Someone must have taken apart a new camera

[Bruce Allen]

You mean like Canon, Nikon, Silicon Imaging and Vision Research?

Haha yes! Whom you should add compressed Bayer RAW files before RED did. And in Silicon Imaging's case, compressed RAW video for cinema purposes before RED did.

I am re-entering this thread with one more important thought:

If this patent BS happens with light field camera stuff, I am going to be mighty pissed off.

Seriously, for the good of filmmaking, does anyone want to come in with me and:

1. Create light field camera (just put some microlenses on an existing sensor... or even a printed grid can work)

2. Do all of the obvious stuff NOW before stupid patents go down.

First priority for me would be lightly-lossy video compression techniques that take into account not just spatial and temporal continuity but also continuity across micro-lenses and are designed to be decompressed by GPU-like architectures.

Anyone else in? I am seriously down to put in some funds and time.

I remember sitting in Pietermaritzburg, South Africa in 2001, marvelling at Canon's DSLRs with my friends and of course after 5 minutes we were like "Wow, these 35mm sensors will be great for video too. You'd want to record the raw of course. And compress it a little to make it fit." I remember talking over this with a friend when he was chatting about buying an old Mitchell 35mm camera and the consensus was it was better to get on the digital thing because this was obviously coming down the line.

I don't want future advances in cinema being hampered by someone telling me some painfully-obvious extension of existing technologies is patented. Seriously, let's build some stuff, create some prior art and stop this horror.

Bruce Allen
www.boacinema.com

 

[Rob Anderson]

Atta boys. With all of the positive Sony mentions lately, I was beginning to think RED may have lost a bit of bite on the end of their knife. Glad to see the REDs are still storming.

That said, this is going to get ugly. Buckle up.

 

[Wayne Morellini]

Pity the other thread was closed, because under the potential list of damage remedies that Red was seeking, was an amount equivalent to loss sales. Sony should jump at that (because they hardly dented Red's sales). Patent stuff is just too much. Destroying all cameras, wouldn't it be better to replace the offending part?

 

[Nick Morrison]

Jim Jannard said here on the forum that he never lost a single law suit just when he sued ARRI.

Well ... There is a first for everything, isn't there?

Peter

Technically he didn't "lose"; just no one "won".

 

[Jeff Kilgroe]

I think trying to academically discuss the substance of this law suit is pointless. Until RED officially presents their case, we don't know what they're suing for. Unless it's a totally, blatantly obvious infringement, they're not going to show their cards until the first hearing or until both sides sit down with an arbitrator, if and when it comes to that. Jim has referenced in the past the REDCODE and REDRAW patents and how they might be a problem for other camera makers. I guess we'll see what happens... and this could take a while. My assumption is that it will settle out of court, as the vast majority of these cases do. And a settlement is not a loss in legal terms. Often times, it is the most beneficial outcome for all parties involved, even if not everyone is happy with it.

We don't know the terms of the Arri settlement. I wouldn't be so quick to chalk it up as a loss as it seems a lot of people want to do.

 

[Nick Morrison]

I think trying to academically discuss the substance of this law suit is pointless. Until RED officially presents their case, we don't know what they're suing for. Unless it's a totally, blatantly obvious infringement, they're not going to show their cards until the first hearing or until both sides sit down with an arbitrator, if and when it comes to that. Jim has referenced in the past the REDCODE and REDRAW patents and how they might be a problem for other camera makers. I guess we'll see what happens... and this could take a while. My assumption is that it will settle out of court, as the vast majority of these cases do. And a settlement is not a loss in legal terms. Often times, it is the most beneficial outcome for all parties involved, even if not everyone is happy with it.

We don't know the terms of the Arri settlement. I wouldn't be so quick to chalk it up as a loss as it seems a lot of people want to do.

I agree, there is so much unknown right now. The best we can do is speculate. What is a fact is RED has promoted and pioneered 4K for years (in face of strong industry-wide skepticism)...only to find their stubbornness rewarded with 4K becoming a "new standard".

In so doing, it is safe to assume that RED's "tech" has been carefully picked over and scrutinized, and if any of their patents and hard work have been borrowed and/or stolen in order to facilite a faster level of parity, then they certainly seem entitled to go to to court and ask to be remunerated for their inventions.

 

[Björn Benckert]

Haha yes! Whom you should add compressed Bayer RAW files before RED did. And in Silicon Imaging's case, compressed RAW video for cinema purposes before RED did.

I am re-entering this thread with one more important thought:

If this patent BS happens with light field camera stuff, I am going to be mighty pissed off.

Seriously, for the good of filmmaking, does anyone want to come in with me and:

1. Create light field camera (just put some microlenses on an existing sensor... or even a printed grid can work)

2. Do all of the obvious stuff NOW before stupid patents go down.

First priority for me would be lightly-lossy video compression techniques that take into account not just spatial and temporal continuity but also continuity across micro-lenses and are designed to be decompressed by GPU-like architectures.

Anyone else in? I am seriously down to put in some funds and time.

I remember sitting in Pietermaritzburg, South Africa in 2001, marvelling at Canon's DSLRs with my friends and of course after 5 minutes we were like "Wow, these 35mm sensors will be great for video too. You'd want to record the raw of course. And compress it a little to make it fit." I remember talking over this with a friend when he was chatting about buying an old Mitchell 35mm camera and the consensus was it was better to get on the digital thing because this was obviously coming down the line.

I don't want future advances in cinema being hampered by someone telling me some painfully-obvious extension of existing technologies is patented. Seriously, let's build some stuff, create some prior art and stop this horror.

Bruce Allen
www.boacinema.com

Patents have been around for quite some time, even back when that mitchell camera was made. I can only imagine how much Mitchell benefited from their patents of their 4 pin register mechanism and also those early Canon DSLR's you liked where of course built upon huge stacks of patents and licenses. So no matter how you slice it, trying to give inventors an upper hand on copycats can not, to me at least, only be seen as bad thing.

 

[Peter Moretti]

You mean like Canon, Nikon, Silicon Imaging and Vision Research?

Nikon and Canon are stills cameras. Silicon Imaging uses CineformRAW, which I remember reading has some kind of an agreement with Red concerning the Redcode patent.

 

[Gavin Greenwalt]

Nikon and Canon are stills cameras.

Stills cameras are just low framerate film cameras. I can't patent "a car that drives faster than 200mph" (while everyone else drives 100mph).

Silicon Imaging uses CineformRAW, which I remember reading has some kind of an agreement with Red concerning the Redcode patent.

Wow! RED must have amazing lawyers if they pulled that off! Seeing as the RED project was publicly announced in in December 2005 as an uncompressed RAW camera and SI introduced their functional RAW recorder in April... 3 months later.

So unless RED invented it in December (decided not to use it) and Cineform coded night and day for 3 months I think you've got things a little backwards.

 

[Jeff Kilgroe]

If this patent BS happens with light field camera stuff, I am going to be mighty pissed off.

Um... I know of at least a dozen different patents applied for in regards to light field cameras and light field lensing arrays. Some by one company that is a client of mine. Their target products and customers have essentially nothing to do with cinema.

1. Create light field camera (just put some microlenses on an existing sensor... or even a printed grid can work)

2. Do all of the obvious stuff NOW before stupid patents go down.

And what is this obvious stuff? There are million different ways to skin this metaphorical cat and there are patents already being filed that pertain to a variety of things such as the incidence angles and surface radii of the micro lenses and the materials they are made from or coated with.

I remember sitting in Pietermaritzburg, South Africa in 2001, marvelling at Canon's DSLRs with my friends and of course after 5 minutes we were like "Wow, these 35mm sensors will be great for video too. You'd want to record the raw of course. And compress it a little to make it fit." I remember talking over this with a friend when he was chatting about buying an old Mitchell 35mm camera and the consensus was it was better to get on the digital thing because this was obviously coming down the line.

Heh. I beat you to it and so did Canon.. I was talking with a Canon R&D guy somewhere about 6 months after the first digital EOS was released. I asked him how long until this thing recorded video. At that time, they honestly thought they would do it within a year or two. And most of us were thinking about the upcoming possibilities of a digital camera, one that could shoot stills and video, surpass film, etc.. Way back when it was science fiction. SciFi stories, films, etc.. have portrayed digital cameras in several forms, or film-less camera might be a better description, since well, uh, forever. Science fiction and artistic depiction or "prior art" have never held up as a way to dismiss patented concepts.

I don't want future advances in cinema being hampered by someone telling me some painfully-obvious extension of existing technologies is patented. Seriously, let's build some stuff, create some prior art and stop this horror.

We'll eventually see what this whole RED vs. Sony business is all about. There is a lot more there than what's on the face of those patent documents, I can assure you of that. Patents don't have to disclose specific methodology, material compositions or anything that can be classified as "trade secrets". That includes manufacturing processes, molecular compounds, software algorithms, etc.. What RED's patents show is that they have been granted an exclusive patent(s) for a certain way of doing things. Now what they have to show is that Sony has infringed on their way of doing something by establishing to the courts the way it is being done, how long they have been doing it that way, and that Sony is now doing it that same way.

Like I said above, there are a million ways to do something. If RED feels Sony is infringing, perhaps they are. Perhaps they [Sony] need to go back to the drawing board and find their own way of doing it. The thing is, we don't know and we're speculating.

As for the whole hoopla over patents and obstruction and all that... Yes, I think our current patent system is in dire need of reform, like I said way back in this thread, we need a global effort to streamline the process and level the playing field. On the other hand, this is the way it is, this is how it exists now. From a business point of view, you either play ball or you go home.


For the record I hold three patents of my own. All three of them I feel are fairly obvious on the surface, one of them stupidly so. Of the three, I have licensed two of them, one is sold pending a product release that uses it, however I don't think that likely to happen now that the tech it is based on is falling out of favor, namely DLP projection...

 

[Bruce Allen]

And what is this obvious stuff? There are million different ways to skin this metaphorical cat and there are patents already being filed that pertain to a variety of things such as the incidence angles and surface radii of the micro lenses and the materials they are made from or coated with.

When I mean "obvious stuff", I mean stuff like "a video camera with a handle that shoots high-resolution compressed RAW"

RED's patent seems like it doesn't even want to nail it down to a specific compression format. They just want to patent the basic idea of compressed RAW video?!

I like RED but it would be a disaster for the industry if this patent were upheld. Compression has always gone hand-in-hand with video and being forced to store uncompressed would be a horrible restriction on anyone trying to do something new. Luckily thanks to people like Cineform doing prior art, I really don't think that one stands a chance.

There are other things such as doing noise reduction at various stages before and after compression. To me that is obvious that you'd want to do that. Pre and post-processing has always gone hand in hand with compression.

To me, certain methods of compression for both acquisition and distribution are obvious consequences of moving to a light-field or volumetric work. It would be similarly disastrous if those things were patented by any one company.

OK this has me galvanized. Once I stop doing 80 hour weeks I'm going to put all of my ideas that I consider obvious and necessary for future filmmaking stuff on a website at the very least.

Or hey... maybe I should just patent it and then use the licensing fees to fund my indie film projects

Bruce Allen
www.boacinema.com

 

[Peter Moretti]

Stills cameras are just low framerate film cameras. I can't patent "a car that drives faster than 200mph" (while everyone else drives 100mph).



Wow! RED must have amazing lawyers if they pulled that off! Seeing as the RED project was publicly announced in in December 2005 as an uncompressed RAW camera and SI introduced their functional RAW recorder in April... 3 months later.

So unless RED invented it in December (decided not to use it) and Cineform coded night and day for 3 months I think you've got things a little backwards.

At the time these patents were filed, there was certainly a vast difference between motion and stills cameras. Remember that the Red One predates even the Nikon D90.

As for who came up with compressing mosaiced data first, I don't know. But I do remember reading that Red and Cineform had an agreement in regards to Red Code.

 

[Bruce Allen]

At the time these patents were filed, there was certainly a vast difference between motion and stills cameras. Remember that the Red One predates even the Nikon D90.

To me as a guy who did a lot of stop motion animation, there was never much of a difference. You could use both for both purposes (stills cameras to do video and video to do stills). The primary difference is just FPS and interface (dictated by FPS).

There has always been a link between 35mm motion and 35mm still cameras.

The 35mm still camera was always just a slow movie camera - seriously, look up its invention!

They looked at 35mm motion picture film cameras and said "hey, what if we use the same film and technology but just allow people to take select frames..."

As for who came up with compressing mosaiced data first, I don't know. But I do remember reading that Red and Cineform had an agreement in regards to Red Code.

You keep stating this but I certainly don't remember it.

What I do remember was that back in 2006, they were already shooting Spoon in compressed Cineform RAW in South Africa.

Bruce Allen
www.boacinema.com

 

[Jeff Kilgroe]

Stills cameras are just low framerate film cameras. I can't patent "a car that drives faster than 200mph" (while everyone else drives 100mph).

Of course. However, you can patent engine modification designs or techniques or drivetrain alterations that could make derivative versions of that 100mph car travel at 200mph. You also don't have to disclose, within the patent documentation itself, enabler methods used to achieve those modifications -- such as gear ratios, surface coatings or lubricating compounds.

I guess what I'm saying is that RED hasn't shown their hand yet and simply analyzing off the surface context of what is on those patent papers only makes it all seem silly. Whatever it is seems to be pertaining to the F5 and/ or F55 cameras. Look at the timing in relation to official release of the new Sony cameras and in relation to the new RAW software tools from Sony. The F65, F35, etc.. as they have existed for some time now, don't seem to be the concern here. On the other hand, it takes time to put something like this together, so those other cameras very well could be and the timing could just be coincidental. After all, it takes time to file for and be granted patents, some of these patents were granted pretty recently.

We don't even know if RED's legal argument here has anything to do with compression or REDCODE.. And when it comes to compression, the concept of using structured algorithms to "compress" data is by no means new and you can't patent it. On the other hand, you can patent the specific method or algorithm used to do the compression.

I'm really eager to see what happens here. And yes, I'm rooting for RED in a way, but to be honest I really don't know who is "right" or "wrong" here and we can only speculate as to how deep this rabbit hole goes.

 

[Gavin Greenwalt]

At the time these patents were filed, there was certainly a vast difference between motion and stills cameras. Remember that the Red One predates even the Nikon D90.

As for who came up with compressing mosaiced data first, I don't know. But I do remember reading that Red and Cineform had an agreement in regards to Red Code.

The 1D was delivering 8.5fps in 2004. That's only 33% of 24p. You can't patent "I'm going to make a camera that shoots 3x faster!"

Also it doesn't matter who came up with the idea of compressing bayer pattern data first... you can't patent something as obvious as "hey I've got an idea, how about we don't shoot uncompressed?" It's like me deciding to patent zipping an excel spreadsheet. "hey nobody ever thought of zipping an excel sheet! Sure people were zipping word documents, but right clicking on an excel sheet hasn't been done!"

http://www.dcimovies.com/archives/spec_v1/DCI_Digital_Cinema_System_Spec_v1.pdf
The v1 DCI spec is from 2005. It specifies JPEG2000 encoded frames as its specs.

Motion picture camera with higher resolution than 1080p: Prior art (DALSA).
More than 23 frames per second RAW: Prior art (DALSA).
Compressing Movies: Prior art (SI-2K/Cineform, The entire history of digital video)
Compressing Bayer Pattern: Prior art (Canon, Nikon, Olympus etc...)
Using a LUT to make data more compressable: Prior Art (Cineon, YUV, sRGB jpegs etc.)

So I can only assume that the patents that they're trying to defend is the exact recipe that makes the data more compressible. In which case even if that exact methodology of prepping the footage for compression is found to be patentable (and very well may be novel and unique) you could get very good results without it by just JPEG2k encoding a straight up Cineon log LUT.

 

[Björn Benckert]

Stills cameras are just low framerate film cameras. I can't patent "a car that drives faster than 200mph" (while everyone else drives 100mph).

You Do not think so? Well, Douglas Trumbull has patent for showing 65mm film in 48p instead of regular 24p, the patent is called "showscan".

Look at his patents here: http://douglastrumbull.com/patents

It's quite crazy what you can file for patents. Like his amusement park VR ride patent I kind or like the most, it's like:

- Make a box.
- put people in the box.
- show the people film inside the box.
- Shake the box.
- file for a patent.
- Make sick amount of money.

 

[Dustin Shepard]

You Do not think so? Well, Douglas Trumbull has patent for showing 65mm film in 48p instead of regular 24p, the patent is called "showscan".

Look at his patents here: http://douglastrumbull.com/patents

It's quite crazy what you can file for patents. Like his amusement park VR ride patent I kind or like the most, it's like:

- Make a box.
- put people in the box.
- show the people film inside the box.
- Shake the box.
- file for a patent.
- Make sick amount of money.

Big B!(Bjorne) I'm still in the clear to buy an epic right? Lol I was feeling a little sketch bout buying an epic cause of my imagination went freaking haywire imagining Sony shutting down red and me not being able to do jack shit with my epic hahah

 

[Peter Moretti]

The 1D was delivering 8.5fps in 2004. That's only 33% of 24p. You can't patent "I'm going to make a camera that shoots 3x faster!"

Also it doesn't matter who came up with the idea of compressing bayer pattern data first... you can't patent something as obvious as "hey I've got an idea, how about we don't shoot uncompressed?" It's like me deciding to patent zipping an excel spreadsheet. "hey nobody ever thought of zipping an excel sheet! Sure people were zipping word documents, but right clicking on an excel sheet hasn't been done!"

http://www.dcimovies.com/archives/spec_v1/DCI_Digital_Cinema_System_Spec_v1.pdf
The v1 DCI spec is from 2005. It specifies JPEG2000 encoded frames as its specs.

Motion picture camera with higher resolution than 1080p: Prior art (DALSA).
More than 23 frames per second RAW: Prior art (DALSA).
Compressing Movies: Prior art (SI-2K/Cineform, The entire history of digital video)
Compressing Bayer Pattern: Prior art (Canon, Nikon, Olympus etc...)
Using a LUT to make data more compressable: Prior Art (Cineon, YUV, sRGB jpegs etc.)

So I can only assume that the patents that they're trying to defend is the exact recipe that makes the data more compressible. In which case even if that exact methodology of prepping the footage for compression is found to be patentable (and very well may be novel and unique) you could get very good results without it by just JPEG2k encoding a straight up Cineon log LUT.

Gavin,

That the 1DC could shoot 8+ fps does not in anyway make it a motion picutre camera. I have a 1999 Saturn with a tow hitch on the back, does that make it a tow truck? I mean, it's got an engine and can pull stuff... at least 33% of the way there!

The DCI spec is irrelevant, as it calls for J2000 compression of non-mosaiced images. Do you really think that DCI players do on-the-fly demosaicing and have debayering settings?

An important revolutionary aspect of Red was compressed RAW.

I suspect Sony believes it can get around the patent becuase, IIUC, they are not using wavelet compression.

 

[Bruce Allen]

We don't even know if RED's legal argument here has anything to do with compression or REDCODE.. And when it comes to compression, the concept of using structured algorithms to "compress" data is by no means new and you can't patent it. On the other hand, you can patent the specific method or algorithm used to do the compression.

OK, Jeff, help me understand. I'm trying to quit this thread but I just can't.

Do the actual specific methods or algorithms used to do the compression need to be contained or described in the patents themselves?

Because here are the patents:

http://news.priorsmart.com/redcom-v-sony-corporation-of-america-l7D8/#pat-8174560
http://news.priorsmart.com/redcom-v-sony-corporation-of-america-l7D8/#pat-8358357

I'm just not seeing anything in there that is specific groundbreaking algorithm stuff.

It would have to actually be described in the patent, wouldn't it?

Most of it really seems to be about "hey, we patent having a visually-lossless RAW recording video camera with a Bayer or other sensor, using JPEG2000 or other compression, that does more than 23fps and 2K or greater" - which seems insanely general for a patent and would be really restrictive on the industry.

After reading the patent... what kind of specific stuff could Sony actually be infringing here? Am I missing something?

Bruce Allen
www.boacinema.com

 

[Peter Moretti]

Bruce, I believe the crux of the matter is that Red is compressing the RAW mosaiced data.