5 ms·

Lib-Ray: Non-DRM Open-Standards HD Video Format

sandGorgon 14y ago
>The player will be written in Python with the python-gst, pywebkitgtk, and python-gtk Why ? It would much more useful to me (and much less work for the developer) if you could write a VLC plugin.
- Sander_Marechal 14y ago
  VLC has Python bindings. You could use the Python code and make a VLC plugin out of it.
- jvc26 14y ago
  Not that everyone uses VLC - I thought the intention was something a little more flexible.
  - sandGorgon 14y ago
    true - but I would rather be forced to use something that has been downloaded 1 billion times[1], than something that is a custom player just for this codec and is relatively untested (with all due respect) [1] https://news.ycombinator.com/item?id=3970611 https://news.ycombinator.com/item?id=3970611
  - dkersten 14y ago
    But the people who don't use VLC are able and willing to use python-gst, pywebkitgtk, and python-gtk?
zoul 14y ago
Free and independent film are experiencing an incredible boom right now, and the Blu-Ray eco-system with its DRM (or TPM) , region codes, and exorbitant patent licensing fees denies us access to that. So, I think it's time to do something about it. Like stop distributing movies on physical media?
- jeffool 14y ago
  As much as I wish you were right, digital does nothing toward stopping region coding/DRM if we're talking completely legal means. Though, the licensing fees, sure. As much as I'd like to say you're right, I think there are still too many people (here in the US) that just don't have access to the kind of pipes make this feasible. Even considering so many places here that DO have DSL or cable as choices, you have to make the choice of EITHER streaming video OR doing something more intense than web browsing. I was at my cousin's today and ran into the trite example of downloading something on Steam while someone else was trying to watch Hulu. First the quality went to crap, then it just stopped, and Steam was only ("only") pulling 600Kb/s. Sure, there's probably enough of a market to make it a niche business, but you'd have to strike gold to make it turn into a reason for better Internet service. I mean bigger than Netflix. And with ISPs trying their best to turn us into a "pay per GB" society, I don't see physical media going away any time soon.
tar 14y ago
I sincerely hope that this project gets fully funded but I think this comic (http://xkcd.com/927/ http://xkcd.com/927/) is very relevant in this context.
- unimpressive 14y ago
  I was under the impression that this is about patents, and not standard-hipsterism. I could be wrong though.
dantiberian 14y ago
How is this better than a standard video file for the job that people hire a movie to do? I don't think I have ever been glad that I had a movie menu to play from, clicking a file is faster, more portable and more future proof.
aw3c2 14y ago
I don't quite understand what this is about. it sounds to me that mkv already does most of the things?
- sparkie 14y ago
  Misleading title - the project is about a player, not the format. The player happens to use mkv etc. As to why he is reinventing menus is questionable though, since they're already supported by the mkv format.
  - 0x09 14y ago
    Matroska does not have any menu spec whatsoever.
    - sparkie 14y ago
      Oh, right you are! I had assumed they would be in there by now, as it was over 5 years ago when there was discussion about adding them. Appears to still be a draft idea.
DigitalSea 14y ago
It's a noble goal, but a little too ambitious for one person to undertake by themselves though isn't it? Why not just start an open source project with the cash and get some developers on board? Use the money to buy tools and hardware needed, then let others help write the format and test.
josephlord 14y ago
If they are trying to create an interoperable standard the work plan seems to be missing the two most important parts. 1) Test suite of sample encoded files with expected decoder behaviour documented. This shouldn't be a random selection of different people's encodes but a carefully selected set of those pushing the boundaries of the spec. This should be used to test decoders rather than having a reference decoder that has to be copied bugs and all. 2) The counterpart to the above is a file validator that checks as much as possible that any given encoded file meets the specification. Reference decoders only really make sense for codecs where you will expect a particular bitstream output for any input. If they don't do both of those things it doesn't seem much more useful than just making some encoding recommendations for compatibility.
jkn 14y ago
Video File Format: MKV container file with VP8 video and FLAC or Vorbis audio (this standard is similar, but a little more permissive than WebM, and is optimized for use on fixed media rather than downloads). It doesn't seem like they have strong arguments to create a new standard that is almost but not quite compatible with WebM. The inclusion of FLAC is not a strong argument, rather a bad idea, cf http://people.xiph.org/~xiphmont/demo/neil-young.html http://people.xiph.org/~xiphmont/demo/neil-young.html
- sparkie 14y ago
  It's not just ability to use flac - there's a bunch of other restrictions on the MKV format for webm. These include chapters & menus (and ability to use separate media streams for each), support for variety of subtitle/closed caption formats, plus the ability to add attachments like fonts, textures and images for use in those streams. Some of these features have been re-invented in the webvtt format, which is inferior to what was already available.
  - ZeroGravitas 14y ago
    If they have good reason to break compatability with WebM, then they might want to use Opus (Xiph's follow-up to Vorbis) as the audio codec. It's possible that its combination of voice and wideband audio could be ideal for video soundtracks. http://opus-codec.org/ http://opus-codec.org/
    - makomk 14y ago
      Opus is optimised for low-delay, relatively-low-bandwidth audio for stuff like VOIP. It doesn't make much sense for recorded media where you can quite easily hide a few hundred milliseconds of latency.
      - ZeroGravitas 14y ago
        Opus is actually designed to span from low to high bandwidth (see the diagram here http://www.opus-codec.org/comparison/ http://www.opus-codec.org/comparison/) and since it performed surprisingly well against aac and vorbis the developers are currently working on tuning the encoder for standard music/soundtrack type workloads at higher bitrates. I was thinking that the mixed speech mode of Opus might be excellent for documentaries and such with lots of dialogue at Youtube or web bitrates but you're right, I forgot in this case we're talking Blu-ray size video, in which case you can just play it safe and ramp up the vorbis bitrate instead and it'll be a drop in the ocean compared with the video bitrate.
PaulHoule 14y ago
I'd have a little more faith in it if he was asking for more money. Also I think he is belittling what Blu-Ray has to offer. Blu-Ray doesn't just provide 'menus', it provides an environment for writing Java applications that can do serious tricks with video (such as picture-in-picture.) In fact, from an authoring pov, a Blu-Ray disk is more like an Android application with 25 GB of storage attached than a media file. Blu-Ray players have streaming media capabilities so it would also be possible to make a disk that is a "season's pass" to stream all the Red Sox games for a year. Now, you can make a case that this is surplus capacity and that people would do fine with DVD menus or no menus at all. Maybe you're right. Certainly the full potential of Blu-Ray is hardly touched, and I know movie industry people are quite baffled about what to do with it. You can certainly find disks that make you watch annoying trailers -- on the other hand, I can point to many Blu-Ray disks that "just play" and that come with 6 hours of documentary content, commentary and other good stuff.

>The player will be written in Python with the python-gst, pywebkitgtk, and python-gtk Why ? It would much more useful to me (and much less work for the developer) if you could write a VLC plugin.

Sander_Marechal 14y ago
VLC has Python bindings. You could use the Python code and make a VLC plugin out of it.
jvc26 14y ago
Not that everyone uses VLC - I thought the intention was something a little more flexible.
- sandGorgon 14y ago
  true - but I would rather be forced to use something that has been downloaded 1 billion times[1], than something that is a custom player just for this codec and is relatively untested (with all due respect) [1] https://news.ycombinator.com/item?id=3970611 https://news.ycombinator.com/item?id=3970611
- dkersten 14y ago
  But the people who don't use VLC are able and willing to use python-gst, pywebkitgtk, and python-gtk?

zoul 14y ago

Free and independent film are experiencing an incredible boom right now, and the Blu-Ray eco-system with its DRM (or TPM) , region codes, and exorbitant patent licensing fees denies us access to that. So, I think it's time to do something about it. Like stop distributing movies on physical media?

jeffool 14y ago
As much as I wish you were right, digital does nothing toward stopping region coding/DRM if we're talking completely legal means. Though, the licensing fees, sure. As much as I'd like to say you're right, I think there are still too many people (here in the US) that just don't have access to the kind of pipes make this feasible. Even considering so many places here that DO have DSL or cable as choices, you have to make the choice of EITHER streaming video OR doing something more intense than web browsing. I was at my cousin's today and ran into the trite example of downloading something on Steam while someone else was trying to watch Hulu. First the quality went to crap, then it just stopped, and Steam was only ("only") pulling 600Kb/s. Sure, there's probably enough of a market to make it a niche business, but you'd have to strike gold to make it turn into a reason for better Internet service. I mean bigger than Netflix. And with ISPs trying their best to turn us into a "pay per GB" society, I don't see physical media going away any time soon.

tar 14y ago

I sincerely hope that this project gets fully funded but I think this comic (http://xkcd.com/927/ http://xkcd.com/927/) is very relevant in this context.

unimpressive 14y ago
I was under the impression that this is about patents, and not standard-hipsterism. I could be wrong though.

dantiberian 14y ago

How is this better than a standard video file for the job that people hire a movie to do? I don't think I have ever been glad that I had a movie menu to play from, clicking a file is faster, more portable and more future proof.

aw3c2 14y ago

I don't quite understand what this is about. it sounds to me that mkv already does most of the things?

sparkie 14y ago
Misleading title - the project is about a player, not the format. The player happens to use mkv etc. As to why he is reinventing menus is questionable though, since they're already supported by the mkv format.
- 0x09 14y ago
  Matroska does not have any menu spec whatsoever.
  - sparkie 14y ago
    Oh, right you are! I had assumed they would be in there by now, as it was over 5 years ago when there was discussion about adding them. Appears to still be a draft idea.

DigitalSea 14y ago

It's a noble goal, but a little too ambitious for one person to undertake by themselves though isn't it? Why not just start an open source project with the cash and get some developers on board? Use the money to buy tools and hardware needed, then let others help write the format and test.

josephlord 14y ago

If they are trying to create an interoperable standard the work plan seems to be missing the two most important parts. 1) Test suite of sample encoded files with expected decoder behaviour documented. This shouldn't be a random selection of different people's encodes but a carefully selected set of those pushing the boundaries of the spec. This should be used to test decoders rather than having a reference decoder that has to be copied bugs and all. 2) The counterpart to the above is a file validator that checks as much as possible that any given encoded file meets the specification. Reference decoders only really make sense for codecs where you will expect a particular bitstream output for any input. If they don't do both of those things it doesn't seem much more useful than just making some encoding recommendations for compatibility.

jkn 14y ago

Video File Format: MKV container file with VP8 video and FLAC or Vorbis audio (this standard is similar, but a little more permissive than WebM, and is optimized for use on fixed media rather than downloads). It doesn't seem like they have strong arguments to create a new standard that is almost but not quite compatible with WebM. The inclusion of FLAC is not a strong argument, rather a bad idea, cf http://people.xiph.org/~xiphmont/demo/neil-young.html http://people.xiph.org/~xiphmont/demo/neil-young.html

sparkie 14y ago
It's not just ability to use flac - there's a bunch of other restrictions on the MKV format for webm. These include chapters & menus (and ability to use separate media streams for each), support for variety of subtitle/closed caption formats, plus the ability to add attachments like fonts, textures and images for use in those streams. Some of these features have been re-invented in the webvtt format, which is inferior to what was already available.
- ZeroGravitas 14y ago
  If they have good reason to break compatability with WebM, then they might want to use Opus (Xiph's follow-up to Vorbis) as the audio codec. It's possible that its combination of voice and wideband audio could be ideal for video soundtracks. http://opus-codec.org/ http://opus-codec.org/
  - makomk 14y ago
    Opus is optimised for low-delay, relatively-low-bandwidth audio for stuff like VOIP. It doesn't make much sense for recorded media where you can quite easily hide a few hundred milliseconds of latency.
    - ZeroGravitas 14y ago
      Opus is actually designed to span from low to high bandwidth (see the diagram here http://www.opus-codec.org/comparison/ http://www.opus-codec.org/comparison/) and since it performed surprisingly well against aac and vorbis the developers are currently working on tuning the encoder for standard music/soundtrack type workloads at higher bitrates. I was thinking that the mixed speech mode of Opus might be excellent for documentaries and such with lots of dialogue at Youtube or web bitrates but you're right, I forgot in this case we're talking Blu-ray size video, in which case you can just play it safe and ramp up the vorbis bitrate instead and it'll be a drop in the ocean compared with the video bitrate.

PaulHoule 14y ago

I'd have a little more faith in it if he was asking for more money. Also I think he is belittling what Blu-Ray has to offer. Blu-Ray doesn't just provide 'menus', it provides an environment for writing Java applications that can do serious tricks with video (such as picture-in-picture.) In fact, from an authoring pov, a Blu-Ray disk is more like an Android application with 25 GB of storage attached than a media file. Blu-Ray players have streaming media capabilities so it would also be possible to make a disk that is a "season's pass" to stream all the Red Sox games for a year. Now, you can make a case that this is surplus capacity and that people would do fine with DVD menus or no menus at all. Maybe you're right. Certainly the full potential of Blu-Ray is hardly touched, and I know movie industry people are quite baffled about what to do with it. You can certainly find disks that make you watch annoying trailers -- on the other hand, I can point to many Blu-Ray disks that "just play" and that come with 6 hours of documentary content, commentary and other good stuff.