>After starting with a large raw file, the finalized version ready for release is around 30% smaller, around 7GB for a 1080p file.
Huh? Are they recompressing it? How else would it shrink? If they're recompressing it, why don't they just use a capture card and avoid the hassle of reverse engineering stuff?
Netflix serves a tremendous amount of video, a large percent of total internet bandwidth. They spend a lot of time optimizing their video compression, which can be seen from their tech blog, and papers they publish on video compression techniques. There's no reason for an encrypted file to be more than 100 bytes larger than the unencrypted file. I highly doubt Netflix is introducing a 30% overhead for no reason.
> There's no reason for an encrypted file to be more than 100 bytes larger than the unencrypted file.
Video engineer currently implementing MPEG Common Encryption here:
Actually, there is.
MPEG CENC (Common Encryption) uses AES CTR and requires conveying one initialization vector for each audio/video frame. Those are generally 16-byte long.
To make the matter (a little) worse, MPEG CENC adds signalization to the file, to specify which parts of the streams are encrypted or not: indeed, for remux-without-decrypt reasons, each video frame must have its headers in cleartext.
In fine, you get some dozen of bytes of overhead per audio/video frame ; do the math.
Of course, this overhead remains negligible and could by no means explain a 30% size reduction when decrypting.
Aside from the quality loss due to re-encoding - which should be identical to the loss encountered when capturing digital video and audio output, assuming the capture is done correctly - I fail to see anything "obvious".
Sorry - clearly wasn't thinking straight - their alternative approach has this same issue (and having now read the article! - They do plan around this)
Presumably this is still present in the encoded stream they decrypt. I believe the way around this is by prepaid accounts etc and throwaway emails. High end piracy like this is big business.
Correct me if I’m wrong about the IDs in the stream though.
Edit: I just saw parents reply. It took a while to type this out on my phone :/
Transcoding like that is very heavily frowned upon. A scene group would not risk their reputation like that.
I suspect what they are referring to there is the process of remuxing the file and removing extra audio streams, ie. a BD50 BluRay of "They Shall Not Grow Old" comes in at 20GB with two audio streams, DTS-HD and Dolby Digital. The remux of that BluRay comes in at 16GB and has an identical bit-rate on the video stream, but the Dolby Digital audio stream has been removed, as well as the menu graphics (though this would not apply to a WEB stream).
No its not, look at piratebay. By far the most downloaded versions are 1.5-3GB 1080p rips, or much smaller 720p ones. And on some 55" screen, they are just fine
The scene releases are usually the highest quality possible released in exclusively closed private trackers - sometimes only paid accounts are accepted as a member (if at all).
Everything else probably 95% you see in the wild are transcoded to lower bitrate and stripped out of all the extra files in the container (sub/multiple audio stream/highbitrate audio) to smalled files - because thats what most people want:
I remember that in DVB-S (maybe T/C too?) streams there is a constant bitrate, with additional empty packets (PID 8191 in MPEG-TS). Removing those usually results in a ~30% reduction. But I can't imagine that Netflix or Amazon uses such null bytes.
Thanks to content owners/distributors we’ll see a resurgence in piracy.
After Fox, Disney et al pulled content from Netflix, coupled with regional restrictions, there’s literally nothing to watch on Netflix in Sweden except Netflix’s own original content.
Hulu isn’t available outside of US.
HBO is just as bad as Netflix outside the US.
Even iTunes Store is subject to regional restrictions (but the selection yhere is still much larger)
Whenever there's a need, there's gonna be someone that fullfills that need.
Piracy decreased because the alternative became more convenient. It's easier to pay a few bucks on a monthly basis than it is to download the torrents. But, as those fees are increasing alongside the number of services that you need to pay for, piracy is bound to fill in the gaps. It's illegal, but the less people could afford the alternative, the more difficult it is going to be to stop piracy. A domain or two could be shut down, but if TPB is of any use now, it's to remind us that it's not easy to stop a highly-motivated group of people. I'd argue that The Web is still hectic enough to accommodate such a group of individuals.
Piracy was significantly reduced by Spotify and Netflix because it was more cost and effort-effective to just pay a monthly subscription to get most if not all content you need.
And then copyright holders decided it wasn't good enough and crippled Netflix without providing any alternatives (as always). Watch them blame pirates again in the nearest future.
I am not so sure about that. It will always exist, but it is getting increasingly difficult to break DRM. So we may have long periods where the rips are inferior compared to the source (i.e., transcodes via capping). All the new 4K stuff is already quite difficult due to SGX/other hardware enclaves. So it's not a question of just time/dedication to reverse engineer a blob.
I'm not so optimistic. That is precisely the threat model. That is not to say that it is impossible. Side channel exfiltration from stuff like SGX has been shown, but that's a different game really.
Yeah, but people in the scene are good. Like, world class good. Like, 'perform fault injection to dump BootROM from SoC to get management core execution, discover CPU cores with a custom ISA doing DRM, black-box reverse engineer the ISA, exploit code running on those cores to get code execution, exploit silicon bugs to escalate and extract DRM keys from there' good.
It's a challenge to circumvent DRM, so people with the skills to do so will attempt it. I expect that for many it will also be a matter of fighting back against something that is inherently a threat to (digital) civic freedoms.
And because DRM is fundamentally broken (i.e., there is always the analogue hole at the end of the spectrum) it will never prevent piracy on its own.
Source quality and video codecs are so good these days that those "inferior" rips will be much higher quality than, say, DVDs. It's all about "good enough".
Right? I remember watching awful cam rips back in the day. I'm always so busy these days, 90% of the long-form content I watch is on my tablet riding the bus on my commute. Movies are different, but I can't remember the last time I was inclined to see a tv show above 720p. This is all personal anecdotes, but still.
> it is getting increasingly difficult to break DRM
Do you have real examples of where this is the case? Outside of a vanishing minority of titles, DRM-stripped releases appear in a timely manner just as they have always done.
Without going into specifics there are multiple attack vectors for breaking DRM or simply bypassing DRM. Many are unfortunately for content providers trivially easy. The only real limits are for naked, original files which are best possible quality available outside of the content provider networks themselves. Every single transaction point from creation, transmit, transcode/transmux, distribution to CDNs, and finally transmission to the end user is an attack vector. If you're assuming that attacking the DRM directly is the only option or even the best option available, I'm sad to say for content providers, you're unfortunately misinformed.
Protection of property and property rights is seen as part of capitalism though.
Intellectual Property is a strange concept, but there is something going on, after all movies and TV shows have enormous budgets and they depend on revenue, which is still basically made from individual cinema goers and TV channels (ads), and subscribers (no ads).
The never expiring copyright is the problem, and the very strict fair use (eg sampling, and fanfics are illegal unless proper licences have been attained).
This leads to relatively few complex and kind of efficient for-profit content producers, and a lot of very small not-for-profit enthusiasts (who are probably hyper-efficient in some dimensions, but since their competence is naturally limited, they will not be that efficient in all the relevant/required domains). So we have Hollywood, a few independent filmmakers doing the small budget creative stuff, the regional markets with their respective woods (eg India, China, Japan), plus usually countries have some funding for local productions, but that's it.
And maybe even this would be more than enough weren't almost all of the content locked up in archives. (And only available maybe on torrent, or maybe on a random on-demand pay-per-view site.)
I'm not arguing against IP and friends here, just pointing out that pirates offer a better service that people increasingly like to use and are punished for it.
Another thing about piracy that isn't brought up enough is the preservation of media. Historically, most companies are awful at preserving history. And copyright lengths have been pushed to such extremes that it will easily outlive the company that owns it.
The article says that "Subtitle files... are not encrypted". Netflix subtitles are really interesting to me, because they're aligned in different languages. A lot of unofficial SRT files have different timestamps.
The reason is that I'm trying to collect data for learning languages, especially Chinese. Kind of like VoiceTube [1] and Youglish [2] are doing for English. So far it's only for personal use, not a revenue-taking company.
I tried investigating a Netflix stream, but the subtitles weren't plain text! They were distributed as PNG files [3].
Does anyone have more information about this? Do the pirates really OCR the subtitles for their MKV files? (I doubt it). Is there another way to get the plain text? Contact me directly if you're afraid to comment publicly.
It's not a carry-over from their DVD business. It's a consequence of how they choose to stream. My assumption is that image-based subtitles are more compatible across a range of devices and clients, as you get to choose how they look beforehand.
You can definitely send subtitles as raw text and metadata – it's what they do on iOS devices IIRC, since you can change the subtitle appearance on the client.
OCR-ing ASCII text is bearable. I've tried Tesseract and get about 80% accuracy for English. But not Chinese. That's about 40% at best. The only decent OCR for Chinese I've used is Google Cloud Vision, and that costs money.
This is similar to idx/sub files from DVD which are just images and they are OCRd to text sub/srt file. I assume it's just easier for them to render the image than to mess up with text rendering/code page issues/whatever complication coming from huge diversity of devices they support.
> The reason is that I'm trying to collect data for learning languages, especially Chinese.
Subtitles sites (opensubtitles, podnapisi, etc.) have similar data and simple alignment algorithms will work for unaligned subtitles.
> Do the pirates really OCR the subtitles for their MKV files?
OCR is really simple for text. No reason why they wouldn't do it. There were free programs decades ago for OCR when you were ripping DVDs and they worked mostly on character patterns (you would start with an empty database and then just tag the pattern, after a minute you'd have the whole subtitle OCR-ed).
Please can you recommend a "simple alignment algorithm"? I can send some example ENG&CHT subtitles your way. There's a different number of subtitles, shown at different times, breaking lines in different places. It's a mess. If you know how to align it quickly, please tell me! Doing it manually takes me a couple of hours: longer than the whole movie. That's why I've only done it for a very few, specifically for movie nights with friends.
You need good dictionaries, giving you all possible translations of a word both ways. Given that subtitles in both languages will have matching sentences very closely you can easily determine (be it some counting criteria, simple linear model or dynamic programming) which sequence of let's say 3 subtitle lines matches a sequence of 5 subtitle lines in the other translation.
People/place/town names align stuff pretty easily. The fact that a sequence of lines matches some other sequence of lines makes finding the optimal alignment very efficient too.
I'm guessing that just making a set of words for each subtitle line, counting the common words and picking a criteria of deciding if line maps to other line is more than enough. Subtitles are much easier than free-flowing documents because they are time constrained.
Please show me your best transcriptions for this channel [1]. There's over 2 TB of video data. No way to get access to the original PPT files.
At least the subtitles on there are already aligned English & Chinese, but they do require OCR. Message me directly if you want my OpenCV script that can pull the yellow text out.
Tesseract gave me this. The W0「d n! Gnd has bP〔‥mE ‥任Sh 嬰孩降生 道咸肉員 Peace has Come for O… Km… 峒 ﹏m US 靜安 來自‵我君手 `
I'm sorry to continue insisting, but seriously, Tesseract gave me terrible results. Identical lines don't even give the same OCR result, so trying to run partial matches on Google Translate-quality word equivalents isn't going to cut it. The OCR can't even give me the right number of characters! In the end I'm trying to do it manually, but it's really time-consuming.
You should not use Tesseract for subtitles that blend with the colorful background. Some leptonica magic can extract that yellow stuff pretty easily. OpenCV is not as powerful for it, although it does have some leptonica functionality.
I'll try to make a simple one in the next couple of hours.
Although, the task you're asking me to do is different from subtitle alignment you could do on .srt files.
What you can do with leptonica is extract the symbols (letters and ligatures) and manually tag them. You could also extract the lines containing letters and then process them through tesseract.
I have just extracted the U component (YUV colorspace), binarized the image and tesseract just loves it. The players in the background are completely removed.
I would probably use some background removal or some morphological operations from leptonica to make stuff more robust (instead of binarization).
Of course, if tesseract does something incorrectly you can do a quick connected components in leptonica, extract the symbols and do your own OCR (manually). For chinese it's a bit of an issue because there are bunch of symbols, but for english you will be done tagging all the unique symbols in no time.
If you could email me your scripts, that would be extremely useful. And then I'll send you some subtitles where alignment is important. I just used this channel as an example of where I'd tried Tesseract and given up.
https://nikse.dk/SubtitleEdit uses tesseract as one method for OCR. Sometimes there are errors, like I and l but there is a "correct common errors" mode you can use.
Please can you help me try to extract the subtitles from this channel? [1] Tesseract gave me this. The W0「d n! Gnd has bP〔‥mE ‥任Sh 嬰孩降生 道咸肉員 Peace has Come for O… Km… 峒 ﹏m US 靜安 來自‵我君手 `
That kind of error ratio isn't in the realm of "correcting common errors". Surely there must be better software, but I can't find anything that runs locally and can read these subtitles. I've resorted to typing them out by hand, which is super slow because I don't yet speak Chinese, so I have to use handwriting recognition to enter each character.
You probably need to preprocess the subtitles before feeding it to Tesseract. Crop them to size, use some thresholds to only get the subtitle in the image. Seperate by languages.
I own subscriptions for both Netflix and Prime Video and (try to) use them most of the times from my Linux desktop on a 21:9 monitor. I have to force proper full hd resolution using a browser extension but can't find a way to fix aspect ratio / black bars problems. Just yesterday I tried to watch the third season of Stranger Things and I had black bars on all the 4 sides. After boring my SO for about 10 minutes messing around with extensions I just fired up qBittorrent and fixed the problem in 2 seconds using the autocrop feature in mpc.
I end up the torrent way so often that I'm starting to doubt the reasons that push me to support these platforms and their broken features.
Is the desktop userbase a niche so small they can afford to ignore?
On two of the three 4K monitors I own, I have no way of watching any films or TV shows at that resolution whatsoever. There is no mainstream content at all that can be consumed on Linux at 4K through any legitimate service that I know of. This is in spite of me subscribing to a ‘4K’ Netflix service, owning multiple movies in 4K in various forms etc. The only device that supports 4K legitimately is my Apple TV, which I bought simply to stream Netflix at native resolution. Now that I have it, I often find myself renting films on it, but the only options there are for watching on my laptop are piracy. And honestly, piracy is higher quality and a better UI (being able to download in advance, the press play with no worries about buffering).
? If you guys are cracking one 10th of a percentile I would be surprised. I also use netflix on linux desktop (regular monitor though) but I cannot pretend that they should have to care about such a tiny portion of their overall userbase.
They are different issues: Windows and Mac OS users using 21:9 monitors will have aspect ratio / black bars problems, Linux users with standard 16:9, 16:10 monitors will have problems with full hd and 4k resolutions.
And even if these are two small user bases please consider the efforts required in order to fix the problems we are talking about. What could possibly take to lift the block on full hd as I am already doing with some browser extension? How hard could it be to copy the crop detection algorithm from mpc?
Tangential, but I've recently noticed that due to huge fragmentation in the streaming industry (Netflix, PrimeVideo, Hulu, Disney+ et al.), I'm only using two - Netflix and PrimeVideo, and a third torrent streaming service (PopcornTime) when I can't find a specific movie/TV series.
I prefer to use the third option as few as possible, but I find myself having to use it more and more often as the years go by - geolocked content, huge swathes of content being removed on the whims of providers (The Office comes to mind), and 'originals' that are beyond even the so-bad-it's-good category that cannot even begin to fill the void left by the good content leaving.
I cannot ever see myself subscribing to a plethora of services - the best case is subscribing to a service when it's absolutely necessary (HBO/Hotstar) and then unsubscribing when you're done, which is not at all conducive to fulfilling long-term revenue plans of such services.
So what is the endgame of such services? Is it simply to build a huge catalog of content that you'll mindlessly devour whenever you're bored? Or something else?
That 'feature' of Amazon's service really annoys me as well - presumably it's to make it look like they have far more content available than they actually have?
They also have interlaced two seasons of some shows together (noticed this on a couple kids shows for my daughter). So it would go season 2 ep 1, season 3 ep 1, 2-2, 3-2, etc. In general it wouldn't matter for a kids show that doesn't have a continual storyline, but when you only pay for season 2, it then stops at the end of each episode instead of continuing on automatically, and you have to go back to the UI and skip one.
I don't know where you are but Netflix streaming in the U.S. has the first three seasons of "Lucifer" as well as the 4th season sold directly to Netflix.
I find myself watching less and less content, which is not available on Netflix. I even gradually dropped some tv-series I liked, because they were not on Netflix. Everything else (pirated or not) just seems so clumsy in comparison.
I want something simpler: to pay for a movie or show when I want to watch it. From the one spot (e.g. Google Play movies and it's ilk). Shouldn't be too hard. I don't understand why studios don't licence things to get a direct payment for them, instead forcing a subscription that I don't want and won't use.
Reliable recurring revenue is better for long term stability of a business. Joe Average on subscription is better for their overall "health" than the ups and downs of "no-one's buying this month because something amazing is coming out next month".
I wonder how many content providers see the popularity of old shows and think they can pull in the audience, without realising that a big part of the appeal is having all the shows in one place with one subscription?
> 'originals' that are beyond even the so-bad-it's-good category
Couldn't agree more. Besides a very few "tentpole" features, and a small number of standup comedy specials, that Netflix Originals logo up the top left may as well mean "do not watch". If I could exclude them at the UI level, I would. I pay to see top-range feature films and high quality productions, not an endless bargain bin of third-rate filler.
I suspect the "endgame" is that after all the studios have tried out their little walled garden plan and most have failed, they will band together and finally create some kind of unified system, same as the RIAA has done. It'll probably take a good few years, though.
I think this is one reason Disney is going to make money hand over fist. Almost all of their shows and movies are at least OK and often very good. You won’t have to wade through heaps of trash looking for something decent.
This is only possible in a world where nonsensical bullshit is rewarded with a big pile of money.
In a world we'd all want there would be ONE SERVICE on earth that has EVERYTHING. As long as they don't get that I'll be a pirate.
At least the big companies don't suffer from the loss of my money as a little band would do. That's why I still buy vinyl and visit concerts but I have no choice to do that with movies/series I'm afraid.
To the people who always argue for the "free market": Is this what you meant by "competition is good for the consumer"?
I used to watch a lot of Monty Python sketches on YouTube, and in the past few months they've been removed. Last time I looked it was impossible to find the super popular Parrot Sketch!
They're not on Archive.org, no decent results on alternative sites such as DailyMotion, just disappeared.
That's absolutely shocking to me, a huge piece of popular culture is unreachable unless you specifically go pay for it or are lucky to find a collection on a torrent site.
If you live outside of the US and use streaming services you 100% still need torrents to get access to a lot of popular shows and movies you might randomly hear about.
Even content that is constantly marketer directly to me by ESPN is US only (ESPN plus) despite my Canadian geo. It’s annoying.
I live in Canada, where regional laws force the majority of even casual users into piracy. One could not even subscribe to enough number of services to get the amount of content Netflix has in the USA alone. This may be an exaggeration, but honestly, it's been this way for years. I've honestly put a lot of effort into seeking out legal sources for material here in Toronto, and had little luck in certain cases I easily find through a proxy.
How do these pirates make money?
It seems like so much work is/has gone into their operations that it can't just be for karma points.
Edit: I'm referring to the content ripping groups. Apart from the technical and legal challenges depicted in the article, they need to seed their files which (must be) very costly. Also, content is released consistently and extremely quickly.
The common answer is that the motivation is non-monetary: Fun, community service, group pride etc. But I find that hard to believe entirely. The product is just too polished, subjectively resembeing in quality a revenue generating operation rather than a(n illegal) side hobby.
Why does everything have to be about money? Some people do things because it is fun and rewarding. I'd go so far as to say that if you are capable enough to consistently pirate things you are capable enough to hold a well-paying full-time job.
It depends on the group. Some deal with Malware. Most it really is just the draw of the community, the recognition and the fun. I say this from having (formerly) been on the inside.
Not really surprising when you look at how much non-monetary development goes into the open source community.
I suppose that some do it as an alternative to Netflix - they rip a few things to become part of the group, and in return they get first-in access to everything they might want, without the annoying DRM/geoblocking/etc.
The big inconvenience of encrypted netflix content is that I cannot run it through madvr for better upscale 1080p->4k, the paradox is that a free torrent content might in the end take less size and look better. And let's not forget about some phones (poco f1) where netflix only allows lq because of missing Widevine L1.
"The Scene" here also have a close relationship with the Demoscene, CrackmeScene and KeygenScene. These groups work/compete on challenging technical problems involving extreme and artistic programming. Almost all of these groups work with similar (and sometimes closely shared) custom compression, graphics, debug and platform code.
So, on PC hardware, outside of Windos and Macos, Widevine it's the only option available and it's capped at 720p, but the same DRM is used on Android and it's OK to stream at 1080p...
P.S. on Netflix, I should verify on Amazon Prime
Edit: after posting I recalled a plugin for Chrome/Firefox that forced 1080p playback on Netflix, the github page has the details https://github.com/truedread/netflix-1080p
112 comments
[ 5.3 ms ] story [ 168 ms ] threadHuh? Are they recompressing it? How else would it shrink? If they're recompressing it, why don't they just use a capture card and avoid the hassle of reverse engineering stuff?
Video engineer currently implementing MPEG Common Encryption here:
Actually, there is.
MPEG CENC (Common Encryption) uses AES CTR and requires conveying one initialization vector for each audio/video frame. Those are generally 16-byte long.
To make the matter (a little) worse, MPEG CENC adds signalization to the file, to specify which parts of the streams are encrypted or not: indeed, for remux-without-decrypt reasons, each video frame must have its headers in cleartext.
In fine, you get some dozen of bytes of overhead per audio/video frame ; do the math.
Of course, this overhead remains negligible and could by no means explain a 30% size reduction when decrypting.
Creates an obvious problem though.
Aside from the quality loss due to re-encoding - which should be identical to the loss encountered when capturing digital video and audio output, assuming the capture is done correctly - I fail to see anything "obvious".
Edit: I just saw parents reply. It took a while to type this out on my phone :/
I suspect what they are referring to there is the process of remuxing the file and removing extra audio streams, ie. a BD50 BluRay of "They Shall Not Grow Old" comes in at 20GB with two audio streams, DTS-HD and Dolby Digital. The remux of that BluRay comes in at 16GB and has an identical bit-rate on the video stream, but the Dolby Digital audio stream has been removed, as well as the menu graphics (though this would not apply to a WEB stream).
Everything else probably 95% you see in the wild are transcoded to lower bitrate and stripped out of all the extra files in the container (sub/multiple audio stream/highbitrate audio) to smalled files - because thats what most people want:
Small Video easily playable on a regular screen.
After Fox, Disney et al pulled content from Netflix, coupled with regional restrictions, there’s literally nothing to watch on Netflix in Sweden except Netflix’s own original content.
Hulu isn’t available outside of US.
HBO is just as bad as Netflix outside the US.
Even iTunes Store is subject to regional restrictions (but the selection yhere is still much larger)
Isn't piracy illegal though? I assume it will get deplatformed
Piracy decreased because the alternative became more convenient. It's easier to pay a few bucks on a monthly basis than it is to download the torrents. But, as those fees are increasing alongside the number of services that you need to pay for, piracy is bound to fill in the gaps. It's illegal, but the less people could afford the alternative, the more difficult it is going to be to stop piracy. A domain or two could be shut down, but if TPB is of any use now, it's to remind us that it's not easy to stop a highly-motivated group of people. I'd argue that The Web is still hectic enough to accommodate such a group of individuals.
Piracy was significantly reduced by Spotify and Netflix because it was more cost and effort-effective to just pay a monthly subscription to get most if not all content you need.
And then copyright holders decided it wasn't good enough and crippled Netflix without providing any alternatives (as always). Watch them blame pirates again in the nearest future.
I am not so sure about that. It will always exist, but it is getting increasingly difficult to break DRM. So we may have long periods where the rips are inferior compared to the source (i.e., transcodes via capping). All the new 4K stuff is already quite difficult due to SGX/other hardware enclaves. So it's not a question of just time/dedication to reverse engineer a blob.
Also, hardware does get hacked too.
When life gives you lemons you JTAG those keys out the enclave if you have to. Don't worry, life always finds a way.
I'm not so optimistic. That is precisely the threat model. That is not to say that it is impossible. Side channel exfiltration from stuff like SGX has been shown, but that's a different game really.
And because DRM is fundamentally broken (i.e., there is always the analogue hole at the end of the spectrum) it will never prevent piracy on its own.
Do you have real examples of where this is the case? Outside of a vanishing minority of titles, DRM-stripped releases appear in a timely manner just as they have always done.
Pirates are going to prison for providing a better service at lower cost, the core essential of capitalism.
Intellectual Property is a strange concept, but there is something going on, after all movies and TV shows have enormous budgets and they depend on revenue, which is still basically made from individual cinema goers and TV channels (ads), and subscribers (no ads).
The never expiring copyright is the problem, and the very strict fair use (eg sampling, and fanfics are illegal unless proper licences have been attained).
This leads to relatively few complex and kind of efficient for-profit content producers, and a lot of very small not-for-profit enthusiasts (who are probably hyper-efficient in some dimensions, but since their competence is naturally limited, they will not be that efficient in all the relevant/required domains). So we have Hollywood, a few independent filmmakers doing the small budget creative stuff, the regional markets with their respective woods (eg India, China, Japan), plus usually countries have some funding for local productions, but that's it.
And maybe even this would be more than enough weren't almost all of the content locked up in archives. (And only available maybe on torrent, or maybe on a random on-demand pay-per-view site.)
Of course it works marvelously, for example there are shows that otherwise are inaccessible to many. (Animes come to mind.)
If there were some kind of pay as you wish backchannel probably a lot of people would happily give more than zero to content creators.
There is Hulu in Japan, but with a different catalogue afaict.
The reason is that I'm trying to collect data for learning languages, especially Chinese. Kind of like VoiceTube [1] and Youglish [2] are doing for English. So far it's only for personal use, not a revenue-taking company.
I tried investigating a Netflix stream, but the subtitles weren't plain text! They were distributed as PNG files [3].
Does anyone have more information about this? Do the pirates really OCR the subtitles for their MKV files? (I doubt it). Is there another way to get the plain text? Contact me directly if you're afraid to comment publicly.
[1] https://www.voicetube.com/
[2] https://youglish.com/
[3] https://www.slideshare.net/RohitPuri23/timed-text-at-netflix...
You can definitely send subtitles as raw text and metadata – it's what they do on iOS devices IIRC, since you can change the subtitle appearance on the client.
Also, their ingest works with TTML (a text-based format): https://partnerhelp.netflixstudios.com/hc/en-us/articles/215... – this is a fairly widely used standard.
Yes. At least as far as I have seen, which is not scene level but in private tracker discussion forums.
OCR-ing digital text is pretty accurate afaik.
Viki has a lot of subtitled Chinese shows. They generally don't offer Chinese subtitles for the Chinese shows though.
The show itself often will, but those subtitles are just part of the video, not even an overlay.
https://www.youtube.com/playlist?list=PLhIooD7mFhphhT5nDdhK0...
Subtitles sites (opensubtitles, podnapisi, etc.) have similar data and simple alignment algorithms will work for unaligned subtitles.
> Do the pirates really OCR the subtitles for their MKV files?
OCR is really simple for text. No reason why they wouldn't do it. There were free programs decades ago for OCR when you were ripping DVDs and they worked mostly on character patterns (you would start with an empty database and then just tag the pattern, after a minute you'd have the whole subtitle OCR-ed).
People/place/town names align stuff pretty easily. The fact that a sequence of lines matches some other sequence of lines makes finding the optimal alignment very efficient too.
I'm guessing that just making a set of words for each subtitle line, counting the common words and picking a criteria of deciding if line maps to other line is more than enough. Subtitles are much easier than free-flowing documents because they are time constrained.
At least the subtitles on there are already aligned English & Chinese, but they do require OCR. Message me directly if you want my OpenCV script that can pull the yellow text out.
Tesseract gave me this. The W0「d n! Gnd has bP〔‥mE ‥任Sh 嬰孩降生 道咸肉員 Peace has Come for O… Km… 峒 ﹏m US 靜安 來自‵我君手 `
I'm sorry to continue insisting, but seriously, Tesseract gave me terrible results. Identical lines don't even give the same OCR result, so trying to run partial matches on Google Translate-quality word equivalents isn't going to cut it. The OCR can't even give me the right number of characters! In the end I'm trying to do it manually, but it's really time-consuming.
[1] https://www.youtube.com/channel/UC8_emPVKZvOwJnoBVpU0ytg/vid...
I'll try to make a simple one in the next couple of hours.
Although, the task you're asking me to do is different from subtitle alignment you could do on .srt files.
What you can do with leptonica is extract the symbols (letters and ligatures) and manually tag them. You could also extract the lines containing letters and then process them through tesseract.
我們傾倒耶穌腳前
This is my output from tesseract (chi_tra) (for https://www.youtube.com/watch?v=RzIh2pamcwU)
I have just extracted the U component (YUV colorspace), binarized the image and tesseract just loves it. The players in the background are completely removed.
I would probably use some background removal or some morphological operations from leptonica to make stuff more robust (instead of binarization).
Of course, if tesseract does something incorrectly you can do a quick connected components in leptonica, extract the symbols and do your own OCR (manually). For chinese it's a bit of an issue because there are bunch of symbols, but for english you will be done tagging all the unique symbols in no time.
tesseract 3.04.01 leptonica-1.74.1 libgif 4.2.3 : libjpeg 9c : libpng 1.6.34 : libtiff 4.0.9 : zlib 1.2.11 : libwebp 1.0.0 : libopenjp2 2.3.0
If you could email me your scripts, that would be extremely useful. And then I'll send you some subtitles where alignment is important. I just used this channel as an example of where I'd tried Tesseract and given up.
That kind of error ratio isn't in the realm of "correcting common errors". Surely there must be better software, but I can't find anything that runs locally and can read these subtitles. I've resorted to typing them out by hand, which is super slow because I don't yet speak Chinese, so I have to use handwriting recognition to enter each character.
[1] https://www.youtube.com/channel/UC8_emPVKZvOwJnoBVpU0ytg/vid...
On two of the three 4K monitors I own, I have no way of watching any films or TV shows at that resolution whatsoever. There is no mainstream content at all that can be consumed on Linux at 4K through any legitimate service that I know of. This is in spite of me subscribing to a ‘4K’ Netflix service, owning multiple movies in 4K in various forms etc. The only device that supports 4K legitimately is my Apple TV, which I bought simply to stream Netflix at native resolution. Now that I have it, I often find myself renting films on it, but the only options there are for watching on my laptop are piracy. And honestly, piracy is higher quality and a better UI (being able to download in advance, the press play with no worries about buffering).
- Linux Desktop - 21:9 monitor
? If you guys are cracking one 10th of a percentile I would be surprised. I also use netflix on linux desktop (regular monitor though) but I cannot pretend that they should have to care about such a tiny portion of their overall userbase.
I prefer to use the third option as few as possible, but I find myself having to use it more and more often as the years go by - geolocked content, huge swathes of content being removed on the whims of providers (The Office comes to mind), and 'originals' that are beyond even the so-bad-it's-good category that cannot even begin to fill the void left by the good content leaving.
I cannot ever see myself subscribing to a plethora of services - the best case is subscribing to a service when it's absolutely necessary (HBO/Hotstar) and then unsubscribing when you're done, which is not at all conducive to fulfilling long-term revenue plans of such services.
So what is the endgame of such services? Is it simply to build a huge catalog of content that you'll mindlessly devour whenever you're bored? Or something else?
Similar gripe, finding out that only the sequel or the last movie in a trilogy is available: why Netflix, why?
I wonder how many content providers see the popularity of old shows and think they can pull in the audience, without realising that a big part of the appeal is having all the shows in one place with one subscription?
Couldn't agree more. Besides a very few "tentpole" features, and a small number of standup comedy specials, that Netflix Originals logo up the top left may as well mean "do not watch". If I could exclude them at the UI level, I would. I pay to see top-range feature films and high quality productions, not an endless bargain bin of third-rate filler.
I suspect the "endgame" is that after all the studios have tried out their little walled garden plan and most have failed, they will band together and finally create some kind of unified system, same as the RIAA has done. It'll probably take a good few years, though.
This is only possible in a world where nonsensical bullshit is rewarded with a big pile of money.
In a world we'd all want there would be ONE SERVICE on earth that has EVERYTHING. As long as they don't get that I'll be a pirate.
At least the big companies don't suffer from the loss of my money as a little band would do. That's why I still buy vinyl and visit concerts but I have no choice to do that with movies/series I'm afraid.
To the people who always argue for the "free market": Is this what you meant by "competition is good for the consumer"?
I used to watch a lot of Monty Python sketches on YouTube, and in the past few months they've been removed. Last time I looked it was impossible to find the super popular Parrot Sketch!
They're not on Archive.org, no decent results on alternative sites such as DailyMotion, just disappeared.
That's absolutely shocking to me, a huge piece of popular culture is unreachable unless you specifically go pay for it or are lucky to find a collection on a torrent site.
Even content that is constantly marketer directly to me by ESPN is US only (ESPN plus) despite my Canadian geo. It’s annoying.
It's literally self-deprecating behaviour.
Edit: I'm referring to the content ripping groups. Apart from the technical and legal challenges depicted in the article, they need to seed their files which (must be) very costly. Also, content is released consistently and extremely quickly.
The common answer is that the motivation is non-monetary: Fun, community service, group pride etc. But I find that hard to believe entirely. The product is just too polished, subjectively resembeing in quality a revenue generating operation rather than a(n illegal) side hobby.
It’s not about money… it’s about sending a message.
Not really surprising when you look at how much non-monetary development goes into the open source community.
https://youtu.be/xIs_5nfJKu4
P.S. on Netflix, I should verify on Amazon Prime
Edit: after posting I recalled a plugin for Chrome/Firefox that forced 1080p playback on Netflix, the github page has the details https://github.com/truedread/netflix-1080p