From The Street...
Tuesday, March 26, 2002
Foveon - National Semi - Gilder
"No camera company will be able to succeed without using Foveon chips any more than a PC company could succeed without using silicon semiconductors." G. Gilder
More of Gilder's recent comments on Foveon and National Semiconductor snipped from his Telecosm Forum. (subscription). Gilder Technology Report: Home
===================
This morning at 9:30, with an appropriate delay after the Foveon letter, I made my first new public company investment since Global Crossing. I put $250K into National Semiconductor. People on this board have been asking what effect Foveon will have on National. My view is that National will become Foveon--that within five years Foveon's revenues and market cap will be at least five times the current National levels, that Foveon will dominate the 20 billion dollar camera imager market the way Synaptics dominates the far smaller touchpad market. Unlike Global Crossing in 1998, National has virtually no debt and some $700 million of cash.
--GG
====================
On my noggin at the moment are National Semi, Corvis, Avanex, Essex, Mirror Worlds, Scale 8, and Narad Networks. The last three are hot but private. The optical companies may not revive for another year or so, since in my view the government has wrecked the telecom sector for the next couple years. If I am right about Foveon, National is the single best investment in all technology.
====================
I am in the middle of the next letter, so I can't explore the subject in depth at the moment. But Foveon has created a fundamentally superior image plane, on CMOS with three buried bipolar junctions to collect all the color image information instantly at every pixel (no digital processing and thus good for videocams as well). The quality is better than film and it enables an entirely solid state, still and motion camera in one box. I cannot estimate the market for such a product, but I can tell you that there is no real competition at present. Could there be competition in the future? I cannot predict the future of three dimensional optics. But I seriously doubt that anyone can excell Foveon for two dimensional photography. As far as I can see, the solution is as elegant as it can be: one chip three colors, no delay, at resolution that saturates the eye. What more do you want?
====================
Foveon has all the patents on the invention, including the imager patent portfolio from National and Synaptics which these two companies assigned to Foveon when it was founded. Beyond mastery of the black arts of Foveon's process, National also possesses patents on generic analog fab technology used in producing the Foveon device. But sharing those patents are all participants in National's patent sharing agreements with myriad other semiconductor companies such as Intel, TI and Sony. In the end, before the standard becomes universal in cameras, the manufacture of the devices will need to be second sourced anyway, but the invention will stay with Foveon. The situation will somewhat resemble Qualcomm's CDMA patents and second sourcing arrangements
===================
The chip is cheaper and better and offers the possibility of solid state cameras that do both stills and video. The resolution will be determined by the printer, as with any camera, but the information from the Foveon will be correct, without distortion or alias, so the pictures and colors are better regardless of resolution. No camera company will be able to succeed without using Foveon chips any more than a PC company could succeed without using silicon semiconductors.
=====================
Monday, March 25, 2002
Automation and Digital Asset Managemnt...DAM
Since my brief fling with the Thomson/Technicolor folks late last year, interest in DAM has blossomed like the spring daffodils - all over da lawn, as we say in Chicago.
Some recent discussions with several automation and server companies show that they find DAM attractive as Their Next Step.
First, Automation.
Ownership Deregulation, Economies of Scale and The Freeing Up of The Market were pushing the Consolidate Operations bandwagon pretty fast, and interest in CentralCasting picked up considerably. But post 9-11, some now think a centralized location of The Whole Network might not be such a good thing. Automation to assure both on-time, flawless switching and Manpower Savings is laudable, but Planning For The Worst is now more common. First step is to get Hubs to be runnable from Other Hubs - if all the Hub people go away, God forbid…
Broadcast automation is no longer a proprietary business. Many suppliers are now attacking the known market to control distribution servers, a limited business. ($200-300M in servers? Control is 25% of that? Broad Assumptions…). And, as most all "video" distribution is now done by servers, the once-proprietary secrets of synchronous machine control are now well known and used by the (6-10 ?) companies in this market. Three smart guys and NT/Windows 2000 PCs are not high barriers to entry, all selling to a printed list of broadcast stations; and the pie they seek to get a piece of, a fixed and relatively small one, is quickly becoming commoditized.
Server manufacturers - and most automation software suppliers - like to remain "agnostic" - which means they say their servers will work with most standard control schemes; and the ASS's feel their software has to work on all popular server solutions. So lots of energy and time is spent working on the interfaces to everything. Not unlike the video editors of old? Some server suppliers are considering taking automation software in-house and incorporating it as part of their total product…sort of adding it to the motherboard….
Users - in this case broadcasters - have been buying and using hardware boxes for years. They sometimes mistakenly view software as something that can be made to do anything, even after it is "built" and delivered, with new features added, in ways never possible with hardware. Well, not exactly….
Large customers are demanding. Yet many suppliers prostrate themselves and promise anything in order to get The Big Network Deal, with the inner hope that it's all really be possible to do; or like their customers, they say, "we'll fix in post" - after the order is accepted. Lots of promises are made; some can't be kept. Making Stuff Do What The Salesman Said is the Bane Of Engineers Everywhere. Some companies that promised all sorts of goodies are slowly digging themselves out of holes like this; some are digging new holes because their goals are short term top line revenue, or "market share" - a now somewhat discredited notion that you can lose money on some deals to get more business.
The Big Order then often goes to the Folks That Promise The Most. These are - also often - smaller companies that need the "visibility" they believe accompanies A Big Network Order. They can sell similar "features" at half price, not being burdened with Big Corporate Overheads and Support Organizations. But Big Never-Been-Done-Before-Systems for a Highly Visible Customer can be a Resource Killer. More than a handful of companies have Foundered/Floundered (take your pick) on the Shoals of such deals; and many say a big order from a big name customer not only will not be "profitable", but by tying up limited engineering time and efforts, Permanently Injure the Firm,
Users can exploit the suppliers to add their specific features, and with onerous terms - "free updates for 10 years.."; again mistakenly believing these can be done easily, and for free, or as part of "Your guy promised me it would do this…". Hmm…
Can a little/few resources company survive a contract with the Big Guys. Several current deals in process will provide some empirical evidence to answer this.
Running components of a system synchronously and switching on time and keeping track seem to be problems largely solved. The need to get deeper into the manipulation and "management" of what's being switched to add some value seems to be the next step. As distribution evolves from framed analog video and audio, to serial digital streams and files; and point-to-point routers give way to data stream routing, the start and end points of a circuit depend on what's in the content - or "essence" as the "information inside" is now called. Managing content seems like a good place to go, and there is general agreement to develop standards for data about the data , or metadata, as a first step towards the Holy Grail -"Interoperability Over Heterogeneous Networks". Whew…
Rest period…..more later on DAM.
Wireless Access...
I've used the Apple Airport/WiFi for several years at home - with an iBook, a PowerBook G3 (using the Farallon PCMCIA card), and the G4 TitaniumPowerBook. When the kids visit, they use my DSL service by plugging in the PC card into their laptops.
On a recent trip, I discovered an 802.11b WiFi network that gave me wideband access from the lobby of Leitch in Burbank! It took them awhile to find the engineer inside doing it. Since then, I've been airports and Starbucks that let you get on for a membership fee, a one time use charge, or other schemes. In the Bay Area there is a group, not unlike Amateur Radio Enthusiasts, who add power and put up antennas to extend the range of their "station". Here's more information, including links to a map:Bay Area Wireless Users Group
I signed up with the one servicing Starbucks stores: MobileStar
A more extensive list of links for all kinds of wireless service in the Bay Area is here: Bay Area Wireless Internet
Ideavirus...
A great extended essay/short book on why and how ideas catch on and spread is available on line - or as a paperback. Lots of similarities to viruses.
This 197-page "essay" is a PDF file of about 850K. It also available as a book through several sources. Free download here; be patient. IdeavirusReadandShare.pdf
It's probably easier to buy the paperback through Amazon, but if you've mastered the art of reading extended pieces from your screen - or are stuck somewhere without Web access, like an airplane or airport - go get it and plough through.
Worse case, email me and I'll send you the PDF file.
Thursday, March 14, 2002
Writing well...
Most everybody has to write now. From emails to presentations. Doing it well, or at least better, pays off. Reading bad stuff is painful and detracts from the thought you're trying to get across.
Brother Tom expands on the basics of writing - He's a working journalist who has also has taught Journalism in Dublin and in Cedar Rapids. TomW@fyiowa.com. Excerpts from a recent chat -
=====================
One of the first lessons is "the shitty first draft" as a requirement of the process.
Writing is something that gets easier (and better) the more you do it. But you have to do it "badly"" in order to get comfortable with doing it well. Accept that most of what you write, until you understand the process, will be marginal at best.
Clear. Concise. Correct. Those are the "gold standards" .... unless you're Joyce or Faulkner. But that's another league of writing. What I do every day is take A LOT of information and reprocess it so that it makes sense to the average mass media moron who wouldn't know how to begin to sort it all out.
Lesson two is "Write backward from the reader."
Lesson three is: Good writing is invisible.
Lesson four is: don't raise any questions that you don't address.
Lesson five is: Find a way to say in one sentence of 12 words what you've just written as three sentences of 20 words each.
Lesson six is: Check everything. Twice.
You must have some feel for who is reading this stuff and then you can structure it in terms of their levels of understanding of your subject matter. Mass media audiences that I write for are morons -- 6th grade reading level, maybe.
I do three things after I draft a story. One check for grammar, punctuation, spelling and style. One check to shorten sentences (make long sentences into two or more sentences). And one check to make sure I've answered all the questions the story addresses (if I don't have answers, I don't suggest, much less pose, the questions).
Before you sit down to write anything, imagine two guys in a bar, one of whom says "I didn't have a chance to read that thing Jim Walsh wrote. What was it about?" The other guy has read it. Okay: What do you hear him saying? What are the next 10 words out of his mouth? That's your key message.
Focus on it. Make the key point quickly and explain as required. Don't use it as a punchline. Don't task the reader with trying to figure out what the fuck you're talking about: They won't bother to make the effort.
That's today's lesson. Tune in again tomorrow.
What I do is not an art, but a craft. One of its advantages is the longer you do it, the easier it gets. I've been doing it for 37 years.
===============
This is from one of a series on writing for blogs. http://www.alistapart.com/stories/writebetter/
"The most essential gift for a good writer is a built–in shock–proof shit-detector."
Tuesday, March 12, 2002
Hardware - then software - then services ?
As the broadcasting business "evolves" and "converges" on both the supplier and customer sides, the direction seems to be this: Most television applications can now be done on non-specific hardware, generic, general purpose computers. With faster CPU speeds and cheaper storage, value added has moved to application software. Hardware is now a commodity. Apple's editing software and G4 PowerBook have become a cost effective editing system for such big users as CNN, who save lots on sending fewer Betacam stuffed anvil cases and their handlers with each reporter.
But, as the referenced article details, software, as a business, is subject to some of the same marketplace realities. It seems the software business is like the box business..where end-of-quarter dialing-for-dollars is the norm.
From an interview with Netscape and Loudcloud founder Marc Andreessen:
Economist.com
"...For a start, says Mr. Andreessen, he did not want the new firm to be in the business of selling software to large companies. The trouble with selling software, he says, is that software vendors end up in an adversarial relationship with their customers. The customers know that the vendor needs to meet its quarterly sales target, and is often dependent on a few large sales to do so. The result is a frenzy of last-minute negotiation and price-cutting at the end of each quarter. An industry joke tells of the chief financial officer of a software firm who calls down to the loading bay on the last day of the quarter to ask how many units have shipped. "Too early to say," comes the reply. "It's lunchtime, so we're only halfway through the quarter." [ We all know video execs who do this…JMW]
The situation at AOL is different. Because AOL is a service business, its revenues are predictable. Once a month, it collects $20 or so from the tens of millions of customers to whom it provides Internet access. Mr. Andreessen decided that his new firm should also run on a service model...."
======
Is the "service business" better? Is broadcasting to become an on-line application? "We'll deliver your message to all these people, both on-line and off-air…and we'll do the hardware and networking and let you, the content owner, participate in the revenue models, whether advertising or subscriptions or PPV…"
Maybe, if grownups do it...
Not Rocket Science…..
My brother Tom is a writer. He's lived in Iowa City for over 30 years. He just did a piece for the Cedar Rapids Gazette (Cedar Rapids: Gazetteonline.com) on James Van Allen, noted space physicist at the University of Iowa, and a pioneer in America's first efforts in space, post Sputnik. The Van Allen Belts, bands of radiation surrounding the earth, are named after him.
Allen, now a spry 87, liked the article and the sidebar snips of video clips of an interview with him. These were compiled and given to him on a VHS tape. His wife called Tom and wondered how to run the tape and who could do it for them.
Tom began, "Well, it's not rocket science…"
I bet he's been waiting for years to say that.
Saturday, March 09, 2002
Digital Television...
From colleagues who make a full-time living in this business:
The TV stations' focus is now on compliance….and the Transmitter part is where the action is. Most middle market stations have a five-year plan for the millions in Capex necessary to comply with the FCC timetable to "go digital":
Year 1.) Tower 2.) Transmitter 3.) Master Control 4.) Automation 5.) Local programming production equipment, if justified by sufficient HDTV owners/viewers.
One manufacturer said it had 12-15 antennas and six transmitters in storage awaiting tower completion.
Local affiliates now, generally, switch between Local in the day, and Network at night for DTV. No measurable viewers means no need for advertising and HD production. And indeed, despite the sales of all the transmitters and towers, there is no revenue model that works for DTV owners. The estimated power cost by one manufacturer is $3K/month for a 50KW DTV transmitter. In fact, a recent FCC word was that it's okay to lower power levels now, and increase them later. This only exacerbates the 8VSB limited reception issues.
The top ten markets generally are in place; the middle markets are now in a rush to get the minimum done. This is not unlike the buzz to get the first ATSC MPEG encoders a few years back. Suppliers "dropped trou" to get their original units placed; now many suppliers are no longer in that business, and prices have plummeted. Stations were clear that they were buying the minimum; they only wanted what would get them by. But the "serious" capital expense for a digital transmission system raises the dollar stakes by several orders of magnitude, and all without a corresponding revenue generation scheme. Most stations are owned by folks with other media interests, all demanding investment money; only the "visionaries" who own their own places can take the technology plunge and go all HD, like WRAL in North Carolina.
I tend to agree with the polemics of Craig Birkmeier on OpenDTV about government regulations being the root cause of most of this evil; and the fact that entrenched incumbents are operating in their best short term interests to thwart a disruption of their established businesses. This applies to telcom as well as broadcast television. And, listening to the regulatory limitations demanded by Hollywood spokesmen like Jack Valenti, on digital distribution of their stuff, the chorus is complete. Please "protect" our franchise from disruption by crippling innovation and market dynamics now enabled by digital technology, broadband distribution, and widespread connectivity. With big bucks to contribute to political campaigns, and influence on how the word gets out to hometown constituencies, Hollywood and Broadcasting are fighting to keep the good times of a departed era, rolling, by imposing restrictive legislation. But the telecosmic paradigm and the markets will prevail. "Listen to the technology…"
Friday, March 08, 2002
Digital broadcasting…
From colleagues who make a full-time living in this business:
The TV stations focus is now on compliance….and the Transmitter part is where the action is. Most middle market stations have a five-year plan for the millions in Capex necessary to comply with the FCC timetable to "go digital":
Year 1.) Tower 2.) Transmitter 3.) Master Control 4.) Automation 5.) Local programming production equipment, if justified by sufficient HDTV owners/viewers.
One manufacturer said it had 12-15 antennas and six transmitters in storage awaiting tower completion.
Local affiliates now, generally, switch between Local in the day, and Network at night for DTV. No measurable viewers, means no need for advertising and HD production. And indeed, despite the sales of all the transmitters and towers, there is no revenue model that works for DTV station owners. The estimated power cost by one manufacturer is $3K/month for a 50KW DTV transmitter. In fact, a recent FCC word was that it's okay to lower power levels now, and increase them later. This only exacerbates the 8VSB limited reception issues.
The top ten markets generally are in place; the middle markets are now in a rush to get the minimum done. This is not unlike the buzz to get the first ATSC MPEG encoders a few years back. Suppliers "dropped trou" to get their original units placed; now many suppliers are no longer in that business, and costs have plummeted. Stations were clear that they were buying the minimum; only what would get them by. But the "serious" capital expense for a digital transmission system raises these dollar stakes by several orders of magnitude, and all without a corresponding revenue generation scheme. Most stations are owned by folks with other media interests, all demanding investment money; only the "visionaries" who own their own places can take the technology plunge and go all HD, like WRAL in North Carolina.
I tend to agree with the polemics of Craig Birkmaier on OpenDTV about government regulations being the root cause of most of this evil; and the fact that entrenched incumbents are operating in their best short term interests to thwart a disruption of their established businesses. This applies to telcom as well as broadcast television. And, listening to the regulatory limitations demanded by Hollywood spokesmen like Jack Valenti about digital distribution of their stuff, the chorus is complete. Please "protect" our franchise from disruption by crippling innovation and market dynamics now enabled by digital technology, broadband distribution, and widespread connectivity. With big bucks to contribute to political campaigns, and influence on how the word gets out to hometown constituencies, Hollywood and Broadcasting are fighting to keep the good times of a departed era, rolling by passing with restrictive legislation. But the telecosmic paradigm and the markets will prevail. "Listen to the technology…"
Sunday, March 03, 2002
Tech Retreat Report...and new color camera technology...
Mark Schubin, iconoclastic yet charming host of the Hollywood Producer's Association Technical Retreat in Palm Springs last weekend, has posted some notes in the 2 March 02 edition of his regular Weekly Memo (usually out on Mondays). These are archived here, on digitaltelevision.com. Mark Schubin's Monday Memo.
I got lucky and won a Nipper, a stuffed toy dog RCA trademark, courtesy of my favorite French Media Conglomerate. They have now completed the purchase of GVG, and have also bought Vidfilm, or so I hear. Leitch released their quarterly numbers last week. New CEO Craig has a real challenge. Cash way down. Investments written down. Still real high Sales and Admin expenses, it seems. But I digress....
One high point of the event was Dave Bancroft of Thomson talking about digital imaging - beginning with explanations of how both Film and Television are Broken. Brilliant, erudite and always humorous, he was a great counter balance to some of the other presenters who seemed to take themselves pretty seriously. If his presentation gets posted, I'll post a link here.
Across the country in Orlando at the same time, Foveon (Foveon - News) was announcing the productization of their X3 color imaging technology at a commercial photography convention. Carver Mead, famous professor from Cal Tech who helped "invent" VLSI, started this company to develop his "massively parallel" analog imaging technology. Foveon sold a bunch of very high end professional implementations of this last year, but now a "better, cheaper, faster" version has been announced by Sigma (SIGMA News). Obvious question, can it do real time video? Benefits are lower costs and complexity, and the use of standard interchangeable lenses. The press release seems to indicate that the camera can do both High Res stills and SD level PAL or NTSC video. See their site above for specific numbers.
Questions from Smart Industry People (i.e. competitors) include signal/noise challenges in trying to gather enough low level signal from the CMOS-based device; and colorimetry. Different colors are detected and picked up at different vertical levels in the silicon. (See the good charts on the Foveon site (Foveon - Why X3 is Better). It looks like "detection depth" would have to be tuned to get the "right" mix for a particular color gamut. In current digital RGB chip implementations, filtering can be done to tweak this "nuts on". However, such a solution would sure simplify (read: cut the costs a lot for the...) lensing and optics.
If I were a Japanese or Dutch-French camera guy, I'd pay close attention to these folks. The barriers-to-entry costs seem to be dropping ...They are literally swamped with interest since the announcement: "...a real tsunami", said a company officer.
