-
Website
http://avc.com/ -
Original page
http://www.avc.com/a_vc/2009/10/open-subtitles.html -
Subscribe
All Comments -
Community
-
Top Commenters
-
ShanaC
1239 comments · 73 points
-
daryn
216 comments · 15 points
-
kidmercury
835 comments · 104 points
-
howardlindzon
207 comments · 71 points
-
Charlie Crystle
205 comments · 36 points
-
-
Popular Threads
-
Top Tracks of 2009
14 hours ago · 49 comments
-
Top 10 Records Of 2009
1 day ago · 73 comments
-
Getting Computer Science Into Middle School
6 days ago · 281 comments
-
Open APIs and Open Standards
1 week ago · 207 comments
-
Thoughts on Blackberry Fail
4 days ago · 77 comments
-
Top Tracks of 2009
Fred did not download the movie legally. It's not yet legally available for download in the US. Before all of you troglodytes start screaming "THEFT! THEFT!", may I remind you that when there's no commercial market for something, there's nobody for Fred to be stealing from. Who has he wronged?
Open subtitles live in a legal grey area -- as soon as there's a subtitle site that's actually making notable revenue (or, one that's taken investment from the VC community) the lawsuits will rain down. For subtitles that are a transcription, not a translation, it's clear that if they're compiled without the "owner's" permission, they'll be considered infringement... and translations will likely be treated as an unauthorized "derived work". That'll provide more than enough ammunition to prevent any decent subtitle site from ever becoming a viable commercial concern.
If you want to invest in this (growing, encouraging) space... invest in breaking down the stranglehold that Big Content has. Work for sane copyright laws. Because what Gotham Gal and Fred did last night should be legal.
Seems like some of the tools are getting there, but it would be nice to add some of the following
1. broader integration of things like Open Subtitles
2. some recognition for contributors (pay, credits, etc)
I love the idea of getting access to new content - there is so much great creative work happening around the world and this just lowers the cost for everyone and should give content producers another way to get the word out about their work.
Adobe has a really cool voice transcription software built into Premier (making it relevant again)
The thing about subtitles/captioning- is that it also makes video more search friendly- a really important function in our video enabled world.
Thanks for sharing this info on opensubtitles Fred- I hadn't heard of it previously.
Closed Captioning is not ADA, it's mandated by the FCC in the US, via the Television Closed Circuitry Act of 1990, please see: http://www.access-board.gov/sec508/guide/1194.2...
Also, subtitles and close captioning are two separate technologies in implementation for the US/Canada, though the two phrases are used interchangeably. In a nutshell they occupy separate "channels" or tracks in broadcast media; thus when encoding for digital distribution, the CC or Subtitle tracks need to be included, often it's not "automatic"
To clarify:
1) Subtitles: this is a pure transcription of dialogue, without any description of stage direction or action on the screen.
2) Captions: In addition to dialogue, captions attempt to indicate all audio content: stage direction, identity of speakers, music, sound effects, etc.
As for accessibility, Apple/iTunes has supported all WCAG/Section 508 guidelines for *years*, IIRC shortly after the guidelines were recommended, in 1998. Most major US vendors are fully compliant with these guidelines as well: Apple, IBM, Microsoft, Adobe, HP, etc. The US Federal Gov't is the largest technology consumer in the world, ergo no Section 508 compliance, so no sales to the Gov't.
You can check Apple's Section 508 compliance here: http://www.apple.com/accessibility/resources/
It's the same for the other major US vendors, just go to their web sites and search for "Section 508" and/or "Accessibility". All vendors are required to post a statement of compliance with Section 508 (there's an actual metric).
To implement closed captioning in iTunes, in Preferences->Playback select "Show closed captioning when available.
As for "Open Subtitles", I'm interested in checking the intellectual property rights here. As mentioned above, subtitles are a direct transcription of the the broadcast dialogue, which in the case of of film, could be construed as content of the "screenplay". In that case copy/property rights would be licensed and/or assigned by geographic region, much like song lyrics.
It's an interesting problem, and once again it's not about the technology.
These tools should be very easy to implement- and almost required.
The FCC ruling is only on shows over 30 minutes btw.
iTunes supports it fine- I was only talking about from the creators side.
I'm with you on he creator's side, which is why I said "not automatic" in my post above. I think Adobe does a better (not great) job on the technology side.
See my post below about working with human transcribers, vis a vis "news". In the MPEG-4 specification, there is description to add text to audio and video via XML. IBM did a lot of fantastic work on that topic circa 2000-2002. Please see: http://www.research.ibm.com/mpeg4/Projects/XMTI...
Essentially, the XMT solution allows a content creator to insert textual content to a video stream via XML.
It's pretty cool stuff, though not fully implemented by any vendor. You can add text programmatically to video via QuickTime especially for 3GPP output.
is money to be made here
There are several companies that provide near real-time transcription for broadcast media, some derived from Gov't research in machine learning / AI.
As an example, Blinx (http://blinx.com) was formed as a technology spin-off from Autonomy (http://autonomy.com) to focus on entertainment, specifically video search. Fetch (http://Fetch.com) is another company focused on content/nuance extraction.
The technology has improved dramatically in the last few years. In 2005, I worked in disaster recovery for Hurricane Katrina. Part of my job was to manage a team of transcribers (crowd-sourced via the Interwebs) that were monitoring broadcast coverage, as well as "closed" emergency response channels.
First and foremost, human transcribers are simple amazing beings. I had a couple of folks that could transcribe at 140 words per minute in the 95% range for accuracy. At that time CNN was still using human transcribers and could post to their web site text transcriptions of their programs within 1-2 hours post broadcast.
Fast forward a couple of years, and I met with several technology vendors indicating that many news vendors had moved to "automated transcription".
So the technology opportunities are out there.
Me, I prefer human transcription. Again, the technology has improved dramatically, but still lacks the ability to infer or interpret nuance. It's a real issue for any form of electronic communication.
Just my .02
As I mentioned, the technology has approved dramatically, it's astounding. I can't or won't quote percentages of accuracy with automation. But here's my recommendation:
Use automation for the bulk of transcription. Please use humans for verification, this is a "must have" for the time being. Maybe two humans for a start-up in an industry without any regulatory/compliance issues. You'll need a good IP lawyer as well.
For entertainment? With the constant online scrutiny you may want to start with 2-3 human reviewers, and double-up on the need for a well-seasoned IP/contract attorney with entertainment experience.
Shazam is pretty amazing, though not widely implemented in the US. Last I checked, Shazam used some form of "audio fingerprinting" technology which is not viable for real-time transcription. In essence, audio/video fingerprinting requires a canonical source/master file for comparison. And, again the human ear is far superior here.
Here's an example for Shazam as I understand it. I'm in a restaurant, I hear a song. I invoke the shazam application on my mobile phone, which transmits audio to a database of acoustic files. After comparison to those source files, Shazam returns an artist/song title.
I'm sure language translation is more processing intensive.
I think even something like google translation could work in movies, because of the type of language (and having video), sure it would work better between some two languages and not so good with some more difficult translations with more challenging structural differences. But I think with what google is doing with their human assisted translation machine, it will continue to get better.
If for nothing else it could be used to do the first version of titles and then continued by human.
How was the movie?
This is a dynamic i also noticed within Foursquare a couple of days ago. Foursquare's database in London is not great, but i saw many people including myself adding new places constantly. I'm pretty sure that the number of places created by users within Foursquare is much higher than in most other LBS services. The dynamic must have something do do with a personal incentive or reward (status, points, checkin) and the knowledge that other users do the same and so, collectively add more value to the entire shared system.
I find this highly interesting.
game" in foursquare
"I dare you to eat a large burrito at El Farolito with Ring of Fire X-tra Hot Habanero Hot Sauce - no water!"
Upload a picture or get one of your friends to confirm that you completed the challenge.
Its a little worrisome though - who does the translations?
And do the films' creators have any say?
I know, its just the movies, but still, wars have started over the correct/incorrect translations of a few words
Here's an example: Clark Gable's famous last line in Gone With The Wind is
"Frankly, my dear, I don't give a damn."
That could be translated as
"Up yours, cunt."
or
"Darling I love you but I can't take the pain anymore."
So it will be interesting to watch as the inevitable happens:
1) filmmakers and writers attempt to assert their rights to control or at least influence translations
2) this technology inevitably gets hijacked for propaganda purposes
get the best ratings
For instance, I have a small grasp of Spanish because of my family's background. I am by no means conversational, but I know enough to at least realize that we are missing out on incredible subtleties of language/character because the translations for films like Motorcycle Diaries, Amores Perros, Maria Full of Grace, et. al., are somewhat lackluster. Issues like this make me wonder what I'm missing in translations from classic films like Bergman's or Truffaut's, let alone the more contemporary, lighter fare.
So, there may be very good reason filmmakers and writers will attempt to influence translation: better accuracy as far as character and dialogue. But controlling the translations? Perhaps... but I would imagine the crowds should eventually win out :-\
And, to your second point, I think that crowd-sourcing translations, even though they could be hijacked for propaganda, may also open the other door to allowing more subversive titles reach a wider audience.
Finally, it may be important to crowd-source translations as many nations become more globalized. Meaning, if we're already missing the subtleties for Spanish-language films from Mexico or Colombia, I wonder what is "lost in translation" in a French-language film from an Iranian director/writer that is subtitled in English...
Okay, I am done with my this-is-for-the-good-of-humanity speech and am now stepping down from the soapbox, hahaha
"Warp" should be "warp", like "Ohm" is, well, "Ohm". Now, some Spanish speakers prefer more spanish sounding words, like turning "Volt" into "Voltio". Should it then be "Warpo"? "Warpio"? The translation is not one to one, but doable:
warp speed = velocidad warp
let's go to warp = saltemos a velocidad warp
warp 2 = warp 2
warp bubble = burbuja warp
Now, the other one is not that hard: core breach = rotura de nucleo, fisura de nucleo.
The phrase I quoted can be translated: "Si saltamos a velocidad warp puede que tengamos una fisura de nucleo y la camara de dilitio explotaria".
The only problem is that I tend to watch movies on my XBox, as it's the only computing device near the projector, and it doesn't support subtitles. I have yet to find a good way to burn subtitles onto a movie (iSubtitle doesn't cut it), so many films remain unwatched, or go through a brutal proxy on the iPhone (Apple has very good subtitle support).
Sigh, maybe it's time to get an AppleTV - what do you run Boxee on?
it is really too bad microsoft won't allow boxee onto the xbox. but you know
how that goes
you can get a used mac mini for about $200-$250 on ebay. that's what i would
do if i was starting over.
my favourote.
J
Also, maybe this is slipping under the radar right now, but will open subtitles go the way of music lyrics, with brain-dead execs and lawyers claiming that subtitles are copyright violations (the movie script is copyrighted) in the same manner that music lyrics are?
I am interested, though, in what you think of the copyright implications of the whole thing. Some years ago, any site that posted music lyrics and had any sort of revenue, even if the revenue was spent wholly on keeping the site up, was shut down by the music industry lawyers. It was their copyrighted work, and only they had the right to make money off it.
Wouldn't you think that the lawyers would go after subtitle sites with even more zeal? Especially if someone dares to put up a translation that the movie studio does not want released yet?
Bear in mind, this was the early 00's, and a lot of underground recommendations about what was legal to own versus illegal to own for rare stuff that no one heard of that was being talked about on Messageboards like AnimeNation. So it all might be folk wisdom being spread around by those who fear no prosecution, because it was terribly unlikely that one's favorite vampire anime/Magna was coming to the US up until recently...
The only "problem" with this whole thing is that the industry doesn't seem to like community made subtitles, and many sites have been took down by lawsuits claiming copyright infrigiment.
Thanks for the kind words. The open subtitle world is certainly full of unanswered copyright and questions of how best to create the subtitles and time coded captions. We think that at dotSUB, we have created an easy to browser based functionality to allow anyone, anywhere, of course with permission of the rights holder, to create time coded captions and subtitles, and have that video rendered anywhere on the web with those multiple languages.
Would be glad to speak with you further about it.
What DL service are you using? I checked with the legitimate ones (vongo, netflix, amazon...) but, none of them had it. That leaves (umm....) Torrents?
Sometimes I think we are so trapped in other people's viewpoints that we don't give ourselves the time to consider the impact of diversity and difference upon on us as a beneficial and self-coordinated raising of our personal awareness. Observing film silences us enough to see how artists see the world. I love what Marshall McLuhan wrote about how it is the artist is who is the first harbinger of change, though he didn't use the word harbinger, he did say that it is the artist who is at the forefront of change, it is the artist who is the first to recognize change way before the majority us, who have not yet crossed that discovery chasm.
The opposite of "open subtitles" for me is "lost in translation", which means that it is our own given intelligence that is the translation mechanism and if we have a desire to raise the bar on our given form of discovery; then the finger of change belongs to us rather than outwards, because then it is our own finger that presses the enlargement of choice revolving in our own headspace. Open subtitles opens far more than just a door to diversity, for foremost it is a conversion towards expectancy of personal exploration and also the joy and calculation of challenge. Diversity is not a finger we point outwards at someone else, for me it is the degree of our own openness, which at least leads me to reconsider my own life behaviour, or at least the singular futility of my own expressed words.
Openness is starkly different in my mind to nakedness, for we do not need to be a stark naked society, but IMHO allow our own personal fashions to cloth us with our own design of life, this personal design is a much more global perspective, a far more patience one that befits our own personal life goals that finds a middle path between being too wrapped up in ourselves and too transparent that we become hollow and formless, redundant of personal meaning. Openness for me is not a tsunami of incoming media but outward choice, a subtitle merely puts a guiding light so we can determine how receptive we are capable of being without becoming an object or structure of media ourselves. Media at worse is an inward missile of interruption, but media at best is an outward growth of our own discovery.
Open subtitles need to be extended to non-fiction media as much as it opens the global media perspective, yet the expectancy of difference or cultural impact is not as nearly as wise or idealistic as I would like to think of it to be, but if entertainment is the first source of that change, then the rest will follow as more and more people become accustomed to a vastly different way of looking at the world.
I welcome understanding Brazilian, South African, Swedish cultures etc because those countries figure in my own personal journey, but I welcome more than that a global dimension we are no longer putting people on alters or pedestals, that one day we won't be consumed by what we think we know, but embrace much more of what we don't know and in that "not knowing" become far more thoughtful and respectful about those who are different from us; and in that difference maybe the subtitle can be viewed as a cute neural pathway for us to personally appreciate life's uniqueness.
Pax Tecum
LiVE #03
[Em]
Incidentally, as an aside, I'd recommend the recent Swedish series of Wallander by Henning Mankell over the recent English-language version starring Kennth Brannagh - strangely both were shot in Sweden with Swedish props, and even Swedish language programmes showing on TV etc. The only difference appeared to be that the main characters spoke English in the UK version by the BBC...
(The Swedish version is available with full subtitles, and was even shown on the BBC in the UK following their adaption, which was fantastic.)
As others have already pointed out, there's a problem and it's inevitably a legal one. In order to legitimately create captions or subtitles, my company has to be granted a licence to do so by whomever owns the copyright in the video. Otherwise it's incontrovertibly a copyright violation for us to use that video's content in our captions or subtitles. Captions or subtitles are a separate copyrightable work, but they incorporate copyrighted content from the video too. It's like a remix of a piece of music.
There may be some wiggle room if you translate the dialogue in a nonstandard way, maybe...? If you could argue that your subtitles did not contain any of the original dialogue, you'd be in the clear, but of course that rather defeats the object. :)
So I applaud opensubtitles.org and its creators, but I have my doubts about it as a viable exercise in its current form. What it really needs is to be run like Spotify, with the active involvement of the video copyright owners. If they just turned around and said it was OK to crowdsub their content, they'd save themselves a fortune and gain millions of extra viewers. Ho hum.
On a more trivial note it's also worth pointing out that there is potential for confusion in the term "open subtitles" - for many decades it has been standard parlance for subtitles which are permanently visible (as opposed to closed subtitles, which the viewer can turn on and off, as on a DVD). Inspired though the folks behind opensubtitles.org may be, they may find it tricky to appropriate a term which already has an important and widely understood meaning.
What's the purpose of translating is often a really good question to ask-
My school got very lucky and was one of the first American Unofficial Openings of Waltz With Bashir because of some deal the school has with Sony.
Waltz with Bashir was already subtitled at that point (having come out of Cannes), My Hebrew may be kind of bad, but I know that there were mistakes made at the direct translation. That being said, it was overall an excellent subtitling because it got at for what I could tell, the flavor of the movie for the dialogue.
OTOH- it could have gone for a much more direct translation of the dialogue. Hebrew has less cursing than the English subtitles implied.
Further on the other hand- why not translate some of the rock music which was there when there was no dialogue? That would have made some of the cultural nuances more extensively explained.
You might need open subtitles just to be able to find out what is the right balance for a variety of people trying to get vast cultural messages across. Translation is an art form as much as explaining a bunch of words.
The other basic though:
Although everyone here is going RaRAra about the idea of open subtitles to movies, ect. I am always going to be slightly annoying on this subject. Cultural authenticity may not be something one wants to lose, and although opening up your culture means that more people can now understand you, it also means you may end up paying the price of becoming more culturally homogeneous with surrounding cultures. It is good to have a few cultural secrets that are inscrutable to outsiders and take some study in order to join into the party. It is what makes translation a joy and a marvel of communication in the first place.
great "foriegn films" i've seen in the last year
Fred downloaded it via BitTorrent. He's said so in the comments.
I know there's a lot of back and forth about in the YouTube debate (e.g., did Google really pay a $1b premium for that site?), or did Google understand that the ability to search video *well* is a possible goldmine...
I not buying those or any DVDs that I cannot share with my wife and other spanish and non-english speaking friends.
In Europe something weird happens. DVDs are released with subtitles in many, many languages except Spanish. I think in Spain they prefer the sound to be in Spanish, or are they using subtitles to divide zones? Spain is Zone 2, but maybe it gets movies after UK, France, Germany, etc.
I don't think people will pay for any entertainment unless it's optimal, and if you have to rip a DVD and find the subtitles, it's easier and faster to torrent it.