| |||||||||||||
More about the Chrome HTML Video Codec Change (The Chromium Blog)More about the Chrome HTML Video Codec Change (The Chromium Blog)Posted Jan 15, 2011 3:33 UTC (Sat) by wblew (subscriber, #39088)In reply to: More about the Chrome HTML Video Codec Change (The Chromium Blog) by Kit Parent article: More about the Chrome HTML Video Codec Change (The Chromium Blog)
However, Apple also cited a lack of hardware decoder support for Theora.
(Log in to post comments)
More about the Chrome HTML Video Codec Change (The Chromium Blog) Posted Jan 15, 2011 16:33 UTC (Sat) by hitmark (guest, #34609) [Link]
a claim that is basically wrong as long as one go for a programmable DSP or GPU based acceleration.
Still, Apple actually managed to talk Youtube/Google into reencoding all the videos for H264 back in the day for use via mobile app. Also, Apple is a member of the MPEG-LA group (i think they provided some container format patents).
Nokia was the real surprise, and their letter looked like it was written by someone with a poor grasp of english.
More about the Chrome HTML Video Codec Change (The Chromium Blog) Posted Jan 16, 2011 22:06 UTC (Sun) by elanthis (guest, #6227) [Link]
It's not "basically wrong" unless you want to argue meaningless semantics.
If you're going to argue that a DSP or GPU based acceleration counts as hardware decoding, then why doesn't CPU based acceleration count as hardware decoding? They're all high-speed general-ish purpose processors with programmable behavior, after all.
The hardware decoding that Apple and others referred to literally means dedicated circuits that decode the video with very, very high efficiency. Significantly higher efficiency than the CPU, DSP, GPU, or any other general-purpose programmable circuitry can do. Multiple streams in 1080p at 30-60 hz with power draw in the mW range for some of the better devices, for instance.
Google has released completely free hardware designs for VP8 decoding, which will be found in some chips this quarter. It's likely that the batch of Android phones coming out by this summer or at least by next Christmas will all feature hardware-accelerated VP8 decoding. Without this support, VP8 is a no-go for serious use on mobile devices. With this support, Apple's and others' arguments go out the window.
Also keep in mind that Apple and other phone manufacturers desire hardware encoding too. Think of Apple's FaceTime (and any other video chatting feature, or video capture support in phone cameras). That requires a high-speed, high-quality, energy-efficient encoder to transmit the "Retina Display" quality video at acceptable framerates without draining the battery dry in 10 minutes. Video encoding is significantly more expensive than decoding.
Google is still working on its hardware encoding design, but it will be published royalty-free as well once done (probably very soon). We may see that in device by the end of this year.
For all the political maneuvering arguments the .h264 proponents make, the real question will be whether Apple incorporates that VP8 hardware into its next batch of A4 chipsets. The Android and even Windows 7 Phone handsets are all using off the shelf chipsets from companies like Qualcomm, but Apple got into the in-house chip design business. They may stonewall VP8 support in iOS devices. The question then will be how much people care given what will likely be overwhelming popularity of the VP8-enabled mobile devices running Android, WP7, etc. (based on the existing sales figures of Android devices vs iOS devices).
More about the Chrome HTML Video Codec Change (The Chromium Blog) Posted Jan 18, 2011 22:29 UTC (Tue) by daniel (subscriber, #3181) [Link]
"Google has released completely free hardware designs for VP8 decoding, which will be found in some chips this quarter."
Wow and in contrast to my comments about ChromeOS I find the whole VP8 effort brilliantly conceived and executed from beginning to end. I just want to give a big shoutout here about that.
More about the Chrome HTML Video Codec Change (The Chromium Blog) Posted Jan 17, 2011 9:24 UTC (Mon) by DonDiego (subscriber, #24141) [Link]
> Also, Apple is a member of the MPEG-LA group
More precisely, they are a licensor in the AVC/H.264 patent pool, see
http://www.mpegla.com/main/programs/AVC/Pages/Licensors.aspx
> (i think they provided some container format patents).
Incorrect, see
http://www.mpegla.com/main/programs/AVC/Pages/PatentList....
Contrary to a statement that often gets repeated around here there are no meaningful patents on container formats.
|
|||||||||||||