Page 3 of 3 FirstFirst 123
Results 21 to 26 of 26

Thread: Upcoming Maxwell GPUs Will Support H.265, But VP9 Is Uncertain

  1. #21
    Join Date
    Mar 2007
    Location
    West Australia
    Posts
    358

    Default

    I feel like the hardware has moved faster than the software for some time now. Why bother making .265 when in 2y time it will probably play back on a phone for less than 1W. I feel like nvidia is just buying time.

  2. #22
    Join Date
    May 2007
    Location
    Third Rock from the Sun
    Posts
    6,583

    Default

    Quote Originally Posted by tuubi View Post
    You missed the point. He said that using 10-bit colour depth and other advanced features most hardware decoders (not to mention encoders) do not support you can get comparable visual quality from the same source material at lower bitrates. That is also fact.
    All h265 encoders/decoders support 10 bit (and beyond). It's part of the spec. Everything is "comparable" btw, you still have degradation and that is fact.

  3. #23
    Join Date
    Aug 2012
    Posts
    384

    Default

    Quote Originally Posted by deanjo View Post
    All h265 encoders/decoders support 10 bit (and beyond). It's part of the spec. Everything is "comparable" btw, you still have degradation and that is fact.
    You have a serious case of selective reading.

    Name one H.264 hardware decoder that supports 10-bit H.264 decoding: None.
    How long has 10-bit Profile H.264 been available? Years.
    How long has H.265 hardware decoders been available, along with H.265 encoders? Exactly.

    H.265 is still essentially not very different from H.264 from a technical standpoint, and offers little benefit over the existing H.264 10-bit profile. VP9 isn't too interesting either. Only Daala is interesting as it's implementing newer, better technologies -- it's an entire generation ahead of H.265.

    Having encoded over 14,000 episodes of media, I can't share the same view as you when you claim degradation as if that's a problem. I've done many blind tests over the last three years and no one can tell the difference between the original source and my encodes. If whatever degradation that occurs is not noticeable, then why does it even matter? Standard BluRay media uses causes incredible bitrate bloat which can be mitigated by re-encoding in a better format. Encoding technologies have come a long way, and with 10-bit, errors made by the encoder/decoder are mitigated to the point where quality looks the same after re-encoding if your CRF value is sufficient enough.

    Want to know what's really degraded? Netflix streams. 4K Netflix is useless since Netflix bitrate starves their content which makes high resolution content practically useless. As it stands today, I still see content getting released on BluRay that's actually just a 720p upscale to 1080p. I'd rather see high quality content compressed at the source with the best methods (Daala + Opus) than the low quality stuff we have today -- at least that would give a reason to not have to re-encode content for sake of getting rid of the bloat.
    Last edited by mmstick; 05-11-2014 at 06:53 PM.

  4. #24
    Join Date
    Mar 2013
    Posts
    44

    Default

    Quote Originally Posted by zanny View Post
    If you live in the US, and are encoding video, say to put on youtube, you are violating MPEG-LA licensing terms by using x264, for example. And how many commercial h.264 encoders are people using? The entire industry seems to be x264 anyway, and it is basically a time bomb waiting to blow whenever MPEG-LA goes all Oracle on everyone, or at least x264 distributors.

    We need an open codec that isn't a legal black hole. I'm still hoping Dalaa can be accelerated with openCL enough to do a good job, because vp9 is more of a stop gap based on old tech and it isn't really competitive with either h.264/5 at all.
    You're never going to get an "open codec that isn't a legal black hole". You're assuming the problem is somehow technical, whereas the problem is the brokenness of the current patent system/ To do ANYTHING complicated means you're going to violate some nonsense that you had no idea anyone would be stupid enough to patent.
    What protects x264 is simply the fact that there's no money to go after. Daala will change nothing. As long as it, too, is not attached to anyone with money, no-one will go after it. As soon as it gets adopted by anyone with money, you'll discover that it's infringing fifteen ridiculous patents that should never have been granted.

    You fix the problem but fixing the patent system, not by imaging that by doing things "right" this one time you'll magically create something that won't attract the trolls.
    Last edited by name99; 05-11-2014 at 10:43 PM.

  5. #25
    Join Date
    Aug 2012
    Posts
    384

    Default

    Quote Originally Posted by name99 View Post
    You're never going to get an "open codec that isn't a legal black hole". You're assuming the problem is somehow technical, whereas the problem is the brokenness of the current patent system/ To do ANYTHING complicated means you're going to violate some nonsense that you had no idea anyone would be stupid enough to patent.
    What protects x264 is simply the fact that there's no money to go after. Daala will change nothing. As long as it, too, is not attached to anyone with money, no-one will go after it. As soon as it gets adopted by anyone with money, you'll discover that it's infringing fifteen ridiculous patents that should never have been granted.

    You fix the problem but fixing the patent system, not by imaging that by doing things "right" this one time you'll magically create something that won't attract the trolls.
    The same people behind x264 were developing a next-gen M4A/AAC but it was shot down by the superior Opus codec developed by the same people making Daala. Now the thing is, Opus is using technology, and is being used, by many big players in the industry. In fact, the entire international mobile network is fueled by opus for making voice calls. Even Steam and Skype are using a technology that is a part of Opus (Silk). No one has gone after Opus though.

  6. #26
    Join Date
    Sep 2012
    Posts
    662

    Default

    Quote Originally Posted by mmstick View Post
    Name one H.264 hardware decoder that supports 10-bit H.264 decoding: None.
    How long has 10-bit Profile H.264 been available? Years.
    How long has H.265 hardware decoders been available, along with H.265 encoders? Exactly.

    H.265 is still essentially not very different from H.264 from a technical standpoint, and offers little benefit over the existing H.264 10-bit profile. VP9 isn't too interesting either. Only Daala is interesting as it's implementing newer, better technologies -- it's an entire generation ahead of H.265.
    So the main difference between 10-bit H.264 and H.265 is that hardware decoders for H.265 actually exist?
    Seems like a good selling point to me.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •