Canon purchases 24p license

[Simon Wyndham]

Thing is though, does that mean that everyone with a telecine machine has to pay a fee to him too? The more I think about this, the more I think this patent is absurd.

[Charles Papert]

I believe his patents are for reverse telecine--60 to 24, whereas telecine is 24 to 60 (traditionally speaking--of course these days there is one to one 24 fps film to 24p DVD)

[Simon Wyndham]

So why did Panasonic need to pay him for the Varicam? Also, all of the cameras that do 24p lay down the footage in PsF form. Ie 24p to 60i, exactly what the patent is for. So telecine comes under that banner too depending on the transfer. Film to 60i video for example.

[Charles Papert]

I can only theorize, I'm not an expert on this. I do know that the basis of the patent specifies video as the originating format, excluding film, which is why telecine is not involved.

[Glenn Chan]

I think you can read the patent here:
http://patft.uspto.gov/netacgi/nph-P...S=PN/5,335,013

SUMMARY OF THE INVENTION

The present invention provides a method and apparatus in which a video camera produces a video signal that has the aesthetically desirable appearance of film but with the economy of video technology. The video camera has an imaging device on which a lens forms an image. The imaging device is typically a solid state charge-coupled device (CCD). A circuit provides timing and synchronization for sampling the imaging device to generate an analog-sampled video image in a 24 progressively-scanned frames-per-second format where each frame comprises 625 lines. The sampled analog video image is processed through a video amplifier and converted into a digital-sampled video image by an analog-to-digital converter. A grain simulator provides random noise with filtering which simulates grain pattern imagery. Preferably, the noise is digital pseudorandom. The random noise is added to the digital-sampled video image and modified by a gray scale modifier. The gray scale modifier preferably includes a programmable read-only memory (PROM) for bit mapping the digital-sampled video image onto a gray scale curve reflecting the non-linear characteristics of photographic film. Circuitry is provided for converting the gray scale modified digital-sampled video image into conventional video formats, The conventional video formats typically include either analog or digital 525 lines/60 fields or 625 lines/50 fields video formats. In the present invention, for converting to the 525 lines/60 fields format, a circuit converts the 625-line to a 525-line format by interpolating adjacent lines that have been weighted. Circuitry then converts the 525-line format from progressively-scanned to interlaced-scanned format. A replication of the third field is added after every fourth field. The resulting digital output is in a 525 lines/60 fields format that can be converted to analog by a digital-to-analog converter. For a 625 lines/50 fields output, the gray scale modified digital-sampled video image is converted from progressively-scanned to interlaced-scanned. The resulting output is a digital output in 625 lines/48 fields format. Alternately, this digital output can be transformed by a digital-to-analog converter into an analog output in a 625 lines/48 fields format. Conversion to a 625 lines/50 fields output is accomplished by recording on a video tape recorder operating at approximately a 4% slower speed and playing back the tape at normal speed. An alternate embodiment of the present invention employs a 25 frame per second clocking for direct conversion to 625 lines 50 fields format.

The patent document also talks about 3:2 pulldown during telecine.

Faber has some other patents, listed on his company's website.

Good lawyer too.

It might be the threat of good lawyers that keeps these companies licensing the patent instead of risking a lawsuit. I'm guessing part of it is because people don't know what the outcome of a case will be (think about it: if people knew, there'd be a lot less cases going to court!).

[Peter Ferling]

The purpose of a patent is sharing of knowledge in exchange for exclusive, first use rights. After the patent expires, the knowledge is already a matter of record and part of the vast data pool for others to spring off from.

Generally, it's a favorable win-win situation. As an engineer, I often consult the patents to ensure that we don't infringe. There are plenty of folks out there whom merely conjure up ideas and commit them to paper in hopes of snaring someone whom actually got their hands dirty, inventing the same thing.

In many cases it can be a long legal battle to see whom truely owns the idea. That idea should go to the one whom did the work, and not merely hire a crafty lawyer and be first to file. Yet, it's cheaper to just pay a fee.

Sure you can't patent a "tire", but you can patent a "tread" design for a tire. However, at the time the patent was issued, if the concept was novel, it's patentable. Some of you may laugh, "24p, indeed", but if this guy was the only one to file a claim, and had device or process in place to back it up, then it's reasonable to see why it was granted.

[Jeff Kilgroe]

Originally Posted by Simon Wyndham

What I find odd though is that Fabers original software converted 60i to 24p. How does a patent like that affect a camera that has a true progressive scan CCD or CMOS and records to solid state or other non-tape format in true progressive scan?

His patent is worded to extend coverage over any form of 24FPS imagery that is captured by electronic means. It's actually a very lengthy and word-heavy patent that is very generalized and all-encompassing. In other words, it's a steaming pile of BS. It all comes down to having a pretty good patent attorney that knows the system and having it reviewed by patent clerks that are uneducated (for the most part) on the technical aspects of what he was applying for. His method of 60i to 24p conversion wasn't the first, nor is it unique, but he was the first to patent with it. This is the bulk of his patent and obviously why it was granted. However, the additional fluff and items included in order to protect his intellectual property are where it gets sticky and the patent goes beyond protecting his conversion method and includes any form of electronic and/or digital means of aquiring 24P video. Sony fought him in court and settled... Even though Faber's patent may be a bunch of crap in regards to most 24p implementations, the fact is the patent was granted. In the eyes of the courts, he holds the patent on such a process, no matter how generalized and can demand compensation.

Faber probably wouldn't be so poorly thought of if he handled the situation differently. But nearly every 24p license is handled confidentially and rumors abound of each company paying differing sums, related to their perceived net worth or financial muscle rather than a legitimate licensing and pricing scale. It's extortion at the fullest extent of still being legal and even that is questionable... But it's cheaper and easier to simply pay than it is to fight it out.

[Jeff Kilgroe]

Originally Posted by Simon Wyndham

So why did Panasonic need to pay him for the Varicam? Also, all of the cameras that do 24p lay down the footage in PsF form. Ie 24p to 60i, exactly what the patent is for. So telecine comes under that banner too depending on the transfer. Film to 60i video for example.

Any device that records 24p into a 60i stream or 25p into a 50i stream, etc.. or vice versa is covered. The Varicam records 24p (and other frame rates) into a native 60i stream.

The sneaky guy also managed to squeeze in coverage for direct 24p recording and playback and direct acquisition of 24p for processing into any other format. So Panasonic had to pay for Varicam, DVX100, HVX200... Sony had to cough up for the CineAlta and even the CF24 modes of the FX1/A1 camcorders.

[Joe Carney]

So when does it expire? Patents are good for 17 years if I remember correctly.

[Jeff Kilgroe]

Yeah, technical patents are 17 years by default, but they can be renewed under various circumstances or new patents can be filed to perpetuate and protect previous patents. Faber's latest patents surrounding 24p and held by his companies (24P, LLC, etc..) are dated '94 and '96. So there's at least 8 more years to go, provided he doesn't keep finding a way to renew or augment them with new filings.

While we're on the subject of frivolous patents, check this one out... It should be classified under the "should never have been granted" category. Nothing truly unique and very poorly documented

[Ash Greyson]

There is no real world difference between 24P native or "real" 24P and 24P progressive pulled out of a 60i stream. The Vari shoots everything at 60p. Someone at Canon told me a while back that the license is specific PER CAMERA, meaning you can buy a blanket license or pay a per unit. There is some formula. The XL2 is a REAL progressive camera that uses the exact same pulldown process as the DVX and even the higher end cameras that embed 24P in a 60i stream.

I understand that the HD-SDI also has a VERY expensive license but it is associated with audio and video. I was told it increases the price per unit by about $6000. That is why you have no audio over SDI on the XLH.

Maybe Canon is purchasing it for a GLH1 camera, or the XLH2 but dont be confused about the XL2 which is a real progressive 24P camera.



ash =o)

[Greg Boston]

From the summary that Glen posted...

A grain simulator provides random noise with filtering which simulates grain pattern imagery. Preferably, the noise is digital pseudorandom. The random noise is added to the digital-sampled video image and modified by a gray scale modifier.

I guess Canon decided not to play ball and this is why they removed the film grain from the XL2. Probably had a confidential settlement based on the number of units sold to that point. I remember when they removed it the reason was, 'because the software is no longer available'.

Hmmm...very interesting indeed.

-gb-

[Daniel J. Wojcik]

Originally Posted by Jesse Redman

Patents are almost universally rejected by the patent office. It often takes several rounds of "explaining" a patent to the patent office before a patent is issued. This often takes two to four years of proving to the patent office that your idea is unique.

And yet they always manage to allow things like this: http://patft.uspto.gov/netacgi/nph-P...&RS=PN/5443036

[Jeff Kilgroe]

Originally Posted by Daniel J. Wojcik

And yet they always manage to allow things like this: http://patft.uspto.gov/netacgi/nph-P...&RS=PN/5443036

ROFLMAO!

Any patent clerk that stamped his approval on that submission should be fired.

[Tony Tibbetts]

Didn't Canon also acquire the AVCHD license? Which has 720p 24fps specs? Maybe a solid state GL3/GLH1. Hmmm... that could be very cool.