Thursday, May 07, 2009

Copyright Office DMCA Hearings: Filmmakers

Today’s notes are not as reliable a summary as yesterday’s, because of my active participation. (And of course the usual caveats apply to the extent that you want to use my reports as summaries; I strive for accuracy, but I am not a transcriptionist.) I believe the Copyright Office intends eventually to release a transcript on its website, where all the proposed exemptions and comments are also available. Fred von Lohmann of the EFF also recorded portions of the Palo Alto hearings along with yesterday’s hearings, including video of the MPAA camcording demo. You can also watch the MPAA demo on Vimeo.

Begin with 11B: motion pictures and AV works in DVD that are not generally available without CSS when a documentary filmmaker is enrolled in a film production course at a postsecondary institution or is a member of a documentary filmmaker organization, when making a specific documentary for which substantial production has occurred, and when the material is in the public domain or its use would be fair use.

Gordon Quinn, Kartemquin Films: Makes nationally broadcast documentaries, including Hoop Dreams and The New Americans (immigration); Terra Incognita (stem cell research) just won a Peabody. Important participants in culture and democracy: contextualize real stories. Most documentary filmmakers belong to at least one documentary organization.

The prohibition is burdensome, intimidating, and impeding fair use and use of public domain materials. The exemption is for filmmakers who are relatively educated about what fair use is and isn’t. This is the bare minimum: only for CSS. This is just the fair use we’ve relied on for decades. The alternatives don’t provide the proper quality, as Jim Morrissette (technical director) will testify. Most documentarians don’t have technical directors with engineering backgrounds. Many things he shows/can do are unavailble to even very good documentarians. It causes Kartemquin a great deal of grief, but others suffer even more.

Fair use is critical to documentarians, which is undisputed. Filmmaking has a long history, and is more predictable than other types of fair use.

DVD is the default distribution format for now. We anticipate it being so for the next 2-3 years. How we use fair use: Refrigerator Mothers (the story of the mothers of autistic children from the 50s & 60s, told by experts that they caused their children’s autism). We show the hold of Freudian psychology on 50s and 60s pop culture, putting a lot of weight on mothers for failing in lots of ways—show excerpts from feature films, off DVDs, magazine articles, etc. It’s critical to the film, but very short, because we don’t take more than we need.

Alternatives are not adequate to achieve the quality necessary for broadcast/theatrical distribution. Also they’re expensive and difficult.

Opponents have suggested solutions that further burden fair use. Licensing as an alternative undermines the purpose of fair use. Center for Social Media study on Untold Stories showed that licensing didn’t work: high and increasing fees; broadcasters demand broader and broader distribution rights from documentarians, but copyright owners are unwilling to license that broadly. Still an unresolved issue.

But fair use is not licensing. The New Americans includes an interview with a Nigerian immigrant who says Coming to America is a favorite movie in Africa, because it shows how little Americans know about Africa. They tried to license a clip, even though it was clearly a fair use, and failed no matter how persistent they were. The film was weaker because of it. Standard studio licensing provisions prohibit criticism of the studio, the industry, the executives, the character, or the film: he showed a standard licensing provision. Along with restrictions on criticism, the license is void if the MPAA rating is more restrictive than “R.” The heart of fair use is that we have to be able to criticize, comment on, and put things in context.

We do license things all the time: we don’t say everything is fair use. We expect to license when that’s appropriate. But we expect not to have to when fair use legitimately applies: ownership can’t prohibit free expression.

Good faith effort to obtain unencrypted copy from copyright holders: has the same problems as licensing. These are the same people who’ve given us so much trouble over the years in licensing. We need the right to participate in the culture of the community, critique things in society; we shouldn’t have to ask permission. Can lead us on wild goose chases—we sometimes go to the artist in cases of fair use, but we shouldn’t be required to do so. This can also create requirements to pay lab fees. You may have the DVD and yet when you go to the copyright owner, all of a sudden the owner tells you it needs to be redigitized at a very high price.

In our work, there are a lot of clips from a lot of things—a montage. Creating a new thing to show the culture of a period, or how people have been treated in certain kinds of media. If each one has to go through a lab process, that adds up.

Suggestion to add “solely” to the class—“solely” for use in a specific documentary film. We don’t object to that change for clarity. We don’t see harm to copyright holders, which is important because we are also copyright holders. We distribute and sell DVDs, some of which have CSS. Our proposed exemption is narrow and could not be used for infringement by consumers watching something they didn’t buy. It would only apply to the established documentary community; the likelihood of misuse by a documentarian is low; the community is professional, thoughtful, and responsible about fair use. (Comment: You must be this tall to make a fair use!) Because of best practices, we can now get E&O insurance for our films when we make fair uses, which used to be impossible.

Copyright owners retain their rights and remedies.

Today, almost all public domain material is bundled with newer, copyrighted material on DVDs, such as extra features or a new soundtrack. Thus, we need to be able to access the public domain material on those DVDs.

We’re not asking for anything broad, vague, or unfocused. We just want fair use itself. As much as we need and no more. Not trying to open up the barn door to random pirates, and we’re a very defined class of people who have spent a lot of time and effort getting educated on fair use.

Jim Morrissette, Kartemquin Films: Technical director; worked there for 30 years. Almost all productions come to him to be fixed. How he deals with analog alternatives to CSS, and resulting problems that affect the ability to get the shows on public TV. It’s been suggested that we could use VHS. We have used VHS, but it’s universally accepted that image quality even on a brand new tape is vastly inferior to begin with. The problem doesn’t stop there. Availability: VHS is no longer being made. Older tapes: for our stem cell movie, we wanted a clip from Sleeper. We first couldn’t find a VHS copy in the Chicago area; finally interns found a library copy. To get it into the program, a zillion steps were necessary to get it broadcast quality.

Play the tape; run through a time corrector to remove Macrovision analog copy control; turned digital. Because of the time corrector, video levels were off, leading to image degradation; must adjust video levels. Recorded output to digital VCR, leading to analog to digital image degradation. Required adjustment of black level. And required cropping head switch, black lines at the bottom of every VHS tape, which consumers don’t see but broadcasters do. Frame size must be adjusted, and then the file must be rendered, causing more degradation.

There are hidden signals: switch, blanking, color—broadcasters (and the FCC!) insist that they must all be within a very narrow spec. If any clips are not within spec, they must be fixed, and fixing them is not easy.

Camcorder alternative: Used to record live TV before videotape was invented, called Kinescope then. The necessary steps raise even more issues in the transfer, setup, and postproduction. Loss of image resolution, warping and other perspective problems, visible scan lines. Setup requires precise camera alignment, audio line-in, color balance, dark room/hooding the camera. If you aim a little too wide, you’ll get the set; if you aim too narrow, you crop and degrade the picture.

More serious technical issues in yesterday’s demo. All HD cameras are widescreen, 16:9 unlike the 4:3 standard def. A SD camcorder could only shoot a square screen, meaning the result is black bars on top and bottom, which degrades the image even further. The source was dark, fast-moving, quick-cutting: an excerpt from Harry Potter. It was impossible to tell whether the image was jerky, flickering, whether the colors were true. If you’d chosen a presidential speech where you could see a human face and judge color, it would be easier to evaluate. We also didn’t get camcorder compared to DVD; instead they compared to a downloaded, compressed, QuickTime file, which of course didn’t look better. That wouldn’t pass public TV standards.

Once the camcorder recording exists, you’re not done. If it’s HD, and you’re making SD, you have to downconvert.

Another alternative: use a black box to record a computer screen. Scan conversion boxes, often used for people making training films. Computers don’t play video the same way TV plays video; they play progressive. Broadcasters want interlaced, which is what comes out of the DVD. The change from interlaced to progressive back to interlaced cuts image resolution in half. There’s also a sync problem with the audio—the delay from conversion can be ½ a second. You have to resync. Scan converters cost $1800 and up.

Third method: use software to record the screen. He’s tested some, including SnagIt, JingPro, and others. They may be acceptable for low-resolution web clips, but unacceptable for our use. They can’t capture the full frame rate of DVD: 30 fps. Even on a quad-core Intel Mac, they’re lucky to get 15 fps. Result: jerky motion. Also very difficult to define precise part of the picture to record. If you get it wrong, your file has to be blown up or cropped or otherwise manipulated to get proper frame size. There are only 3 frame sizes allowed for broadcast; you can’t be wrong, and resizing degrades image quality.

What do we do now: Analog hole. It’s similar to the VHS process. DVDs have Macrovision, so analog output is not directly recordable but must be run through expensive time base corrector to strip that off. Record on digital VCR. Problem: you’re starting with an analog signal out of the DVD player. If you compare analog to digital, it’s night and day: composite signal is blurry, colors are washed out. There’s also an issue with black levels, which affects not only the look but also whether public TV will accept it. Analog black level is higher than digital. You thus have to modify the black level coming out of the DVD player, carefully so you don’t degrade image quality more than necessary. Digital tape must be transferred to computer file. Rendering again creates image degradation.

Even with all his expertise, we still have to send our programs out to a high-tech gatekeeper, one of the approved public TV quality control intermediaries. These issues are grounds for rejection from broadcast. We have to fix them, which is an additional burden in time and money. It’s a key point that it’s not just that we need “good enough.” If we were showing a clip in a classroom, some of these other methods might be good enough, even though they’re a heck of a lot of trouble in time and effort. But if we want national TV broadcast or cable dissemination (many cable channels are equally picky) we need better. Analog conversions stand out as visually and aesthetically awful, and often have to be replaced.

A lot of filmmakers can’t afford this and just give up. If we get this exemption, we can extract the right frame rate, aspect ratio, black level, etc. and just use them. When we make fair use, we only need short clips. We never convert more than we need. It’s not like we have a decrypted copy of a movie sitting on a hard drive for an intern to take home. It’s a lot of work.

Bruce Turnbull, DVD-CCA: Keep yesterday’s comments in mind, especially with regard to the standard for a class of works. Primary concern: viability of DVD-CCA’s technology. CSS enabled the development of DVD market and is critical to its commercial viability. DMCA was enacted with CSS as an example of technical protections available at the time.

These uses are commercial in nature, not educational. The educational exemption was deemed necessary for pedagogical purposes, but here we don’t have that same quality.

Almost all DVDs today do not enable Macrovision, so in the future one of those steps won’t be necessary for most movies.

We understand the limitations of MPAA’s presentation on camcording. The point being made was that the camcorder option has become a viable one in light of the development of technology both as to the nature of the screens being filmed and the type of cameras being used. Professional filmmakers shouldn’t have great difficulty in terms of sizing the image. We compared the result to what the film studies professors said they wanted to use in their class. (Actually that’s not true: Decherney doesn’t use the compressed QuickTime version to teach—he sent it in for display on the Copyright Office website.)

Also documentarians don’t have the same time constraints as educators. The film studies profs have 50-minute class times and make revisions to lectures the night before. Filmmaking is lengthy, and the filmmakers can use the alternatives. (Of course filmmakers have other constraints, as detailed above.) In each rulemaking, DVD-CCA has said it welcomes people to come talk to us about specific needs; no one has ever done so. The market should work this out by negotiation with DVD-CCA. (What does DVD-CCA propose to do? Can it authorize circumvention of particular works? Will it hand over clear copies?)

Steve Metalitz, joint commenters: The proponents have made a good effort to bring the proposal within the existing framework, in contrast to 11A. Still problems: Noninfringing use; alternatives; narrowing.

The Office needs to decide whether the prohibition substantially impedes a use that is in fact noninfringing—a consistent standard. Are clips in documentaries clearly noninfringing? Sometimes yes, sometimes no: Elvis Presley case—documentary was found not to be fair use. Not suggesting that Kartemquin is an infringer, but the office isn’t in a position to recommend an exemption unless it’s in a position to determine that the use is noninfringing. This is inherently more difficult when the basis for the claim is fair use, inherently a case by case determination. The proposal uses proxies: membership in organization or enrollment in film course. That is supposed to up the chances that the result is fair. But why does that make you more likely to make a fair use? In many of Decherney’s film programs, students apparently never touch a camera; we don’t know what they’re taught about fair use. (He’s conflating film production with film studies. It’s like conflating potterymaking with art history.)

Alternatives: One is licensing or nonobjection; the second is scan conversion/camcording; the third is use of other digital formats. It’s important to clear up that fair use is not the same as noninfringing use. Fair use doesn’t require permission, but a permitted use is noninfringing. If people who wish to circumvent can get permission, the test of the statute is satisfied because they can make noninfringing use. You can make fair use without obtaining clips. Corley again: fair use isn’t a guarantee of access to preferred techniques/format. You don’t need permission to make fair use, but you also don’t need to circumvent.

Time Warner: 368 instances in which they permitted/didn’t object (to what? They didn’t give clear copies). If lab fees are the cost of a higher quality clip, that’s not an impediment flowing from the prohibition on circumvention. Lab fees are sometimes charged by studios.

Scan conversion: We’ll look at Morrissette’s information and have tech experts react to that.

Final point on alternatives: the submissions overlook an increasing number of digital formats that aren’t protected by access controls used to distribute at least some of the material. Hulu and network sites: much of the video provided there is in the clear. Increasingly other digital formats will be accessible without circumvention.

Narrowing: We don’t think it’s appropriate to refer to public domain material, because no exemption is needed. We don’t think it’s right to refer to fair use. If that becomes an element of these exemptions, then we’re down the road to saying a noninfringing result is ok. That’s not what Congress authorized. Any exemption should have a necessity requirement: documentation of a good-faith effort to obtain the material without circumvention. Should clearly say that the exemption is limited to DVDs protected by CSS. And should be clear who can exercise the exemption. Defining exemptions by use and user is distressing.

David Carson: Who can exercise? What do you mean?

Metalitz: Exemption is drafted in passive voice. Is the documentarian the person who can circumvent? Or can his agent?

Carson: Reactions to proposed conditions—public domain, §1201 doesn’t cover PD material?

Quinn: You have a DVD. It has PD and non-PD material, and the whole thing is encrypted. That’s more and more common. Also: the idea that there are new digital sources—these things on the internet are not broadcast standard. They are incredibly problematic.

We educate on the Presley case all the time, expanding understanding of fair use. We are concerned with linedrawing, and the best practices for documentary fair use are about that.

Rob Kasunic: Give us some idea of what kind of PD material you need to get from a protected DVD. News reels? Why not go back to source?

Quinn: for certain old feature films, might not be available in any other format. We often will go back to source if we can. Historical documentaries have a big issue here. Other filmmakers use stuff from early days of film, old news footage: some is available and some isn’t. It’s often bundled.

Metalitz: Bundling occurs, but that didn’t justify an exemption in 2006. Circumventing access control on public domain material is not a violation of the statute.

Quinn: But breaking the encryption on the bundle is what concerns us. Also, the alternatives look like crap. When you’re commenting on culture, particularly on feature films: feature films are very sensual, very visual. Sometimes it’s important to preserve that quality so people understand why the films created the emotions or attitudes they did. It doesn’t have to be what you see in the theater, but it has to approximate that.

Turnbull: 1201 doesn’t cover public domain works.

Carson: OK, so what about the bundle to which CSS has been applied?

Turnbull: Yes, the DMCA applies to the bundle. But if you had a straight PD work, no prohibition.

Carson: So, there’s a consensus on your side that circumventing a bundled work to access the PD material does not violate 1201?

Metalitz: It’s not within the scope of this proceeding.

Carson: Answer the question?

Turnbull: Turnbull was answering the question as applied to a non-bundle. Your question is more complicated. And of course the “tools” prohibition applies anyway.

Carson: We’re not dealing with tools; that’s not productive. We want evidence of the magnitude of the problem: how often are PD works bundled and you can’t access them otherwise? In prior proceedings, we had a paltry showing. Maybe the situation has changed.

Another question: Metalitz argued that any exemption shouldn’t refer to fair use. But putting it in the proposal on the face of it narrows the exemption, because there are other noninfringing uses. Are you saying don’t mention fair use in an exemption, so documentary filmmakers can simply circumvent?

Metalitz: If in fact the use is infringing, it may still be a violation. Our concern is that by defining an exemption in terms of a class used by certain people in a certain way if it’s fair use, we’re concerned about the short (or long) step between 11B and 11A. 11B has elements attempting to serve as a proxy to ensure the use is noninfringing, which is the standard you’re supposed to apply. One red line you can’t cross is to say “circumvention for the purpose of making a fair use.”

Kasunic: Yet this is an attempt to narrow a noninfringing use.

Metalitz: If a documentary filmmaker who met these qualifications was going to make a movie, but ends up being an archival copy in a library, then would it not apply? It’s true, the only noninfringing use they’ve talked about are fair. But their burden is to show that the uses enabled by the exemption will be fair/noninfringing.

Carson: What about requiring a good faith effort to obtain material without circumvention?

Quinn: That’s one of those vague things. Who would supervise a good faith requirement? We make a lot of good faith efforts to reach people for licensing, and that becomes very difficult. We went to a lot of effort to find a copy of Sleeper. We’re researchers, but documenting it for a litigation purpose, making us vulnerable to a lawsuit, would make it very intimidating. In many situations we go directly to an artist and tell them even when we think it’s fair use. What kind of vulnerability would that requirement open us up to? We live with the fear of going back to the situation 5 years ago, which is that fair use was off the table because of fear of liability. Contracts required everything to have been licensed, the only exceptions being under the table/“don’t ask don’t tell.” That happened because the people telling us now to make a good faith effort were so aggressive that we thought we’d lost the right to make fair use as they threatened broadcasters. We’ve pushed back as a field and the broadcasters have some backbone. The alternative is damaging to our democracy.

Carson: what about asking permission? If they say no or impose unreasonable restrictions, you can circumvent.

Quinn: Fair use isn’t about permission.

Carson: 1201 has no fair use provision.

Quinn: If you have asked permission and have been denied, you haven’t weakened your fair use claim at all. The thing he’d resist would be a permission requirement. As part of professional conduct, it’s not bad to let people know what you’re doing. (Permission to do what? To circumvent? Do you have to ask DVD-CCA and the copyright owner, or will just the copyright owner’s permission suffice? Or is this permission to use a crappy copy a substitute for circumvention to get a quality copy, which still has the same problems?) Fair use is not as fuzzy as it used to be, at least in our field.

Carson: Can you circumvent in such a way as to get only a short clip?

Morrissette: We do it every day with non copy-protected materials. There are several software programs that let you mark an in point and an out point; it makes a new file. There are programs that don’t require you to copy the whole disk. We never have a whole copy sitting around. We only take what we need.

Turnbull: His understanding of how these programs work is that they allow you to play back the movie, decrypt it, and store the entire thing, which you may then later mark to only use 30 seconds. You can’t on the fly decide only to decrypt 30 seconds at the 90-minute mark. (This is wrong, by the way.)

Quinn: The point Morrissette is making is that these programs could copy the whole film, but the original DVD is still encrypted. In our field, we don’t take the whole thing, we take the piece we need; otherwise it creates storage burdens and it’s just not necessary to take more than what we’re going to work with.

Carson: Assuming it’s possible to decrypt only what you need, would there be any problem with including a condition that you decrypt only so much as you need?

Quinn: No problem.

Turnbull: That’s the kind of restriction we’d like to see if you go down that road.

Kasunic: Are we talking about software that documentary filmmakers get? Or would it be possible for the DVD-CCA to provide a key to a limited portion of a DVD?

Turnbull: No, that’s not how it works.

Morrissette: It has to do with the user. If the organization only pulls out the clips we need, we will never have a copy of the whole thing.

Kasunic: What could you do for documentarians if they’d come and asked you?

Turnbull: He can’t say. But if someone had asked to license the technology such that it does X (what is X?), we would consider it, but he can’t say they’d agree or what conditions they’d put on it. We’d prefer that dialogue rather than a widely available exemption.

Kasunic: DVD-CCA couldn’t help with the underlying content, though, right?

Turnbull: Right. You could definitely get tripped up.

Quinn: That would be problematic. Ownership should not determine what you can use for fair use. If your concern is that we’re taking unencrypted pieces, then work with us and we’ll make sure we erase the pieces when we’re done. They always say “there must be a marketplace solution” but this is a fundamental free speech right. It gets complicated to go ask. They start asking you for publicity rights, etc. Fair use doesn’t require any of that. We have an interest in not creating a situation in which the exemption is misused to steal copyrighted material. And if we’re creating a problem, we’ll work with copyright owners.

Turnbull: §1201 has been upheld against First Amendment challenge.

Carson: Another condition—only available when the access control is CSS. Is that acceptable?

Morrissette: Yes.

No comments: