modules in the data receiving station, right? A. The -- the memo ry modules in the data receiving Page 14 7

:48 :48 :48 :48 :48 :48 :48 :48 :48

:10 : 11 :13 : 16 : 21 :25 :29 :35 :40

1 2

2 :48 :4410 2 :48 :4811 12 :48 :531 2 (12 :48 :5513 ;12 :49 :0414 !12 :49 :071 5 12 :49 :1116 '12 :49 :1717 1249211 8 :12 :49 :2519 12 :49 :2720 12 :49 :3121 ;12 :49 :3322 :12 :49 :382 3 12 :49 :4024 12 :49 :412 5

Q. Okay . A. Now, you know, I would like to add that the data receiving station -- again, we'll colloquially use a teen for today, a "set top box" -- this is a piece of consumer electronics equipment . One can reasonably anticipate that this piece of equipment is going to sit in some reasonably sized cabinet and that the -- the operator, the customer, is not going to be taking pieces in and out . But I suspect that as today, the -- the companies get somewhat upset if you try to mess around with the inside of your box . On the other hand, the central data station is at the cable head end or certainly at a location which is operated by presumably the cable company; and they have an interest, I think, in -- in providing more than one set of programs for the rest of eternity . So, from the standpoint of these things being preprogrammed elsewhere, this is the mechanism by which they can change their programming. Otherwise we have sort of a video-on-demand system which is a video-to-exhaust system. Q . Is -- is there any engineering obstacle or anything unpredictable about making a memory module in the receiving station removable? A . I'm sorry, the second part was unpredictable . The first part was -Q . Let me just use the word "unpredictable" instead . Page 149

24708 1 124710 2 124712 3 124716 4 124719 5 124723 6 124723 7 124725 8 124728 9 2473110 2473211 12473612 12 :47 :3913 12474014 12474415 12474716 12474917 12475018 12475419 12 :47 :5520 2475821 12480222 12480423 12480624 12481025

station are the local storage . So, certainly. Q . nd it's your opinion, I think you said in your declaration, that that memory module is of the same type that's in the central data station, right? A . believe that the Walter patent explicitly makes
that statement .

124 9 12 : 4 9 = 12 : 4 9 124 9 12 : 4 9

:4 4 :48 :52 :55 :55

1 2 3 4 5
6

12 : 4 9 : 5 9

125 0 Q . Okay . Is the memory module in the data receiving 125 0 station also removable? 12 : 5 0 A . I do not expect that the memory module in the data 12 : 5 0 receiving station is removable. 12 : 5 0 Q . But the patent doesn't say one way or another about 12 :5 0 any of the memory modules, right, being removable? s 12 :5 0 MR . PAYNE : Objection, form . 2 :50 A. Well, again, upon reading the specification, looking 2 :50 at the preprogramming, the extensive use of the 2 :5 0 "preprogramming" term -2 :50 Q. (By Mr. Stephens) Let me just be clear. I'm not -'m not asking whether or not you concluded based on what it 12 :50 2 :50 does say, that they're removable . I understand you did. I'm 2 :50 asking about whether it says they're removable anywhere . 2 :50 A. There is no sentence in the patent that says this 2 :50 memory module is removable . 2 :5 0 Q. Okay . And that's true for both the data receiving 2 :50 stati on and the central data stati on, ri ght? A.
It

:00 7 :03 8 :07 9 : 111 0 41 1 :17 1 2 : 2 013 :2114 : 2 515 : 3 01 6 :3 117 :3 418 : 3 619 :3720 : 4 0 21 :4022 :4323 :4 9 24

So, is there anything unpredictable about changing the data receiving station to make the memory module there removable as you say the ones in the central data station are ? MR . PAYNE : Objection, form . A . I'm sorry, I'm still getting a little hung up on this question. Can I ask you to repeat it one more time? Q . Yeah. Is there any reason why you would not be able to or would expect any unusual engineering difficulties that would make it unpredictable or lead to an unpredictable result if you were to modify the data receiving station to make the memory module removable? MR . PAYNE : Objection, form. A . I don't know what unpredictable mean -- can -- can -let me ask you this. And I'm not trying to be difficult. Maybe it's just the metabolism is kicking in after lunch. Can I ask you to start the question again an d pause before you get to the predictable part because I digest the question, and then all of a sudden we're on predictabl e and I don't know what predictable means and I get hung up and then I've forgotten what the beginning part of the questio n is . So -Q . (By Mr. Stephens) Okay . Could you modify the data receiving station to make the memory module removable? MR . PAYNE : Objection, form . A. So, I don't know. I mean, for one, I don't know why

is .

12 :50 :5025

Merrill Legal Solutions -888-513-980 0 www .merrillcorp .co m

sQ A Q A

M S

`s. 1 =1 % j' 1 11 1

1

w

A

A S I

1
Page 3 of 18

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007
9/4/07

Sheila Hemem i

39 (Pages 150 to 153 )
Page 12 :50 12 : 50 12 : 5 0 12 : 5 0 12 : 5 0 12 : 51 1251 12 :51 12 : 51 12 : 51 12 : 51 12 : 51 12 : 51 12 : 51 150 ; 12 12 12 12 12 12 12 12 12 12 12 12 12 12
:53 :24 5326 2 5329 3 5331 4 5335 5 5338 6 5341 7 5341 8 5342 9 534310 534511 534712 534813 Page 15 2

:52 1 anybody would want to do such a thing. : 54 2 . (By Mr. Stephens) Well, we'll get to that in a :553 econd ; but if you wanted to, is there any particular : 5 7 4 engineering difficulty in doing that? :595 . I don't know. I don't know. That depends on how the : 02 6 system is designed. :0 5 7 . Okay . Well, if it was designed so that they fit in :10 8 the same kinds of sockets that they did in the central data :11 9 station, presumably it would be doable, right ? :14 10 . Well, actually, we already have a very differen t :17 11 setup from simply looking at the memory modules . I mean, I -: 2 212 1 believe -- and I think I said this and I hope I'm no t : 2 513 misspeaking when I say that Walter tells us that the memor y : 2 814 module is of the same type, but the memory module as shown in
Figure I in 102 has both inputs and an output and also a li ttle computer control input . The memory modules in the central data station that we see in the upper left-hand corner have one - single arrows coming out of them. We don't see input data going into those . o, this already, even though the memory module

: 5 3 : 5 0 14

1251 :3 315 12514 016 12 : 51 : 4 517 12 : 51 : 4 918 12 : 51 : 5 519 12 : 51 : 5 6 2 0

12 :53 :5115 12 :53 :5416 2 :5 3 : 5 9 17 12 :53 :S918 2 : 5 4 : 0219 12 : 5 4 : 0 6 2 0 : 12 : 54 : 10 21 2 :5 4 :112 2 12 54 : 1123 12 :54 :132 4

12 : 51 : 5 9 21 itself may be of the same type, says to me that these are 12 : 5 2 : 0 3 2 2 integrated into their respective positions in a different 12521023 125 210 2 4 manner. . Do you have any reason to believe that it would be

12 : 5 2 :14 2 5 anything other than a straightforward exercise to make the
Page 15 1 12 : 52 : 16 1 memory modules the data receiving station removable? 12 :52 :21 2 12 : 5 2 : 24 3 . What is a straightforward exercise? . Simple matter of ordinary day-to-day engineering

2 :5 4 :13 2 5

Q . 'm not asking -- I'm specifically not asking about motivation . I'm asking about doing it. If you're instructe d to do it, all you have to do is say: I'm going to put a socket on there instead of soldering it to the board, ri ght? MR. PAYNE : Objection, form. A. don't even know if a socket exists for these things . Q. (By Mr . Stephens) Do you know what they were ? A. o, I -- that is a potential -Q. Do you know what these memory chips are? A. nd we are told that they are recirculating shif t registers . Q. nd what is that? Is that a chip that you could g o out and buy? A . Well, Walter has not given us a specific part ; and I do not know if there were integrated units of these things available. Q . Okay . But you could fabricate one using DRAM, correct? A . No . I don't think one would fabricate one using DRAM. Q . I'm not asking whether you would. A . No . I don't -Q . I'm saying you could. . No .
Page 15 3

2 :5 4 :14 1 2 :54 :15 2 2 :54 :2 3 3 2 :5 4 : 2 7 4 2 :54 :3 0 5 2 : 54 :3 1 6 25 4 : 3 2 7 254 :34 8 2 : 54 : 3 4 9 2 : 54 : 4 010 2 : 54 : 4111 12 : 54 : 4 412 12 : 54 : 4 713 12 : 5 4 : 5 2 14 2 : 54 : 54 15 2 :54 :5416 2 : 54 : 5 717 2 : S 5 :0118 2 : 5 5 : 0 519 2 :5 5 :10 2 0 2 : 5 5 :13 21

12 : 5 2 : 2 6 4 judgment, not involving the exercises of unusual creativity . 12 :52 :31 5 . I think it would involve the exercise of unusual 12 :5 2 :3 5 6 creativity to make the memory module in the data receiving 12 : 52 :3 9 7 station removable. 125241 8 12 : 5 2 : 4 2 9 12 : 5 2 :4 310 12 : 52 : 4511 . I'm not -. I see absolutely no reason -. Again, I'm specifically asking you not to answer about motivation. I'm asking you, if you were asked to do so,

12 : 5 2 : 4 812 if you were asked to make it removable, do you have any reason 12 : 5 2 : 52 13 to believe that it would require anything other than ordinary

12 : 5 2 : 5 514 day-to-day engineering activities to do that? 12 : 5 3 : 0 015 12 : 53 : 0216 12 : 5 3 :0 217 12 : 5 3 : 0 519 . What are -R . PAYNE: Objection, form . . -- ordinary engineering day-to-day activities? I

12 : 5 3 : 0 518 mean, I can envision -(By Mr . Stephens) Saying: I'm going to put a socket 12 : 5 3 : 0 7 2 0 on this board instead of soldering it to the board. Is 12 : 5 3 : 16 21 anything else required? 12 : 5 3 : 18 2 2 12532023 12 : 5 3 : 2 2 24 12 : 5 3 : 24 2 5

. One needs a socket, one needs to understand why would ' 12 : 5 5 :18 2 2 one want to -. I'm not asking about motivation. . -- increase the cost, increase the complexity -255 :1923 2 :55 :2624 2 :5 5 : 2 6 2 5

Q . Why not? A. Because DRAM is a random access memory technique -technology. As we discussed or as I mentioned specifically hen I was talking about semiconductor RAM, each individual piece -Q. Okay . Go ahead . A. -- location needs to be addressed. Q. Okay . A. Now, if I start off with a DRAM and 1 want to make this a recirculating shift register -Q. No, I'm not asking you to change the DRAM into a recirculating shift register . I'm saying it's pretty simpl e to design a circuit that functions as a recirculating shift register using DRAM as the storage component, correct? A . I disagree that it's simple . Q . Well, let me -- let me give you a very specific example . So, if you create a address generator that simply makes the address change by one location with every clock pulse that you apply and use that to change the place where memory is stored and retrieved from, you can generate - you can turn a DRAM into a recirculating shift register in function by doing that, right ? A. You used the word "simply," and I completely disagree that this is a simple operation . Q. Okay . I --

Merrill Legal Solutions w ww .merrillcorp .co m -888-513-980 0

1

WA tB Q dA oA Q Q

M

'1 1 =1 11 .1

aQO I

1
Page 4 of 18

t

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007
9/4/07

Sheila Hememi
Page
12 : 5 5 : 2 6 1 12 : 5 5 :2 8 2 12 : 5 5 :3 6 4 12 : 5 5 : 4 0 5 12 : 5 5 : 4 3 6

40 (Pages 154 to 157 )
154 '
2 : 5 7 :13 2 :5 7 :18 2 :57 :2 3 2 :57 :26 2 :57 :30 2 :57 :32 2 :57 :35 2 :57 :36

Page 156 long -- long answers to simple questions, if you can avoid it . I - I'm not -- I'm not questioning whether you're intending to do it the way it's done. I'm not saying that because I don't believe that you are . I'm simply saying we don't have much time and I have a lot of material to cover so I'm asking you, please, to answer the question I'm asking and not give me long answers. A . Just so you know, as I told Les, this is the relaxing part the day because I have to fly later. So . . . Q. kay. I fly a hundred thousand miles a year . I can definitely relate to that . Well, no, I've flown a hundred thousand miles -A. I'm sorry .

. For one, one of ordinary skill would never, would not o this . This is starting off with a system that is overly hat sits on top of it or in conjunction with it . . Is a counter -- is a counter a complex circuit? . Given that the recirculating shift register does not

12 :55 :32 3 complex for what is needed and then having to design logic

12 : 5 5 : 4 6 7 require the random access, the DRAM already has all of the 12 : 5 5 : 5 0 8 write and read lines and apparatus on it to do the access

12 : 5 5 : 5 6 9
2560210 2560211 12560212 12560613 12560914 2560915 12561116 12561417 12561618 12561919 2562020 2562321 12563022 12563023 12563324 12563425

2 :5 7 : 3 8 e're essentially, then, designing more circuitry to bypass 12 :57 :4210 that -'12 :57 :4611 Q . Look -12 :57 :4812 A . -- and on top of it, then, I have to actually route '.12 :57 :4813 the output of my DRAM to the input. 12 :57 :4914 Q . Okay . Sorry to interrupt -;12 :57 :5115 A . Now I have capacitance issues . 112 :57 :5216 Q . You keep changing my question. I'm not asking you 12 :57 :5517 about motivation . I'm asking you about whether it's something 12 :57 :5618 that's easy to do if you're instructed to do it . A . I don't think it's easy. Q . So, it's not easy to make a counter that would increment the address with each clock pulse ; is that right? Is that your testimony ? A . I'm starting off with a recirculating shift register . Q . No . I'm not. I'm starting out - I'm asking you whether -Page 15 5
12 :58 :0019

;12 :58 :0420
12 :58 :1021

Q . -- as of last month. A. I'm sorry . Q. kay. So, is it a difficult thing to design a counter? A. n the abstract, designing a counter -- well, i t depends who you're speaking to . To one of ordinary skill as described in my various reports, I think designing a counte r is easy . Q . Okay. And you could use a counter to make a DRAM function as a recirculating shift register from the outside, right? In other words, it would look like a recirculatin g shift register to the circuits that interface with it; is that
Page 15 7

:12 :58 :1222 :12 :58 :1523 12 :58 :2024 12 :58 :2625

25 8 : 2 9 1 12 :5 6 : 34 1 . No, no, no . I'm going to compare to that, okay? 25 8 : 3 0 2 12 : 5 6 : 3 6 2 ear with me . I have a recirculating shift register -12 : 5 8 : 3 3 3 12 : 56 : 3 9 3 . I would bear with you, but your counsel has severely 25 8 : 3 9 4 12 : 5 6 : 4 2 4 limited the amount of time we have . So, I can't allow you to 25 8 : 4 2 5 12 : 5 6 : 4 5 5 go off on these lectures . 2 :58 :476 12 : 5 6 : 4 6 6 R. PAYNE : What do you mean I've limited? R. STEPHENS : We're going to stop at 3 :00 25 851 7 12 : 5 6 : 4 9 7 12 : 5 6 : 5 0 8 'clock, right? 2 :58 :558 25 8 : 5 8 9 12 : 5 6 : 51 9 R. PAYNE : Yes . 12 : 5 6 : 5110 R. STEPHENS : Okay . 2 :59 :0710 12 :59 :1111 12 : 56 : 5111 R. PAYNE : Right. 2 :59 :19 12 12 : 56 : 5212 R. STEPHENS : That's severely limiting 2 : 5 9 : 2 6 13 12 :56 :5 313 R. PAYNE : Well, wait a second . You 2 :5 9 : 2 9 14 12 :56 :55 14 specifically agreed to that 25 9 :3 2 15 12 :56 :571S R. STEPHENS : Right. R. PAYNE : Had you wanted more time, you should ` 12 : S 9 : 3 3 16 12 :56 :5 816 12 : 5 6 : 5 917 have told me. 2 : 5 9 :3 3 17 12 :5 9 : 3 5 18 12 :57 : 0 218 R. STEPHENS : I'm going to ask for more time 12 : 5 9 : 3 9 19 12 : 5 7 : 0 319 because you instructed her not to answer a key question . 25 9 : 4 0 2 0 12 :57 :0 5 2 0 R. PAYNE : One question, Garland? Give me a 2 : 5 9 : 41 21 12 : 5 7 : 0 7 21 break, please . 2 : 5 9 : 4 4 22 12 : 5 7 : 0 8 2 2 R. STEPHENS : A key question . 2 :5 9 : 4 9 2 3 12 :57 :0 923 R. PAYNE : Please. Garland 12 : 5 9 : 5 0 2 4 12 : 5 7 : 0 9 2 4 . (By Mr. Stephens) Okay. Well, whatever. We don't 12 : 59 : S3 25 12 :57 :102S have much time . So, I'm going to ask you not to give me

right? A. I would say that while one of ordinary skill woul d not do that -- and I'll cease the pontificating for the rest of it -- certainly we could put a recirculating shift register in a black box and have another black box with DRAM and ssociated circuitry and wiring and potential other stuff o deal with excess capacitance . And to the outside observer, the inputs and the outputs would be identical . Q . Okay. Now, if you were tasked with replacing the -replacing FedEx for distributing video to central data stations, you could simply use the system described in Figure 1, right? In other words, you could -- at the place you compressed the video -A. I have to ask you to start over . I'm completely lost. Q . Okay. A. So, first off, let me say that 1 didn't say that FedEx was - I said that was one option. Q . I think you said courier or -- well -A . Courier could be foot or bicycle, too . . Okay. So, if you were out to replace the manual delivery of the chips, you could use the system described i n Figure I to do that, right? A . I do not see how that would happen . I following you .

Merrill Legal Solutions -888-513-980 0 www .merrillcorp .co m

1 1

bQ fA A Q Q A Q A

#3 1 : 0 10 10

cA fQ Q A Q Q Q A Q A

1

P

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007

Page 5 of 18

Sheila Hememi - 9/4/0 7
41 (Pages 158 to 161 )
Page 158! age 16 0

12 :5 9 : 53 1 . Well, if my point is to get video from Point A to 12 :59 :S7 2 Point B, I can use the system in Figure I to do that, right? 13 : 0 0 : 04 3 . A courier could deliver video from - - let's just 13 : 0 0 : 13 4 stick to, you know, Figure 1 . So, suppose that we have some 13 : 0 0 : 15 5 other location which either requires a FedEx courier or some 13 : 0 0 : 19 6 type of mechanical transport. To use the system in Figure I 13 : 0 0 : 2 9 7 would require that first the central data station hav e 13 : 0 0 : 3 7 8 receiving capabilities which it doesn't have ; and secondly , 13 : 0 0 : 4 4 9 the origin, the location of the compression would need to have 13 : 0 0 : 5 310 transmission facilities . And we don't know anything about the 13 : 0 0 : 5 911 location that does the compression . So, even writing tha t 13 : 01 : 02 12 off, our central data station does not have receptio n 13 : 01 : 0 613 capabilities . We're not -- that's lacking as described in the 13011114 specification . 13 : 011115 . Okay. So, I just want to go back to the courier for 13 : 01 1516 a moment. So, the manual delivery of the chips that you 13 : 01 1917 described was intended to -13 : 01 : 2 318 . Module. 13 : 01 : 2 319 . I'm sorry, modules, memory modules -- was intended to 13 : 01 : 2 7 2 0 transport compressed video from the place it's compressed to 13 : 01 : 3 0 21 the central data station, right? 13 : 01 : 3 3 2 2 . Yes . We could call the compression location where 13 : 01 : 3 6 2 3 that preprogramming occurs, yes . 13 : 01 : 3 9 24 . Okay . So, the reason for using that courier is to 13 : 01 : 4 2 2 5 move the video, the compressed video, from the place where
Page 159

3 : 0 3 :16 1 rom another level of servers . 3 : 18 2 . Look, I'm not asking about that. I'm simply askin g 0 3 : 2 0 3 about whether you could use the system designed in Figure 1 to 0 3 : 2 5 4 transport video from the place where compression occurs to the 03 :285 entral data station ? 0 3 : 31 6 . I -- I don't know that the place where compress on 13 : 0 3 : 3 6 7 occurs would have such a system . 13 : 0 3 :3 8 8 . But if they didn't, this is one way to solve the 0 3 : 41 9 problem of getting video from the central data -- from th e 3 : 0 3 : 4 410 place where compression occurs to the central data station ,
034811 right?

3 : 4 812 . This is a video-on- demand cable system , 13 : 0 3 :5 513 video-on-demand system for use -- it's called a cable 13 : 0 3 : 5 814 provider. The problem that this addresses is very, very 130 4 : 0 315 different from the problem of a location needing to transfer 13 : 04 :1216 or transmit all of what it's producing to a single location . 3 : 04 :1817 . But it could be used for that, right? And, in fact , 04 : 2 518 all that's shown in Figure 1 is transporting it to a single 04 :3 0 19 location, right? 3 : 04 :34 2 0 . All that's shown as opposed to 3 :0 4 : 3 7 21 . Well, there 's only one receiving station shown in 13 : 04 4022 Figure 1, right? 3 : 04 : 412 3 . So, you're contrasting that with having multiple 13 : 04 : 4 3 2 4 receiving stations ?
3044425 . Yeah .

Page 16 1

3 : 01 : 4 6 1 compression occurs to the central data station where it would 13 : 0 4 : 4 6 1 . I guess I don't see how that's relevant to -Ol:482 3 : 04 : 4 8 2 e installed in the system, right? . Well, you said it's -- it's designed to solve a 01 :4 9 3 . It has to get there somehow . 3 : 04 : 5 2 3 different problem, but what's shown in Figure I is :01 :S2 4 . Okay . So, given that problem of moving video from < 13 : 04 : 5 5 4 transporting video from one location to one other location , 01 : 5 6 5 one place to another, a possible solution to that problem is 13 : 04 : 5 7 5 right? 3 : 0 2 : 0 2 6 the system shown in Figure 1, right? It does move video from . 13 : 0 5 : 01 6 . Walter teaches us a system for use in a 13 : 02 : 06 7 the central data station to the data receiving station, right? 13 : 0 5 : 04 7 video -on-demand scenario . The point of, for example, the 13 : 0 2 : 12 8 . The latter half of your question, it does move video 13 : 0 5 : 0 8 8 call-back line is so that different users can request whatever 13 : 02 : 16 9 : 0 5 :17 9 video they want, that being the purpose of video-on-demand . rom the central data station to the data receiving stations, 13 : 02 : 1910 correct. Maybe I can ask you to clarify a li tt le bit exactly 13 : 0 5 : 2 010 You watch Shrek, and I watch Monsters, Inc . To say that this 13 : 0 2 : 2 2 11 how that would be used in conjunction with the compression 13 : 0 5 : 2 611 is simply a point-to-point delivery system because Figure 1 13 : 0 2 : 2 5 12 location . 3 : 0 5 : 3112 only shows one data receiving station, I think, is a 13 : 0 2 : 2 613 : 0 5 :3 4 13 misrepresentation of what is -- what this patent addresses . . Okay . So, if you were to locate the apparatus we see 13 : 0 2 : 3 0 14 at the central data sta tion, a copy of it, at the place where 3 : 0 5 : 3 914 . I'm not asking about whether that -- that it' s 13 : 0 2 : 3 3 15 you did the compression and put a copy of the data receiving 13 : 0 5 : 4 215 limited to a point-to-point transmission system . It could be 13 : 02 : 3 7 16 stati on at the central data station, you could then replace 0 5 : 4 4 16 used for a point-to-point transmission system, correct? 13 : 02 : 4117 the courier by transmi tting the compressed video from the 5 : 5 3 17 . It's extreme overkill, at best, for a point-to-point 13 : 0 2 : 4 718 place where the compression occurred to the data receiving 055618 transmission system. 13 : 02 : 4 9 19 station, right? 5 : 5 8 19 . Okay. Well, it could also be used for broade r 13 : 0 2 : 5 0 2 0 0 6 : 0 0 2 0 distribution, right? And I think that's your point, that i t . So, effectively what you've described is almost a 13 : 0 2 : 5 3 21 tiered system . 13 : 0 6 : 0 3 21 is, in fact, designed for broader distribution of compressed 13 : 02 : 55 22 13 : 0 6 : 0 8 2 2 video, right? . Yes . 3 :02 :5523 . A second tier system . So, I - I worked on video on 113 :06 :0923 . What's described is a central data station that can 1 :03 :0124 demand in 1994; and there certainly was no discussion of 3 : 0 6 : 11 2 4 serve a variety of users. This is not the same scenario as a 0 3 : 0 9 2 5 having the video server itself be a receiving station as well 3 : 0 6 :18 2 5 single compression location needing to transmit thi s

Merrill Legal Solutions www .merrillcorp .co m -888-513-9800

1

cA id Q A Q A Q

W

1 :0 : :

Ia Q Q A Q A A Q

M

1

P

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007

Page 6 of 18

Sheila Hememi - 9/4/0 7
42 (Pages 162 to 165 )
Page 16 2 13 : 0 6 13 : 0 6 13 : 0 6 13 : 0 6 13 : 0 6 13 :0 6 13 : 0 6 13 : 0 6 13 : 0 6 13 : 0 6 13 : 0 6 13 : 0 7 13 : 0 7 :221 nformation to a single end point. . Now_ you said that the central data station is like a :252 :303 able head end, right ? :324 . I'm using this colloquially so. . . :3 5 5 . But most cable systems have lots of head ends, right? :396 . Well, it depends what we call the "head end" so . . . :437 . Most cable systems have lots of things that those : 4 6 8 cable systems refer to as the head end, right? :499 . Yeah . And let me -- you know, what I have in mind is ; 5 210 the single Time Warner unit address, really, building i n : 5 911 Ithaca . Yeah, the central data station is at some -- some : 0 412 location that can access many users so that many users can : 0 813 access it.
. And most cable systems in the Eighties have multiple communities that they served and each community had its own cable head end, right ?

Page 09 09 9 09 09 09 :09 :09 :0 9 : 09 :09

164

3 3 3 13 3

:14 1 . And those places, those central data stations have t o : 19 2 get their content from somewhere, right ? :223 . Well, in the case of Walter, we see the central dat a : 26 4 station has these memory modules which are installed in the : 2 9 5 case of, for example, the -- the Time Warner building in :34 6 thaca, they certainly have cable receivers . :387 . Okay. :3 8 8 . I'm sorry . satellite receivers . :409 . So, the normal way to do it is electronically rather :44 10 than sending things by courier to the cable head end, right? : 4 911 R. PAYNE : Objection, form .
. I don't know what's normal in distribution these days .

3 : 0 9 :5 012 13 : 0 9 :5 313

13 : 0 7 : 0 614 13 : 0 7 : 1215 13 : 0 7 : 1516

13 13 13 13 13

: 0 7 : 1617 . That, I don't know . I don't know how cable systems : 0 7 1718 were configured in the Eighties . : 0 7 : 1819 . It's certainly true today, right? :07 :2020 . I don't actually know how cable systems are : 0 7 : 2 2 21 configured today.

3 :09 :09 9 3 :09 13 :10 13 :10 3 13 13 13

. (By Mr . Stephens) Okay . And you don't know what was :5 314 : 5 515 normal in 1985, right? : 5 616 . In terms of cable head ends, that's correct . : 5 917 . Okay . So, you don't know whether a cable operator : 0 518 would have been motivated to use the same kind of system that : 0 719 we see in Walter to distribute video to its cable head ends ,
right?

13101220

13 : 0 7 : 2 2 2 2 Q . Okay. So, you just pulled the cable head end idea 13 : 0 7 : 2 6 2 3 out of your head -- I'm just trying to understand how you 13 : 0 7 : 2 8 24 decided that the central data station would be located in the 13 : 0 7 : 312 5 cable head end. Page 163+

: 1 0 1521 R. PAYNE : Objection, form . : 10 : 15 2 2 A. So, I would like to see, I guess, a diagram of the : 10 : 2 0 2 3 type of cable system that you're referring to with what's : 10 : 2 2 2 4 called head end in order to answer that . 10 : 2 6 2 5 . (By Mr . Stephens) So, you just don't know without
age 16 5

13 : 07 : 31 1 . Well, because colloquially, I -- I would like to 13 : 0 7 : 3 6 2 elineate between using these as -- okay . 13 : 0 7 :3 7 3 hen I write these declarations in my reports, I 13 : 0 7 : 4 2 4 am attempting, I hope, to write these things in a manner tha t 13 : 0 7 : 4 6 5 people who are not tremendously technical can understand ; and 3 : 07 : 57 6 let me just clarify something here. And my understanding of 13 : 0 8 : 01 7 cable head end, which I think is a general person's 13 :0 8 : 0 6 8 understanding of cable head end, is a location which can 13 : 0 8 : 0 9 9 branch out and serve multiple users . 13 : 0 8 1410 . Okay . And are those often connected to other 13 : 0 8 1611 locations? So, just, for example, cable head ends often will 13 :0 8 :2112 have satellite down link, right? 13 :0 8 :2 313 . Cable head ends, yes, I would agree with that. They 13 : 0 8 : 2 714 do receive information, yes . 13 :08 : 2 815 . And they receive it so that they can retransmit it to 13 :08 :3 016 their subscribers, right?
13 : 0 8 :3 317 . Yes .

13 : 10 3 : 10 10 3 : 10 3 : 10 : 10

:271 having more information ; is that right? :292 . Yes. :473 . Okay. Would a person of ordinary skill in the : 5 0 4 Eighties have understood that you could use a disk drive in : 5 2 5 place of the memory modules in Walter ? :556 . I missed the very beginning verb. Did you say "woul d
7 "?

3 :10 :59

3 :10 : 5 9 8 . Would a person of ordinary skill in the art in 1985 3 : 11 : 0 2 9 have understood that you could use a disk drive in place o f : 0 510 the memory module described in Walter ? 11 : 0 811 . I think that one of ordinary skill would have 13 : 11 : 0 912 understood that you would not use a disk drive in place of the 13 : 11 : 12 13 memory modules. 3 : 11 : 1514 . Why? 3 : 11 1515 . So, first off, Walter has explicitly taught us tha t 11 1916 those are recirculating shift registers . But outside of that, 11 : 2 4 17 those units had part icular speed requirements in order to ge t 13 13 13 13 13 3 13 : 11 : 11 : 11 : 11 : 12 : 12 : 12 :3 819 :4220 : 4 6 21 : 5322 : 012 3 :0524 :0725 the data out at the speeds required for the -- the -transmission . And, in fact, Walter specifically talks about doing the setup of the data cells to somewhat reduce thos e rates a li tt le bit. And I do not believe that that data coul d have been accessed quickly enough using hard drives with the speeds that they had available in the Eighties . . But you -- in other words you might not be able with a single magne tic disk drive to get it out fast enough to send

13 : 0 6 :3 418 . And a typical cable operator like Time Warner will 13 : 0 8 :3 819 possess many head ends, right?
13 : 0 8 : 4 7 2 0 A. I was not using the term head end as one or many . 13 : 0 8 : 4 9 21 I'm - I'm using that term from the standpoint of from a -13 : 0 8 : 5 9 2 2 let's use Walter's tern -- data receiving station, that data

13 : 0 9 : 0 3 2 3 is going to come from some central location where there are -13 : 0 9 : 0 7 24 there is equipment located designed to approp ri ately get the 13 : 0 9 : 112 5 material to the data receiving stati on .

Merrill Legal Solutions -888-5 3-980 0 www .merrillcorp .co m

1 1

l rQ Q A Q A Q Q A

1 1

1

aQ WA Q A

1
Page 7 of 18

Case 3:06-cv-00019-MHP

Document 157-9
-

Filed 09/06/2007 9/4/0 7

Sheila Hemem i

43 (Pages 166 to 169 )
Page 166
Page 16 8

13 : 12 :11 1 a two-hour movie in 31 seconds ; is that right ? 13 : 15 : 21 1 alter, right? 3 : 12 :13 2 . Well, a single magnetic disk drive is certainly not a 13 : 15 : 2 5 2 . I don't know that that would be possible in that one 3 : 12 :16 3 ecirculating shift register. 13 : 15 :3 1 3 would have a substantial issue in coordinating what I expect 3 : 12 :16 4 . No, I'm -13 : 15 :3 6 4 to be a fairly large number of drives . I mean, we're no t 3 :12 :16 5 . But yes . :3 9 5 talking, like, five drives here to get 300 megabits . 3 : 12 : 19 6 . -- I'm specifically asking about if you replaced the :15 :42 6 . Do you know how many drives would be required ? 3 : 12 :21 7 ecirculating shift register with a disk drive . 13 :15 : 46 7 . I am sure it would be at least double digit; but, 3 :12 :2 5 8 . If you replaced a recirculating shift register with a 13 :15 :4 9 8 gain, I just don't have those disk capacities on the tip o f 13 : 12 :2 5 9 disk drive - first off, let me note that Walter states that 13 :15 : 5 3 9 my tongue right now . So, I'd rather not do some hypothetical 13 : 12 :3 410 the compressed data rate is, if I remember correctly, 4 4 13 :15 : 5 710 calculation which may or may not. But I am confident that it 13 : 12 :3 811 megabits per second. So, we see that on Column 7 around 13 :16 : 0 011 would be double digit and possibly even over a hundred . So , 13 : 12 : 44 12 line 39 . So, given the size of single disk drives in the 13 : 16 : 04 12 having said that, the data could certainly be stored. One 13 : 12 : 5 013 Eighties, which I don't have on the tip of my tongue, but I am 13 : 16 1213 could have the fiber optic lines, but there's still a matter 13 : 12 : 54 14 certain would not be able to store more than several seconds 13 : 16 1714 of getting the data off and coordinating it, for lack of a 13 : 12 : 5 8 15 of video -- so, from the standpoint that we're -- Walter is 3 : 16 : 2 315 better way to put it, in such a manner that it can be 13130 2 16 referring to a two-hour movie,l think one of ordinary skill 13 : 16 : 2 816 appropriately multiplexed and packaged for transmission over 13 : 13 : 0717 would understand that . 3 : 16 : 3 517 the fiber optic links as taught. 13 : 13 : 1518 . So, you - it's your testimony that disk drives 3 : 17 : 0 918 . Now, earlier today we were talking about audio 13131819 weren't big enough to store a two-hour movie at that data 3 : 17 1219 compression and you said that a person of ordinary skill in 13 : 13 : 2 2 2 0 rate; is that right? 3 : 17 : 18 2 0 the art would have been aware of statistical characteristic s 13 : 13 : 2 3 21 . Well, at 44 megabits per second, two hours is 7,200 3 : 17 : 2 2 21 of audio and the way human ear processes audio information, 13 : 13 : 2 8 2 2 seconds . And actually, we can do the multiplication here, 31 13 : 17 : 2 8 2 2 right? 13 : 13 : 3 8 2 3 seconds times 10,400 megabits per second . So, this gives 3 : 17 : 2 8 2 3 . Yes . 1313412 4 us 300 -- 300 megabits, and I don't believe in the 3 : 17 : 2 8 2 4 . Now, humans hear audio in a way that makes them more 13 : 13 : 512 5 mid Eighties there were 300 megabit single magnetic drives . 13 :17 :402S sensitive to lower frequencies across the human hearing range ;
Page 167 : Page 16 9

13 13 13 13 13

:13 :13 : 14 :14 :14

:5 5 1 . Okay . But you could stripe disk drives together in a : 5 9 2 RAID array to get that size, right? :08 3 . Can you give me a little bit more details about what :10 4 you mean by RAID? I think I mentioned -:14 5 . Redundant -. Yeah, I understand what Redundant Arrays of -- I'll et you finish that .

13 : 17 : 4 4 1 is that right? Or maybe I should say more finely able to i 13 : 17 : 4 9 2 distinguish between tones at lower frequencies? 13 : 17 :53 3 A. That's very well put. Yes . We essentially have a
13 :17 :S7 4

13 : 14 : 14 6 13 : 14 : 17 7

13 : 14 : 18 8 I think it's variously called Inexpensive Disks or 13 : 14 :22 9 something else, which I don't remember . 13 :14 : 2 510 A . So, my familiarity with RAID also came from when I 13 :14 : 2 811 worked on video on demand, which was in 1994 . So, all the 13 :14 :32 12 nuances involved in -- in something specific known as RAID, 13 :14 :3 713 don't -- I'm uncomfortable answering . I would agree that one 13 : 14 : 4 214 could assemble a large group of disk drives such that the sum 13 : 14 : 4 715 of the capacity was 300 megabits . 13 : 14 :5 016 . And it was known that you could stripe data across 13 :14 : 5 317 them and to increase the UO, the data rate, at which you 13 :14 :57 18 could read them, right? 13 :14 : 5 819 A . I don't know for a fact that that was known in the 13 :15 : 012 0 mid Eighties, but I think it's reasonable to expect that it 13 :15 :0321 was . 13 :15 :04 22 . Okay . And if your only concern was to send it 13 :15 : 0 8 2 3 quickly but maybe not send it in 31 seconds, you could use a 13 :15 :13 24 RAID array of disks in 1985 to store digital video and send it 13 :15 :192 5 faster than real-time over the optical network described in

13 :18 13 :18 13 : 18 13 : 18 13 : 18 13 : 18 13 : 18 13 :18 13 : 18 13 : 18 t 13 : 18 13 : 18 13 13 13 13 13 13 13 13

:00 5 : 04 6 :08 7 : 13 8 :14 9 41 0 81 1 : 24 1 2 : 2 71 3 : 3 21 4 : 3 31 5 : 3 61 6

filter bank in our ears, and the bandwidths of the lower frequency channels are narrower . So, therefore, if we have two tones that are some distance apart, they may fall into separate channels and we can delineate that . Where if we move them up to higher frequencies, they may fall within a single filter bandwidth . Q . And that's reflected in the way musical instruments are configured in octaves, right? A . I think that's a philosophical question . I don't -I don't know even what configured for a musical instrument means . Q . Well, for example, you've referred to the keys on a piano keyboard as being in different octaves, right ?
A. Yes, but that -- okay.

13 : 18 : 411 7

: 18 : 4 41 8 : 18 : 4 91 9 : 18 : 5 3 2 0 : 18 : 5 4 2 1 : 18 : 5 6 2 2 : 19 : 0 0 2 3 : 19 : 042 4 : 19 : 0 8 2 5

Q. And an octave in ordinary acoustical meaning refer s to doubling frequency . If you move up one octave, you double the frequency? A. Oh, okay. Now I understand what you're saying . So, I would say that, yeah, our -- as humans, our representation of music or acoustical information is tied in with -- with our hearing sensitivity, yeah . That statement, I -- that I just said, I would --1 would agree with .

Merrill Legal Solutions www .merrillcorp .co m -888-513-980 0

1

oA r kQ Q A Q Q Q A

S

P

>f 1 1 1

2 2

2

tQ eI A I T A Q A (

1

P

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007
9/4/0'7

Page 8 of 18

Sheila Hememi
age 170 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 :19 :10 1 . Okay . And, so, the entire human hearing frequenc y : 14 2 ange from 10,000 hertz to 20,000 hertz is really just the top : 19 : 19 : 18 3 ctave; is that right? : 19 : 2 3 4 . Now, let's delineate -- octaves are terms that we use : 19 : 2 9 5 as humans to describe music . With respect to the ear, we : 19 : 34 6 refer to what are called critical bands . And the critical : 19 : 4 0 7 bands are effectively -- there's no one single answer for : 19 : 4 3 8 exactly where the edges of the critical bands are . I think : 19 : 4 7 9 there's sort of general agreement. but specifically people can : 19 :5 110 move them around a little bit -- is the critical bands that we : 19 :5 4 11 refer to that actually serve as the frequency delineations in : 2 0 : 0 012 the ear. Now -: 2 0 : 0113 . How many critical bands are there ? : 2 0 : 0 414 . I don't know how many critical bands there are, but : 2 0 : 0 715 to map the critical band back to your question or the are a : 2 0 : 1016 that you talked about between 10,000 and 20,000, I do not knov : 2 0 1617 whether there are critical boundaries, critical ban d : 2 0 : 2118 boundaries in that range or not . . Do you know approximately how many critical bands : 2 0 : 24 19 : 2 0 : 2 6 2 0 there are? . They're on the order of tens -- "tens" meaning order : 2 0 : 3 0 21 : 2 0 : 3 3 2 2 of magnitude so -Q. But not a hundred, right?

44 (Pages 170 to 173 )
Page 172

3 :22 :051 . You know, I have to say that I have no idea what unit 13 : 22 : 10 2 you have, how it is computing, how much it should light up o f 2 2 :14 3 those, what type of preprocessing or signal conditionin g 2 :18 4 they're doing, whether that is reflecting what your graphic 2 :225 qualizer settings are or not . So, I -- I would not like t o 2 2 : 2 5 6 make any type of guess based on what you've just told me . 3 :22 :297 . Do you have any reason to believe that it -- do you 13 : 2 2 :3 5 8 have any reason to disbelieve the proposition that I'm making hat there is relatively little musical content above 1 3 3 :22 :389 13 : 2 2 : 4 310 kilohertz? 13 : 22 : 4311 A. Well, I'll make two comments . One is I don't know 13 : 2 2 : 4 612 what "relatively little" is ; and secondly, I do believe that 13 : 2 2 : 5 213 the standard that exists for digitizing audio for compact 13 : 2 2 : 5 614 disks was designed by smart people and it is a reasonabl e : 2 3 : 0215 standard and the fact that they sample at 44 .1 kilohertz, :23 :0816 which would be twice of the highest frequency of 22 .05, I 13 : 2 3 :1317 believe that decision was made on sound engineering 13 : 2 3 :1518 principles . And if they really believed that, indeed, w e 13 : 2 3 1919 needed to -- we could get away with cutting off at 13 .5 or 15 13 : 2 3 : 24 2 0 or 17, that we would have that because the end result means 13 : 2 3 : 2 9 21 that we would -- this goes back to our storage. We would have 13 : 2 3 : 3 3 2 2 more data on a compact disk. 13 :2 3 : 3 4 2 3 Q . s it fair to say, though, that that decision does 13 : 2 3 : 3 7 2 4 not mean that information is uniformly distributed across th e 13 :23 :422S audio spectrum? age 173

13 : 2 0 : 3 5 2 3 13 : 2 0 : 3 6 2 4

13 : 2 0 : 3 9 2 5

A. Not a hundred, no . Q. Do you know offhand statistically how much musical Page 171'

13 : 2 0 : 4 4 1 information is represented in a band from 10,000 to 20,000 i 13 : 2 3 3 :23 13 :2 0 : 4 8 2 hertz? 3 :23 13 : 2 0 : 4 9 3 . So, can you clarify what you mean by "statistically" 3 :23 132 0 : 5 2 4 and how much -132 0 : 5 2 5 . If you take a large sample of popular music, how much ; 132 3 3 :24 13 : 2 0 : 5 6 6 information content, how much power relative to the entire 3 : 24 13 :21 :0 2 7 signal is being delivered in that band from 10 to 20 13 : 24 13 : 21 : 0 6 8 ilohertz . 3 : 24 132107 9 . Okay . So, now we've, okay, clarified with respect to 3 : 24 13 210 910 power. 3 : 24 13211011 o, were we to take, say, eight FM radio 3 : 24 13 : 21 : 17 12 stations, hopefully one of which would include classical 13 : 21 : 2113 music, and average out, essentially do a histogram for power 13 : 2 4 3 : 24 13 : 21 : 2 6 14 across various frequency bands, I don't know what that 3 :2 4 13 : 21 : 2 815 histogram would look like at -- let me just emphasize that it 13 : 2 4 13 : 21 : 3 3 16 is important to average . We c an 't take a single piece of 13 : 2 4 3 : 21 : 3 6 17 music and say that it's -- or an y audio content, a symbol 13 : 2 4 13 : 21 :3 918 crash, and say that it's representative for the symbol class 13 : 2 4 13 : 21 : 41 19 as a whole . 13 : 21 : 4 2 2 0 . Well, let me just give you an example of personal 13 : 21 : 4 3 21 experience. I have a rather elaborate stereo systetn that has, 13 : 21 : 4 9 2 2 in part, a sort of frequency analyzer display which I can 3 13 : 21 : 5 2 2 3 watch while I'm listening to music ; and I see very little 13 : 21 : 5 6 2 4 content above 13 kilohertz ever -- do you have any reason to 13 3 13 : 2 2 : 012 5 disagree with that -- over the type of music that I listen to . 7 27 :27 :27 :27

: 44 1 . Information is certainly not, in the content of any : 4 7 2 type of audio with the exception of white noise, uniformly :S2 3 distributed across the audio spectrum . :554 . Okay . And in fact, it is concentrated in the bands : 5 7 5 in which human speech is also concentrated, right ? . I have a particular plot in mind from one of my :056 : 0 8 7 lectures, but I honestly don't remember the axes so well . So, : 12 8 'm a little hesitant to nail it down to that, but I will say : 15 9 that certainly in terms of our sensitivity, we have -- on a : 1910 log/log plot, we have a peak that goes up at frequencies below : 2 511 10 kilohertz and then falls off. And I'll just add, we have : 2 912 an extreme sensitivity at a high frequency range which is : 3 3 13 where babies scream. It's true . : 3 414 . And what is that frequency ? : 3 715 . I think that's around 7,000 hertz, but it's very , : 3 9 16 very clear on the plot . : 4 2 17 . We need to change the tape . So, let's take a very : 4 4 18 brief break . : 4 519 HE VIDEOGRAPHER: Off the record at 1 :24 . 0 Recess taken) : 0 0 21 HE VIDEOGRAPHER: This is the deposition of : 2 2 2 2 Dr . Hemami . The time is 1 :27. We're back on the record. :2723 . (By Mr . Stephens) Dr . Hemami, if you were to take : 2 9 24 Claim 1 of the '995 patent completely apart from th e : 3 3 2 5 disclosure of the specification or the drawings --

Merrill Legal Solutions -888-513-980 0 www .merrillcorp .co m

cA 'A tQ A Q Q Q A

M M T

11 : 0 1 1

bQ tr A lA Q Q A Q Q A A Q

M T

1
Page 9 of 18

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007
9/4/07

Sheila Hemem i

45 (Pages 174 to 177 )
Page 174 Page 17 6 :30 :431 . Looks like --

13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13

: 2 7 :3 8 1 3 . Could I just seethe patent? I don't want to go on :27 :402 my memory of the claim . 3 :27:503 . All right. Just a second. : 2 8 : 04 4 his has previously been marked as Exhibit 1 . : 2 8 :13 5 R . STEPHENS : Les, I'm assuming you've got 3 : 2 8 :13 6 hat ? : 2 8 : 14 7 R . PAYNE : Yeah, thanks. : 2 8 : 15 8 . Okay . :2 8 :16 9 . (By Mr. Stephens) If you took Claim 1, would you be 13 :2 8 : 17 10 able to build that -- a person of ordinary skill in the art, : 2 8 : 21 11 wou ld they have been able to build that just based on what 13 :2 8 : 2 5 12 they knew as a person of ordinary skill in the art? :2 8 : 2 8 13 R. PAYNE : Wait, this is a -- is this an : 2 8 : 3 0 14 enablement question? :2 8 :3 3 15 R. STEPHENS : No . It's actually abou alter; : 2 8 : 3 3 16 so, bear with me . :2 8 : 4 0 17 13 R. PAYNE: Can you repeat : 2 8 : 41 18 . (By Mr . Stephens) So, let me -- let me just ask the 3 : 2 8 : 41 19 question differently. Would a person of ordinary skill in the, 13 : 2 8 : 4 5 2 0 art in 1985, if they were presented with Claim 1 of the '995 13 : 2 8 : 4 8 21 patent and the disclosure of Walter been able to build what's 13

:30 30 30 :3 0

:452 . - I've mentioned that -:483 . Okay. Go ahead. :494 . I don' t know what the host -- I don't know enoug h : 5 0 5 about the host computer to say whether it was capable of doin g : 53 6 hat or not. :557 . I think your misapprehending my question a litt e : 56 8 it. I'm not asking whether all the elements in Claim I ar e :3 1 : 0 0 9 present in Walter. I'm asking whether if you had the :03 10 information in Walter, would you be able to build what's in :05 11 Claim 1 if you were looking at Claim 1 so it tells you yo u :3 1 :0 9 12 need something to do compression and it tells you you nee d :12 13 random access memory . With that and the knowledge of a person :3 1 : 15 14 of ordinary skill in the art and Walter, would you be able to :31 :2215 build what's in Claim 1 ? 31 2 8 16 . I -- I -- I genuinely do not understand the question . :31 :3517 . Well, so, for example, when it talks about a mean s :31 :42 18 for compressing audio/video source information, Walter :31 :4719 describes an algorithm for doing that, right ? :3 1 : 51 2 0 . Walter mentions an algorithm for compress on of : 31 : 5 5 21 video .
. Okay . And random access storage means is something

13 : 2 8 : 51 2 2 claimed? 132 8 : 5 6 2 3

3 : 31 : 5 6 22

. Okay. I'm going to repeat this to make sure 113 : 31 : 5 9 2 3 that people of ordinary skill in the art were very familia r

13 : 2 8 : 5 9 2 4 understand this. This person has Walter . 13 : 2 9 : 0 2 2 5 . Yes .
Page 175

3 : 3 2 : 0 2 24 with in the mid Eighties, right ? 3 :32 :0625 . Random access storage would have been known to one of
Page 17 7

13 : 2 9 : 0 3 1 . And this person has just the text o Claim 1 from the 13 :32 :10 1 skill . 13 : 2 9 : 0 9 2 13 :3 2 :11 2 995 patent? . Okay . And Walter describes a way of outputting 13 : 2 9 :11 3 13 :32 :16 3 audio/video information over an optical fiber faster than . That's right. 13 : 2 9 : 2 0 4 13 :3 2 : 2 0 4 . Well, my first off-the-cuff remark is that Walter eal-time, right ? 13 : 2 9 : 2 5 5 hasn't given us any compression means . 13 :3 2 :2 7 5 . Walter describes -- in fact, most of the stuff in the 13 : 2 9 : 2 9 6 . Okay. So, let me ask it slightly differently then . 13 :3 2 : 31 6 central data station is to get the data onto the fiber optic 13 : 2 9 :3 3 7 Let's look only at the functions for the means-plus-function 13 :3 2 : 3 6 7 ines . 13 : 2 9 : 3 7 8 laims. 13 : 3 2 : 3 6 8 . Okay. So, is there anything that's in Claim I tha t 13 : 2 9 : 3 8 9 . Let me also add, we don't have any random access 3 :3 2 : 4 0 9 you would not be able to figure out how to do if you had 13 : 2 9 : 41 10 storage means . So -- okay . So, you would like me to look 13 :32 :4510 Claim 1 and the disclosure of Walter? 13 : 2 9 : 4 3 11 just at what's going on -3 :3 2 :5 011 HE REPORTER : I'm sorry, "disclosure of -13 : 2 9 : 4 4 12 . Right. 3 :32 :5 3 12 R . STEPHENS : Of Walter in 1985 . 13 : 2 9 : 4 5 13 3 :3 2 :54 13 . -- not the thing itself. . Well, Claim 1 also requires an input means . 13 : 2 9 : 4 6 14 Q . Right . I'm not asking you to perform the 112 3 :3 3 : 0 014 . (By Mr . Stephens) Okay . 13 : 2 9 : 4 8 15 analysis . If you'd just look at the functions that are 3 :3 3 :0 015 . And I look at Walter and we've discussed that our 13 : 2 9 : 5 0 16 recited. 13 :3 3 :1116 central data station does not have an input means . 13 : 2 9 : 5 0 17 . I don't know what that means. 3 :33 :13 17 . But the receiving station does, right? 13 : 2 9 : 5 2 18 . Fair enough. If you just look at the -- the 3 :3 3 :16 18 . The receiving station does receive . 13 : 2 9 : 5 3 19 functions that are recited and you take a person of ordinary 13 :3 3 : 19 19 . So, if you need an input means, the -- the structures 13 : 2 9 : 5 6 2 0 skill in the art that has that list of functions in Claim 1 3 : 3 3 : 2 3 2 0 for creating one are described in Walter, right ? 13 : 3 0 : 0 0 21 and Walter, would they have enough information to build 3 :33 :32 21 . Walter -- there are two other things here, and I 13 : 3 0 : 04 22 Claim 1? 3 : 3 3 : 3 4 2 2 think I'm still not exactly understanding what it is you'r e 13 : 3 0 :2 6 23 . I guess I'm not exactly -- I mentioned that Walter 3 :3 3 :3 7 23 asking me . Claim 1 starts off with an audio/video transceiver 13 : 3 0 : 2 8 24 doesn't give us any compression mechanism . Walter doesn't 13 : 3 3 : 4 2 24 apparatus . So, we have an apparatus, not two apparatii or - 13 : 3 0 : 3 2 25 teach us that anything in Figure 1 is a compressor, and -3 :3 3 : 4 8 25 . Okay .

Merrill Legal Solutions www .merrillcorp .co m -888-513-9800

Q Q A A Q A

1: 1:6

9A rA tQ aA mQ Q

M

1

P P

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007

Page 10 of 18

Sheila Hememi
Page 1781

/4/0 7
46 (Pages 178 to 181 )
age 18 0

13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13

13 :33 :491 . -- or two central - two separate locations . ; and it tells you, you : 3 3 : 52 2 . So, you're looking at Claim 1 : 3 3 :55 3 want them both in a transceiver, you're trying to build a :33 :594 transceiver . Walter has both a transmitter and a receiver . : 3 4 : 0 2 5 Now, according to you, it doesn't disclose or suggest putting :34 :056 them together into one box, but Claim 1 does, right? So, if : 3 4 : 0 9 7 you had Claim 1 and the disclosure of Walter, you'd have a 3 : 3 4 : 12 8 transmitter and you'd have a receiver and you could combine :34 :16 9 them in a way to make a transceiver, right ? 3 :34 :1710 . So, at the risk of sounding as if I'm trying to :34 : 1911 agitate you -- and I'm truly not; this is a genuine legal : 3 4 : 2 2 12 question -- isn't this a classical case of hindsight? I have 3 :3 4 :25 13 ingredients in a box - -13 :3 4 :2 8 14 . It absolutely is, yes . 13 :3 4 :2 8 15 . -- that I didn't intent to -13 :34-.3116 . I'm saying, you have the claim. 13 : 3 4 :3 117 . -- put together and -13 : 3 4 :3 218 . So, I'm asking you, if you have that hindsight, if :34 :3S19 you have the claim, do you have all the technology that you i 13 13 :34 :3 9 2 0 need to build it? So, yes, I agree with you . This is 13 : 3 4 : 4 2 21 hindsight I'm asking you to engage in . :34 :34 :34 :34 : 4 5 2 2 A. Well, again, we don't -- we don't have -- Walter has : 4 9 2 3 given us a compression algorithm . Walter -- Walter has said : 5 6 24 the data is compressed using this algorithm. Walter has not : 5 9 2 5 given us any mechanism by which -Page 179

:36 :10 1 of hindsight and you were ooking at the claim? 14 2 . Well -- -:15 3 R. PAYNE : Objection, form . : 16 4 . -- I feel like I'm still missing something ; but let :205 e say, I look at Walter . First off, I want a single :24 6 pparatus . :247 . (By Mr. Stephens) Okay . And the claim tells yo u :3 6 : 2 5 8 hat, single apparatus? 36 :279 . And I don't see that in Walter. The -: 3 6 : 3110 . But -- but could you build it, given the disclosure :3 6 : 3 311 of Walter and the suggestion from the claim ? 3 6 : 3 512 . Well, let me work through the four elements .
:3 6 : 37 13 . Okay .

:3 6 :3 814 A. So, okay, we have our apparatus requirement. Input : 3 6 : 4 315 means : So, I look at Walter and Walter as a data receiving : 3 6 : 4 716 station certainly has input means . The signal gets on it : 3 6 : 5 217 through the fiber optic. Compression means : Now, I think I'm :36 :5 818 still confused about what's going on here, but one of ordinar y : 3 7 : 0 419 skill certaintily understands -- would understand, as we just :37 :072 0 discussed, that this thing could be implemented in a variety :3 7 :10 21 of ways, which doesn't really pertain to what Walter provide s 3 7 :16 2 2 us with other than a single -- an algorithm.
. But that algorithm would be enough to create the

3 :37 :2023

13 : 3 7 : 2 3 24 compression means of the claim, right?

37 :2925

. I feel like that might be a legal question, and I age 181

3 :37 13 : 3 5 : 0 3 1 . Now, you told the Court -3 :3 7 13 : 3 5 : 04 2 . . -- that could be done 13 : 3 5 : 0 5 3 . -- in the -- in the tutorial that a system designer= 13 :3 7 13 : 3 5 : 0 8 4 wouldn't be so concerned with the details of that because the ? 13 :3 7 13 :3S :11 5 system designer would understand that there were a variety oT 13 : 3 7 13 : 3 5 : 14 6 ways to implement a compression algorithm once you had it, 13 : 3 7 3 :3 7 133 518 7 right? 3 :3 7 13 : 3 5 : 19 8 . Yes. 3 :3 7 13 : 3 5 : 21 9 . You mentioned, like, an ASIC and a -- maybe you 13 : 3 5 : 2 3 10 mentioned a field programmable gate array . I forget, but you 13 :3 7 3 :3 7 133 5 : 2 811 mentioned a number of different approaches? 3 :37 13 : 3S : 30 12 . Yes, that's correct. 33 7 13 : 3 5 : 3 013 . Okay . So, a person of ordinary skill looking at 13 : 3 5 : 3 414 Claim 1 and looking at Walter would understand those ways 13 :3 8 3 :3 8 13 : 3 5 : 3 6 15 would be available to them to implement the algorithm 13 :3 8 13 :3 5 :3 8 16 described in Walter, right ? 13 :3 8 13 :3 5 : 43 17 . I think in the - in the --this goes back to could 13 :3 8 13 :3 5 :4 5 18 we do this on a computer . In the abstract, an algorithm , 13 : 3 8 13 : 3 5 : 4 8 19 provided that it can be implemented, is -- one has options to 13 : 3 8 13 :3 5 :54 2 0 do that, yes, that's certainly true. 13 13 13 13 :3 5 :35 :3 6 :36 13 :3 6 :56 21 . And those options were understood to persons of : 5 8 2 2 ordinary skill, right? : 0 0 23 . One would hope, yes . :0024 . Okay . So, is there anything in Claim I that would : 05 2 5 not be enabled by Walter, if -- again, if you had the benefit

: 31 1 don't know how to answer that . :3 2 2 . Okay . I'm just talking about the function . :3 4 3 . So -:3 4 4 . So, there is an algorithm that is present in Walter : 3 7 5 for compressing audio/video source information : and one of : 4 3 6 ordinary skill would be able to built it based on the : 4 6 7 disclosure in Walter. I think you've already said that, :488 ight? :489 . Okay. Yeah, let's not use the word "means" there . :5 110 . Okay . :5 111 . Okay. Random access storage means: Walter does : 5 3 12 not -- not only does Walter not have random access but, as I : 5 813 have mentioned in my declaration, I believe that Walter :0 1 14 actually gives us a specific alternative. The system is :0 515 designed around using these recirculating shift registers. We :0 916 talked a little bit, had a disagreement as to whether on e :12 17 would want to create such a thing from a DRAM or a random :1618 access . So, while random access storage would be known to one : 2 219 of skill, I think looking at Walter, random access storage i s : 2 7 2 0 not something that one of skill would understand would be : 3 0 21 appropriate given the system, so, just simply looking at what 38 :3522 we have available in Walter . :3723 . Okay . But, again, a person of ordinary skill in th e : 4 0 2 4 art would know about RAM, right, so when they read it in the 8 : 4 3 2 5 claim. they'd know what it is ?

Merrill Legal Solutions -888-513-9800 w ww .merrillcorp .co m

sa A iQ Q Q A Q A

O W

;1 :1 1

4

A Q

1
Page 11 of 18

Case 3:06-cv-00019-MHP

Document 157-9

Filed 09/06/2007
9/4/0'7

Sheila Hememi

47 (Pages 182 to 185 )
Page 18 2

Page 184 random access memory, is there anything else that you would need in order to be able to do it; Or is the suggestion enough? A . I'm not sure what you mean by -Q . Do what the claim requires . In other words -A . I'm still not sure I understand what you're asking . Q . -- random access storage coupled to compression for storing a time compressed representation . So, would a person of ordinary skill in 1985 who had that suggestion along with the rest of the claim be able to build a random access storage means that stored the compressed video as we've talked about? A. Okay. Can you repeat that again? Q. Yes. Would a person in 1985 looking at the claim language that talks about a random access storage coupled to the compression for storing a time compressed representation, would they be able to build such a random access storage device? A. Well, I think certainly they could go with a bunch of chips and set them off. Now, again, you know, Walter doesn't give us any compression means. So, this is all very nebulous, right? We've got some way that some system designer is going to implement the compression means . Can that be interfaced to random access storage means? I guess that depends on how they implement the compression means . Q . Okay . But if you were - if your boss handed yo u Page 185
3 : 42 :3 4 1 13 : 4 2 : 3 8 2
3 :42

13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13 13

:3 8 :44 1 . They would certainly have an understanding of what 3 :41 : 05 1 :3 8 : 4 6 2 random access referred to . 3 :41 :0 9 2 :3 8 : 4 7 3 . And they would know that you could store video, 3 :41 :12 3 :3 8 :4 9 4 compressed video in ran dom access memory, ri ght? 3 :41 :15 4 :3 8 :52 5 . I don't know if compressed video was stored in random 13 : 41 :16 5 :3 8 :5 6 6 ccess memo ry at that time . I -- I don't know . 3 : 41 :24 6 :3 8 :5 9 7 . Certainly it was done in the process of compressing 3 :41 :2 6 7 : 3 9 : 01 8 t on a computer, correct? 3 : 41 :