A Treatise on Patent 4,963,995

The author is none of the following: lawyer, patent examiner nor degreed engineer. The author holds a position in the shares of Burst.com. The author has no other current or past financial interest in Burst.com. The author is not, nor has ever been, an employee of, director of, or consultant to, Burst.com. The author has never met, nor is affiliated with, any persons in those capacities at Burst.com. The author has not been paid, or received any other consideration, for this writing.

Investors should consider this personal opinion. Investors should do their own due diligence before making a decision on whether to invest in the shares of Burst.com

Qualifications of the author:

1. held shares since the stock symbol for Burst.com (current BRST) was IVDO;
2. read each and every single one of the posts on the Silicon Investor threads re IVDO;
3. read and reviewed the many information sources and references posted at Burst investors website and  the Burst.com squidoo lens
   
4. read the entirety of both first and second transcripts from the Markman  Hearing in the Burst v. MSFT case, which was held to construe the claims of Burst's patents
   
5. have searched circa 500+ patents at uspto.gov related to early (pre 1988) digital video and audio compression
6. have read and reviewed all prior art cited by the four primary Burst patents: 4,963,995; 5,057,932; 5,164,839; 5,995,705
   
7. worked in high technology  from 1976 to 1999 and invested in tech stocks during that time including video conferencing firms VSVR and PCTL
8. specialized in marketing at major min-computer vendors at a time (important) when client server computing was just getting underway
9. served as webmaster for a NASDAQ listed telecom from 1996 to 1999.
10. tested IVDO technology in ~1998 (as we were trying to find a way to implement video on the web)

■ Why the ‘995 patent is important

In the case of Burst we have many imponderables and unknowables. Confidential negotiations. What Apple does or does not do in Court . And what Burst does as well. What future business there might be for Burst in its new recently disclosed patents at the 2006 SHM (stake holder meeting). Who might be future defendants. Etc.

To me, what is said by the patents in litigation is real and can be analyzed. And the Apple case is probably at least about 95% about their validity. (The other question – is infringement of the Burst patents by Apple going on – to my view looks very probable – see the smoking gun posts #10084 and 10179.) So perhaps there is some predictive value here. And that is important because everything (including the above imponderables) is levered to the early patents. And the first patent – 4,963,995 – most of all. This patent is owned by Burst.com and appears to be the primary basis for its claim of infringement upon Apple. As part of its defense (and I believe the main part) , Apple is seeking to invalidate this patent, among others. So a lot is riding on the validity of '995. To understand its validity, it is necessary to investigate '995 itself, the state of the art it deals in at the time of its initial (1988) and amended (1990) application, and what other inventors were doing at that time. That’s why I have been burning the midnight oil at the website of the US Patent Office. Here is my take.

■ The big picture.

The ‘995 patent makes key claims about "time compression" of audio and video. It is therefore important to understand the following.

There is a fundamental difference between ordinary electronic data – such as text files, images, spreadsheets, even applications themselves – and audio and/or video programs (what ‘995 called "source information"), or data files that represent such audio/video programs. With ordinary data there is no metric of playback time associated with that data. If it is an image, it just sits there for you to look at. If it is text, you can read it as fast or as slow as you want. But because audio and video represent a performance recorded (created) originally in real time – of an event that took place over an interval of time (as a function of time) – there is a corresponding metric of time when it is viewed, or played back.

So an image, for example can be described as having a certain number of pixels and/or colors and/or resolution. But a video program or video file has all those characteristics and more. It has others uniquely related to playback: sampling (or bit) rate, frame rate and playback time. Similarly, an audio file has a metric of sampling rate and duration of playback. If you wish realistic playback, these time metrics are fixed. (The playback rate, especially, must be the same as the original recording rate. Otherwise, for example one can get a keystone-cops effect in video or a chipmunk-song effect in audio.)

Keep this time basis for audio and video in mind. As we shall see in a moment it becomes very important.

At the time of the applications for the ‘995 patent, the primary way to deliver audio/video to the user’s home was to broadcast it in a communications "band" – over the air or a coaxial cable – for immediate playback, or immediate recording on a VCR. The program streamed into the home at the correct playback rate, you watched or recorded it, and that was that. But because the audio/video had to arrive at a fixed rate – to assure correct and realistic playback – it had to continuously occupy a fixed part of the band that was available over the air or on coax. There was no way to send it but continuously in a stream. Thus the program continually and fully occupied a section of bandwidth. And, bandwidth is not unlimited.

Now, at that time, broadcast TV and radio sent programs over Mother Nature's infrastructure -- the airwaves. Because the entire frequency spectrum was broad (VHF, 1-13, UFH, 14-69 for TV, for example), bandwidth was relatively plentiful -- and free if you had an FCC license for your channel. The audience, for example, tuned into Channel 13 and could only watch one program there, one sequentially after the other. If you wanted to see something else you had to wait until it came on, or change the channel, AKA the frequency. You had to move to another part of the band to watch another program. Since there was a lot of bandwidth, the user had available to them many programs playing concurrently that could be selected and watched.

But cable/pay TV was a little different. Cable broadcasters began to offer many channels over coaxial cable to compete with free broadcast TV. And they could not send them out over the free and very broad band airwaves; they had to use their own dedicated cable infrastructure -- which had a limited bandwidth and which had to be built. Now, as noted, each audio/video program needed a continuous part of the band available over coax. The more programs they offered, the more bandwidth they needed. They began to reach capacity. They began to run out of bandwidth. What was their solution? Add more bandwidth? Very expensive. Too expensive.

In the 1980’s inventors began to patent various ways for more channels to run on that limited bandwidth. Computers to the rescue. They began to digitize audio/video. To a certain extent, digital audio/video let more be delivered over the existing infrastructures. In a technique called multiplexing several streams of one and zeros (each representing a program) could be processed so they could then be interleaved and sent out at the same time. Multiplexing is, by definition*, a way to allow multiple signals or  streams to occupy one single, particular channel. (* Newton's Telecom Dictionary.)

All data can be digitized and compressed, of course. But when audio/video is digitally compressed two  things happen to it. The first is fairly ordinary. The second is extraordinary because as you will remember audio/video is time based.

First, its bandwidth is reduced. In this event, material was still  broadcast continuously. But now each program now occupies less of the common "band" being used to transmit it. Less bandwidth per program, more simultaneous programs  per coax. The cable guys had their solution. Two hundred programs at once per wire for a media-hungry public. But....they did not notice something else! And this is the extraordinary part.

Because, second, if (note: this is a critically important if) you have the right equipment, by means of compression the time to deliver a fixed length audio/video is also reduced. Now, given a fixed bandwidth, the time to deliver any compressed hunk of data can be less than it would be to deliver its uncompressed counterpart. No big deal. But remember that for ordinary data this reduction of delivery time has little importance because there is no playback time associated with such data. But there IS a playback time associated with audio/video. And so what if:

• the data is question is audio/video source information, and
• the delivery time is less than it would take to conventionally broadcast
• the delivery time is less than the playback time
• the delivery can be done direct to the user's home on a consumer device
• the audio/video becomes a digital "OBJECT" that can be saved, edited and resent
• audio/video can be delivered and possessed  in less than the time it would take to view it / play it back
  

Now that is a big deal for 1988-90. And that is what the Burst ‘995 patent invents -- and specifically claims.

Because, as of the application for the first Burst patent, only the first thing – program material with a compressed video bandwidth – was being broadcast – streamed – commercially to user’s homes or offices. In my experience, the only uses of video compression that were commercially occurring then were:

1. to allow live video conferencing back and forth over special communication lines – usually ISDN or T-1. Because the expense of the line was (and is) a function of its bandwidth, video compression let a better quality video be sent over a narrower economical "band"
2. to allow video broadcasters to send more concurrent programs of video over a given "band" – primarily cable.

Over and over and over again in patents prior to the Burst patent, video compression is used to conserve bandwidth for normal broadcast delivery. So that more programs are available for viewing as they played back.. (See appendix for examples.)

■ Why the Burst patent changes all that

Bandwidth compressed vs. time compressed. One might think that these are the same facets of one thing. Not at all. If you do not have some means of storing (caching) compressed audio/video program information (and in a random access manner), there is no such thing as time compression.. You are only capable of (a) viewing the video as it arrives or (b) recording as it arrives. It arrives at only one rate – the playback rate. No time compression is possible. Compress the video and the provider can stuff more material in the "band" to the viewer – or user a narrower (cheaper) band. But it still all comes in at the SAME RATE.

Now hypothetically, it is possible to deliver pre-programmed (prerecorded) video in less time than the video takes to play – whether it is compressed or not. If there are dedicated, leased very-high-capacity fiberoptic lines between two places you can send very large files quickly. Hollywood does that now for review of its dailies shot on the set. Or, in an analog example, you could courier a three hour movie reel across town in fifteen minutes if the physical route/trip was short enough. But there was no way for the mass of home users to:

a) receive audio/video source material in a time compressed manner over a general purpose network in less time that it took to view it, and,

b) then adjust the video to improve its rendition and resnd to other areas within the home.

There was no patent that had as its primary goal, the use of a "time compressed representation" of audio/video in a consumer device.

The ‘995 patent changes all that. It makes time compressed video delivery possible -- direct to the users’ home. It invents an affordable audio/video transceiver that, for the first time, IS capable of storing digitally compressed video/audio program material in random-access storage such as semiconductor memory. And if the memory is loaded at a rate faster than the playback rate, then video time compression now becomes a reality. Users can acquire a program in less time that it takes to view it.

Thus, from the ‘995 patent, we find the following key claims:

"…compressing said digital audio/video source information into a digital time compressed representation thereof having an associated time period that is shorter than a time period associated with a real time representation of said digital audio/video source information;…"

"…An audio/video transceiver apparatus comprising:

input means for receiving audio/video source information as a time compressed representation thereof, said time compressed representation of said audio/video source information being received over an associated burst time period that is shorter than a real time period associated with said audio/video source information;

"…random access storage means, coupled to said input means, for storing the time compressed representation of said audio/video source information received by said input means; …"

"…said random access storage means comprises a semiconductor memory…"

"…editing means, coupled to said random access storage means, for editing the time compressed representation of said audio/video source information stored in said random access storage means and for restoring the edited time compressed representation…"

And the patent goes on to say:

"..A still further object of the invention is to provide an audio/video recorder utilizing a data compression technique for efficient storage, transmission, and reception of a digitized audio/video program over telephone lines or by other external digital means such as satellite transmission or reception…"

"…For example--a video program may be communicated at an accelerated rate from the first VCR-ET to a second VCR-ET in less time than it would take to view the program..…"

■ The Pre Burst world

At the time of the  application(s) for Richard Lang’s ‘995 patent, telecom was very different than today. What now seems obvious, even mundane, was practically science fiction.

More specifically, until that 1988 application, digital video and audio compression was used in telecom to conserve bandwidth. Many patent claims exist from that era regarding compressed video/audio bandwidth. Or, to put it another way, compressing data or signal so that it could be broadcast or transmitted over a network (or recorded on media) with an inherently limited bandwidth. An invention called time compression multiplexing existed, but as a multiplexing technique, it goal was still to use bandwidth more effectively, not to speed the actual delivery of audio/video programs.

But what the ‘995 patent does is make the claim, for the first time, of a representation of the audio/video in question that -- because of the configuration of the audio/video transceiver -- is time compressed, instead of bandwidth compressed and that can be sent over a general electronic or telephone network and that can be saved and edited as an object (i.e. a discrete file) in the receiving appliance. The ‘995 patent has as its goal that compressed video/audio information can be transmitted and received so rapidly that the time of transmission and delivery was less than the time that the piece took to play at normal speed (or, real time).

It changed the broadcast paradigm. So that a three minute song could be obtained by the audience in seconds. A feature length movie in minutes. How? And what is the difference between bandwidth compression and time compression?

■ Bandwidth as a river

Imagine bandwidth as a river. Depending on the width, the river could be coaxial TV cable, telephone twisted wire pair, fiber, satellite downlink or over-the-air transmission, etc.

Imagine audio and video as boats in the river.

Until the ‘995 Burst patent, digital compression was generally used to make the boats narrower because it reduced audio/video bandwidth (size) of each boat. This, in turn, conserved occupation, or consumption, of bandwidth. Why was this an advantage?

A) So you could send more boats down a given river at the same time. One beside the other.

Without compression, 20 boats side by side. With compression, 200 boats side by side.

Without compression, 20 program channels of video over coax. With compression, 200 channels over the same coax.

And, cable providers, for example, could charge more for 200 channels than 20.

B) Or you could put a bigger (better) boat in a given narrow river, that otherwise wouldn’t go.

Without compression, one rowboat. With compression, a cabin cruiser.

Without compression, a 160 x 180 pixel 15 fps video over twisted wire pair. Or an monaural song sample at 8K hz. With compression, 480 x 330 pixel 30 fps video over the same pair. Or the same song in stereo sampled at 44K hz.

And providers, for example, could charge users more for a better service, for higher quality audio/video -- while paying with the same infrastructure fees for use of the same limited bandwidth.

■ A new broadcast paradigm

Until the ‘995 Burst patent, the primary purpose of audio and video program material was to entertain, inform or communicate as it played to the audience on TV- from a roof top antenna, from a laser or audio disc, from a video cassette, from a coax cable. Or, to be recorded (or re-recorded in the case of the Go-Video VCR, a pre-Burst Lang invention) as it played. And the primary purpose of digital compression was to let more stuff -- or higher quality stuff -- be delivered to the audience -- broadcast -- over a given infrastructure at the same time. In real time. Just in time. At playback time.

Everyone was thinking about bandwidth capacity. No one was really focusing solely primarily about delivery time -- because, to the audience, delivery and play were largely the same thing. It was all structured to real time viewing at playback. Or real time recording at the time of playback. (Oh, you could go out and rent or buy prerecorded media to play, but that model was even less immediate. No over-the-wire or over the-airwaves delivery at all.)

But, what if you could actually deliver the program material, digitally, electronically, via some method of telecom broadcast, BEFORE it played. To the home. Over phone lines.

And what if you could actually have this broadcast/delivery process somehow accelerated so that program material was broadcast to, and recorded in, the audiences’ homes, in LESS TIME than it took to play it.

That’s what the Burst ‘995 patent does. It does what no one had thought to claim before. It invents:

(1) a new method of broadcast :

(2) a new type of consumer electronic device:

and, (3) a new type of digital editable, viewable electronic object:

It uses compression to not conserve bandwidth, but to speed delivery of a compressed downloaded file. It uses compression not to make the bandwidth act as if it were wider, but to act as if it were faster – to compress delivery time.

It uses compression to not make the boat narrower, but make it able to move down the river faster.

It uses the river, not to send more selectable/tunable channels of signal at once, but to download program material to the audience as fast as possible.

It configures a server and a client system with large high speed memory/storage/caches, audio/video compressor and decompressor, broad internal busses and fast connectivity to exemplify and make possible this new method of broadcast.

It is careful to say in the primary claims, "source material". It does not say "video programs." That means of course the patent applies to video program information of any length – from feature length movies, to clips, to snippets, to segments delivered in bursts. Thus, the patent envisioned internet streaming, in which a series of little program pieces were downloaded one after another into a section of memory. And because they each arrived at a rate faster than the program was played out of memory, the memory is never depleted until the final segment is done.

It is careful to claim "random access" for a variety of storage methods – including fast semiconductor memory. And, it is careful to make variety of clams to cover the use of an "information transfer network" and "one or more communications links comprises a telephone transmission line." And it makes claims for an internal "high speed bus." These claims all allow the patented invention to have sufficient performance to (1) quickly select, access and play one program out of many stored on the audio/video transceiver; and (2) accept the delivery of audio/video in packets over cost-effective telephone networks which are carrying other types of traffic – and which extend over the last mile to the user’s home– as opposed to a dedicated high capacity lines that are used only within the network.

It is careful to claim "audio/video source information stored in said random access storage.." so that the received audio/video becomes a saved digital object, a file.

It is careful to claim an "editing means" so that, for example, the video file can be adjusted for color and saturation to suit users’ preferences – just as all the major media players allow users to do today.

It is careful to claim both an integrated means for "monitor" and "RGB converter" so that video can be viewed directly on the transceiver without additional equipment such as a television.

It is careful to claim "transmission away from said audio/video transceiver apparatus." So that the transceiver can also be configured to be a file server.

■ Patent validity assessment

First point. You will note we repeatedly stress what the '995 patent claims. This is very important because claims, if sufficiently clear, are the primary and only means of enforcement. Thus prior art (prior patents), which merely in passing discuss or describe intellectual property do not necessarily represent what was known to persons skilled in the general art. And thus cannot necessarily enforce the described intellectual property on later patents. This is because courts cannot rely solely on upon "descriptions" in a prior patent to enlarge the claim of that prior patent to invalidate a later patent:

“…The claims are the only truly enforceable part of a utility patent, and they define the property right owned by the patent holder….Topic maintained by University of Maine School of Law….”

“…RULES OF CONSTRUCTION…1.3 Where the Language of the Claim is Clear and Distinct, the Courts Cannot Alter or Enlarge the Claim…”

“particularly 35 USC §112 (relating to the requirement …that the patent claims particularly point out and claim the invention)….”

 "In construing claims, the analytical focus must begin and remain centered on the language of the claims themselves." "If the claim language is clear on its face, the our consideration of the rest of the intrinsic evidence is restricted to determining if a deviation from the clear language of the claims is specified."

Second point. There is a presumption of  validity in the US patent system, just as there is a presumption of innocence in the US Justice system. Once granted, patents are presumed to be valid. The presumption re patents is, however, stronger since all patent applications  undergo a rigorous review by the US PTO before being granted. This is analogous to having your tax return audited each year.

Third point. There are three key measures by which the validity of all patents is assessed. These ask the following questions:

(A) – is the invention truly novel?

(B) – is the invention not obvious to someone who is skilled in the general subject matter of the invention?

(C) – can the invention be practiced without undue experimentation?

The ‘995 patent appears to pass all three tests with flying colors. At the time of  application it seems clear it was (over and above the mere prior art cited by the patent):(a) novel and (b) non-obvious. Why?

One. As referenced in the Appendix, there is an exceptional uniformity of claims in preceding patent art focusing exclusively on bandwidth compression, which clearly differentiate this collective art from that of '995, which focuses on time compression.

Two. There is similar and corresponding a paucity of the '995 patent's claims (as described in this treatise) of  (a) time compression, (b) editable audio/video object (c) random access audio/video storage and (d) player with integrated video display, in preceding patent art, which clearly sets apart '995 from preceding patent art.

Three. The first formal primary claim in '995 -- that of a time compressed, editable, integrally viewable, random accessible and general-network-deliverable representation of audio-video source information -- requires a very specific, sophisticated and detailed set of hardware, software and network platform requirements that are also claimed in '995. This specific set of requirements is found nowhere in preceding patent art in any small part, let alone as the entire set.

Four. This combination of the "representation" claim, the "transceiver" platform claim and the savable/editable object claim in one patent further adds to the novelty and on-obviousness of the intellectual property claimed in '995, because it integrates not one but three novel ideas.

And Burst’s (then, Explore Technology) a/v transceiver demo system at the Consumer Electronics Show at Las Vegas in 1991 proved it was (C) useful and feasible to perform the invention without undue experimentation.

CONCLUSION

It is my further opinion that, even if we put novelty, etc.,  aside, because of the sophisticated and complex interweaving and interdependency of the '995 claims, it will be extremely difficult to argue invalidation of any single one claim, let alone a larger number.

Infringers ignore these facts at their peril.

-- Td

■ Appendix: Until 1988 audio/video compression focused on conservation of bandwidth to enable more program material

Andrews & Pratt, "Television Bandwidth Reduction by Fourier Image Coding," Paper, 103rd Tech. Conf. SMPTE, May 5-10, 1968. .

Andrews & Pratt, "Television Bandwidth Reduction by Encoding Spatial Frequencies", SMPTE Journal, vol. 77, No. 12, Dec. 1968, pp. 1279-1281.

Pratt & Andrews, "Application of Fourier-Hadamard Transformation to Bandwidth Compression", Proc. Polytechnic Institute of Brooklyn, 1969, pp. 515-554.

H. Whitehouse, E. Wrench, A. Weber, G. Claffie, J. Richards, J. Rudnick, W. Schaming, J. Schanne, "A Digital Real Time Intraframe Video Bandwidth Compression System", SPIE vol. 119, Application of Digital Image Processing, IOCC 1977, pp. 64-78.

4,394,774 Widergren , et al. July 19, 1983
...Another object of the present invention is to combine novel circuits and subsystems into a digital video compressor and expander which effectively compresses the bandwidth of a television picture in accordance with novel methods and techniques.

4,665,436 Osborne , et al. May 12, 1987

…A method and apparatus for compressing digital data for transmission and reception in a narrow bandwidth…It is an object of the invention to optimize the efficiency of digital data communications by reducing the number of bits to be transmitted and reducing the required transmission bandwidth, without deterioration in transmission quality.

(Revised '995 claims filed 3/12/1990).

#####################