How are AMV's legal?

[shumira_chan]

Well, yes and no. You also need the "assets" (the audio, video, any still frames, etc.). You need the "index" files (generated as a preprocessing step) if you're editing MPEG-2 (like from DVDs) because Cinelerra can't do frame-accurate seeking by itself. In my experience, it can't edit from purely ripped VOBs anyway, so I extract just the video stream with a demuxer, index that, and edit from there. Probably the thing to do would be include a script with the EDL to do the ripping/preprocessing; in theory someone could just be told "rip it in this way with these tools" but that would be error-prone if we're trying to get a bit-identical result. Other issues: different versions of Cinelerra are not as compatible as they should be, and EDLs normally include full paths (so we'd probably have to standardize a directory structure for storing assets).

The program I'm using (LiVES) is very interesting because
it uses a Perl program as a backend, and can perform all
the steps needed, from ripping audio and video to the final
encoding (however, it is still very new, and very much under
development). The point is, if one saved all the calls to
the backend in a plain text file all one would need to do is to
simply execute this file in order to generate the video
(of course, the Perl program calls a bunch of other programs,
so in a sense it's like having a modularized Cinelerra i.e.
instead of having one big video editor you use Perl as a
glue language which calls a bunch of smaller programs).
The problem with the glue approach is that you need a
lot of smaller programs which would be very hard for
Windows users to put together (heh, heh :) The good side
to this is that it uses the unix philosophy of having small
pieces do one thing very well (it's my understanding that
Cinelerra is not very stable -- I haven't tried it myself --
but LiVES is a rock).

I'm discussing the above with the developer (who started
LiVES with a very different focus -- video DJ'ing -- but I'm
pushing hard to make it an AMV-making tool). One problem
is that I'm actually a Python programmer, so I'd really have
to learn Perl better.

But speaking of Python, my vision of a video-editing
language would be something like PIL -- the Python
Imaging Library -- call it PVL (Python Video Library):

http://www.pythonware.com/library/pil/h ... uction.htm

Python is cross-platform, free, and has a very clean
syntax. A PIL-like PVL would be something I'd really
like to see. However, for the time being I'll press with
the LiVES' Perl backend (called "smogrify", BTW).

Well, we have some highly skilled technical people... the war isn't over yet.

In a sense it's hopeless because, right or wrong, few
people will confront a lawsuit to defend their hobby against
almost limitless resources. But then again, strictly speaking
_everyone_ who has a computer is,
violating some law (it's still laughable to read some
webpages' TOS which, in fact, don't allow you to look at
the page at all). Sadly, it boils down to keeping a low
profile. "Script AMVs" do that, but could also have some
very attractive advantages (and admittedly, some very
serious disadvantages). But the idea is there, and I think
it's worth exploring...

Cheers!

[Kalium]

What the editing index idea issues come down to is this: it's very difficult to impossible to ensure that:
1) The viewer has a bit-identical copy of the source.
2) The viewer has exactly the same system configuration. (A different video card, for example, could throw some things off. Or more/less RAM.)
3) The viewer has the same software configuration.

Basically, the current methods are really optimal. Otherwise, there's no good way to ensure that what I see is what the editer wants me to see. If the editer can't perfectly control what I, the viewer, sees, then there goes the whole basis of video editing.

From a legal standpoint, nobody is really making or losing money from these directly (although the companies may be making money from increased sales). They really don't care about the fan work, as it only helps them. The result? While technically illegal, AMVs fall into the 'nobody cares' legal area.

[trythil]

Does an EDL contain all the information which Cinelerra
needs in order to recreate any video which was made
with it?

No. My projects integrate data from many sources, such as Blender render output, CinePaint-tweaked imagery, and exports from AVISynth + VirtualDubMod and Adobe Photoshop.

I could give you the EDL to my latest video, but it'd tell you essentially nothing, even if you had the Escaflowne Movie VOBs and Escaflowne TV series VOBs indexed and in the right locations, as so much of it is composed of tweaked frames.

It would be possible to include all the source files and EDLs that were used to generate dependencies, but by then you'd end up with an archive somewhere in the hundreds of megabytes. You'd need approximately 65 GB of space to render temporary files, and another 3 GB to render the final product.

In short -- I don't think it's very efficient :P

[shumira_chan]

Does an EDL contain all the information which Cinelerra
needs in order to recreate any video which was made
with it?

No. My projects integrate data from many sources, such as Blender render output, CinePaint-tweaked imagery, and exports from AVISynth + VirtualDubMod and Adobe Photoshop.

Ok, but you did't read the question quite right. If you
use Cinelerra, and _nothing_ else, can you remake the
AMV using the EDL (and the source material, of course)?
Maybe not, but because Cinelerra may not be able to
generate an AMV from source material, but that's a
limitation of Cinelerra, not the idea of what an EDL
could accomplish.

It would be possible to include all the source files and EDLs that were used to generate dependencies, but by then you'd end up with an archive somewhere in the hundreds of megabytes. You'd need approximately 65 GB of space to render temporary files, and another 3 GB to render the final product.

In short -- I don't think it's very efficient :P

I never disputed that. My points are: efficient distribution,
easy to modify, and, most importantly, very possibly
legal (undoubtly more so than AMVs, which cannot
be defended in any way). Distributing a 100KB file
is much more efficient than distributing a 100MB AMV.
I'm not sure about the 100's of MB you mention, remember
I'm saying you have to own all the video/music sources
to avoid (traditional) copyright infringement.

Resource issues are of a
different nature altogether, 100GB of disk space is not
an astounding amount nowadays, in two years' time you
may not even be able to find hard drives smaller than that.
In any case, what I'm suggesting would be -- currently --
for experts such as yourself, and you already have those
resources ;)

[trythil]

Ok, but you did't read the question quite right. If you
use Cinelerra, and _nothing_ else, can you remake the
AMV using the EDL (and the source material, of course)?
Maybe not, but because Cinelerra may not be able to
generate an AMV from source material, but that's a
limitation of Cinelerra, not the idea of what an EDL
could accomplish.

Oh. You're right, I didn't see that. In that respect, it is possible to accomplish, and requires little additional code. One could write a small shell or Perl script to rip DVDs, generate MPEG indexes, and call Cinelerra via its command line interface to start rendering.

I still wouldn't have much use for this, as I rarely limit myself to exclusive use of one program, let alone one platform. That, is my problem, and does not reflect on any facet of the idea.

I never disputed that. My points are: efficient distribution,
easy to modify, and, most importantly, very possibly
legal (undoubtly more so than AMVs, which cannot
be defended in any way). Distributing a 100KB file
is much more efficient than distributing a 100MB AMV.
I'm not sure about the 100's of MB you mention, remember
I'm saying you have to own all the video/music sources
to avoid (traditional) copyright infringement.

Would it be legal to distribute these? You could make the argument that they are still essentially unauthorized derivatives of a copyrighted work, as they were created from direct manipulation of a copyrighted work.

--

The hundreds of megabytes I mention stemmed from the individual frame-by-frame changes that I referred to earlier, which I now realize were never in the scope of the problem.

If you're still interested, here's a file listing that may help to illustrate what I was talking about.

[shumira_chan]

OK, my last post on this matter before I'm labeled a troll.
Please read carefully my previous posts if you want to
respond, I'm probably not making my case well (but that's
nothing new

I'm not arguing against the current status quo, but I think
AMV makers should expand their horizons a bit beyond
using technology just for eye candy, and perhaps focus
a bit on some other aspects of video making, distribution,
and the possible legal consequences (which, I reiterate,
I hope will never surface, but keeping one's head buried
in the sand is not going to help much if it does). Doing
this, I'm afraid, will at first, without a doubt, impact on the
quality of AMVs, but then again the first cars were slower
than a horse.

(Where I'm coming from: I don't care about stars, ratings,
competitions, eye candy is not very important to me
and I suck at making videos, but I do enjoy making them,
sharing them, understanding the technology behind them,
and I'd like to continue doing so... if this makes me an
AMV community outcast so be it, but as I said, I'm not
out to damage the current status quo).

What the editing index idea issues come down to is this: it's very difficult to impossible to ensure that:
1) The viewer has a bit-identical copy of the source.

Agreed, although I'd be curious to know just how much
variation there is, and how relevant it is. Is there not,
for example, variation amongst the quality of current
decoders? I've made AMVs which do look different
when I play them back with MPlayer, or Xine, or WMP,
or QuickTime... if the codec happens to be available
for them all.

Yes, there are more things that could go wrong with
source AMVs, I won't deny that. However, you could
fix that (not a trivial thing, you'd need to be a programmer,
but life's like that sometimes

2) The viewer has exactly the same system configuration. (A different video card, for example, could throw some things off. Or more/less RAM.)
3) The viewer has the same software configuration.

Think of what I'm suggesting a bit like the difference between
downloading a binary vs. a source program, and some
of these points apply to that case also... yet source
distribution is seen by many as invaluable. Admittedly,
AMVs are more complicated, but that's also largely in
part because we don't have the right tools to make them
the way I suggest (nor are these tools trivial to create).

I must say, however, that when thinking about software
problems someone could make a bootable Linux CD
with all the necessary software (think of a video-oriented
LSB-type compliance -- If your AMV satisfies it it is
guaranteed to "compile"). In fact, there's something like
it already:

http://www.dynebolic.org/

You don't even need to install anything (although you
would need a 100GB partition somewhere for temp
files).

Basically, the current methods are really optimal. Otherwise, there's no good way to ensure that what I see is what the editer wants me to see. If the editer can't perfectly control what I, the viewer, sees, then there goes the whole basis of video editing.

I hope that the irony is not lost upon you with respect to
the fact that this is exactly the same argument the MPAA
and the DGA would use against AMV makers

From a legal standpoint, nobody is really making or losing money from these directly (although the companies may be making money from increased sales). They really don't care about the fan work, as it only helps them. The result? While technically illegal, AMVs fall into the 'nobody cares' legal area.

I absolutely disagree with you. It's not only about money,
it's about _control_. So far we're under the radar. So far.
(ADV may be cool with it, but what about Disney? What
would they think about a "Spirited Away" AMV? Maybe
they don't care, I hope it stays that way -- realistically,
I'm much more worried about the RIAA/CRIA).

Alas, it's gotten to the point of "put up or shut up". I will
not discuss this topic any further unless I can link
to an AMV-generating script people can actually run and
test for themselves...

Cheers!

[trythil]

http://www.dynebolic.org/

Holy crap, that's neat

[shumira_chan]

OK, the _last_ one for real...

Would it be legal to distribute these? You could make the argument that they are still essentially unauthorized derivatives of a copyrighted work, as they were created from direct manipulation of a copyrighted work.

That's arguable, and in fact there's a case about this
mentioned above in this thread. If you listen to:

http://www.law.berkeley.edu/institutes/ ... rplay.html

you will hear lawyers arguing that if I lend you
a pair of scissors and tell you to cut out pages
200 and 201 of "The Hunt for Red October"
because they're boring I am, in fact, violating
Clancy's copyright. There are arguments on both
sides, I strongly suggest AMV-makers listen to this.

Where does it stop? Is there
anything left of "fair use"? Ultimately, you will
_always_ be doing something which is potentially
liable, society has become structured in such a
way that you could be sued for anything no matter
what. So, it comes down to trusting common sense
and your own moral compass (BTW, I'm not talking
about murder here, folks, just putting bit of anime to
music). There are two issues I'm thinking about:
what I think is wrong and what will get me into legal
trouble. I don't think making AMVs (for no profit)
is wrong, I believe that culture flourishes thanks to
derivative works, and that the industry is taking the
definition of a derivative works to absurd extremes (all
the while cranking sequels like crazy, hmmmm, not
much creative process there, is there?) I think the
laws are wrong because they allow them to get away
with it in the first place. But that's the _video_ part,
the audio is a bit trickier since it usually remains
untouched (i.e. you still need to watch "Spirited Away"
no matter how many AMVs you see, but maybe I
don't need to buy that Britney Spears CD anymore ;)
So, both morally (to a lesser degree) and legally I'd
feel more comfortable offering AMV scripts, and plus
I think it would have a few inherent advantages.

This is personal stance, though, nearly impossible
to agree upon in any significant manner. Thus,
objectively, I'm just focusing on discussing the
technical feasibility of script AMVs.

Cheers!

[DemonSpawn]

Personally, I think the whole reason everything is so f*cked up is because it's legal to sell (or be forced to sell) your copyright. Generally, that's how the RIAA and MPAA work, they buy the copyright from the artist, then give a fraction of the money they make as royalties for performing "their song" for them.

Aside from that, there's nothing I can really say that would be new. Excellent discussion, guys.

[OtakuMegane]

What I've figured out on the legality is this: AMVs could be argued as legal or illegal, depending on the amount of editing, and especially on who has the best lawyers.

As for whether anyone will do a damn thing about AMVs...anime companies almost certainly won't. The music videos probably give a boost to sales from people who see one and say "ohh, I wanna see that anime". Also, the makers might enjoy making videos with the DBZ TV series they bought, then go buy the movies as well just to see if there's some good footage to use from them. And I am not aware of any negative impact from AMVs on anime/distribution companies.

Music companies benefit in a similar way. I would have skipped about 7 or 8 CD purchases over the past half year if not for hearing music I like in AMVs. However, since the music is usually the full, complete song, *some* people may just extract the music(if they like it) and not bother to get the CD. Same thing as downloading from kazaa. Music companies ten to be very anal retentive lately because they still don't understand why people don't like to pay the same price for CDs as they paid when they were newer technology(I remember buying a couple NIN CDs a decade ago. They are approximately the same price as they were back then, if not slightly higher. So if anyone takes issue with AMVs, it'll be the music companies, almost certainly.