This is a note to announce the availability of "Ogg Theorarm", an optimised ARM implementation of decoding libraries for the Theora video code, and Vorbis audio codec from xiph.org. Full details of this code release can be found at <http://www.wss.co.uk/pinknoise/theorarm>, but highlights include: * Full speed playback of a 320x240x25fps clip with a 48kHz stereo audio track on an 416Mhz Xscale based wince device. (39fps if audio sync is sacrificed). * With post processing (deblocking/deringing) enabled, this drops to 23fps. * Uses no hardware acceleration features (e.g. dithered YUV2RGB is done in software, so on devices with such features available, it could go even faster). * Uses ARMv4 instructions only (no reliance on Xscale SIMD or ARMv5/v6/v7/NEON). * Compatible with films encoded with the latest encoder versions. For more information, see the website above, or contact me. Robin -- Robin Watts, Email: <mailto:Robin.Watts at wss.co.uk> Warm Silence Software, WWW: <http://www.wss.co.uk/> P.O.Box 28, Woodstock, Tel: 01608 737172 (or Mobile: 07885 487642) Oxfordshire, OX20 1XX Fax: 01608 737172
On Thu, Sep 17, 2009 at 10:21 AM, Robin Watts <theora at wss.co.uk> wrote:> This is a note to announce the availability of "Ogg Theorarm", an > optimised ARM implementation of decoding libraries for the Theora > video code, and Vorbis audio codec from xiph.org. > > Full details of this code release can be found at > <http://www.wss.co.uk/pinknoise/theorarm>, but highlights include: > > * Full speed playback of a 320x240x25fps clip with a 48kHz stereo > audio track on an 416Mhz Xscale based wince device. (39fps if audio > sync is sacrificed). > > * With post processing (deblocking/deringing) enabled, this drops to > 23fps. > > * Uses no hardware acceleration features (e.g. dithered YUV2RGB is > done in software, so on devices with such features available, it > could go even faster). > > * Uses ARMv4 instructions only (no reliance on Xscale SIMD or > ARMv5/v6/v7/NEON). > > * Compatible with films encoded with the latest encoder versions. > > For more information, see the website above, or contact me. > > Robin >Whoa ! That sounds excellent ! With all those portable multimedia players out there it will be perfect. Do you think a player compatible with the Android platform could be done easily ? Congratulations ! -- ZikZak -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/theora/attachments/20090917/490d6ba3/attachment.htm
Any performance data on running your decoder on the iPhone? Presumably, because the iPhone supports floating-point calculations natively and has a SIMD unit, it might be better to go that route for best performance? Thanks, Claus 2009/9/17 Robin Watts <theora at wss.co.uk>:> This is a note to announce the availability of "Ogg Theorarm", an > optimised ARM implementation of decoding libraries for the Theora > video code, and Vorbis audio codec from xiph.org. > > Full details of this code release can be found at > <http://www.wss.co.uk/pinknoise/theorarm>, but highlights include: > > ?* Full speed playback of a 320x240x25fps clip with a 48kHz stereo > audio track on an 416Mhz Xscale based wince device. (39fps if audio > sync is sacrificed). > > ?* With post processing (deblocking/deringing) enabled, this drops to > 23fps. > > ?* Uses no hardware acceleration features (e.g. dithered YUV2RGB is > done in software, so on devices with such features available, it > could go even faster). > > ?* Uses ARMv4 instructions only (no reliance on Xscale SIMD or > ARMv5/v6/v7/NEON). > > ?* Compatible with films encoded with the latest encoder versions. > > For more information, see the website above, or contact me. > > Robin > -- > Robin Watts, ? ? ? ? ? ? Email: <mailto:Robin.Watts at wss.co.uk> > Warm Silence Software, ? WWW: ? <http://www.wss.co.uk/> > P.O.Box 28, Woodstock, ? Tel: ? 01608 737172 (or Mobile: 07885 487642) > Oxfordshire, OX20 1XX ? ?Fax: ? 01608 737172 > _______________________________________________ > theora mailing list > theora at xiph.org > http://lists.xiph.org/mailman/listinfo/theora >
In message <a960cd1d0909170422l85637e9tf202a85f1143088e at mail.gmail.com>
Claus H?fele <claus at claushoefele.com>
wrote:> Any performance data on running your decoder on the iPhone?
I don't have an iPhone, so I've not tried it. Buy me a Mac and an
iPhone and I'll try and produce some timings >8*)
Alternatively the source code is available so please feel free to
give it a whirl yourself.
> Presumably, because the iPhone supports floating-point calculations
> natively and has a SIMD unit, it might be better to go that route
> for best performance.
I suspect that floating point won't help - however good the iPhones
FP is it's unlikely to be faster than Tremor/Tremolo.
There are potentially operations that could be accelerated using
SIMD, but the single largest consumer of runtime is the YUV2RGB code
(50%+ of the time). I've spent a few days reading up on the SIMD
options available to us (varying from Xscale to ARMv5E, v6, v7 and
NEON), and I can't immediately see how any of that can accelerate the
YUV2RGB process (when we are going to RGB565 at least).
I'd imagine that the iphone probably has some nifty hardware to
accelerate that process though.
The next largest consumer of time is the postprocessing code (if
enabled). I have started on a version locally that might be faster
using ARMv6 SIMD, but sadly, I don't have a ARMv6 device to hand to
test it on.
Robin
--
Robin Watts, Email: <mailto:Robin.Watts at wss.co.uk>
Warm Silence Software, WWW: <http://www.wss.co.uk/>
P.O.Box 28, Woodstock, Tel: 01608 737172 (or Mobile: 07885 487642)
Oxfordshire, OX20 1XX Fax: 01608 737172
In message <bcbc269c50.Robin/Watts at wss.co.uk>
Robin Watts <theora at wss.co.uk> wrote:> This is a note to announce the availability of "Ogg Theorarm", an
> optimised ARM implementation of decoding libraries for the Theora
> video code, and Vorbis audio codec from xiph.org.
>
> Full details of this code release can be found at
> <http://www.wss.co.uk/pinknoise/theorarm>, but highlights include:
>
> [Snip]
>
> * Uses ARMv4 instructions only (no reliance on Xscale SIMD or
> ARMv5/v6/v7/NEON).
I've started to revisit the hotspots, rewriting the code to use ARM
v6 SIMD extensions, but I am unable to test it due to not having an
appropriate device. If anyone out there has a suitable device and
would be prepared to test my code on it, please get in touch!
I hope to look at using v7 or NEON instructions as appropriate too.
Thanks,
Robin
--
Robin Watts, Email: <mailto:Robin.Watts at wss.co.uk>
Warm Silence Software, WWW: <http://www.wss.co.uk/>
P.O.Box 28, Woodstock, Tel: 01608 737172 (or Mobile: 07885 487642)
Oxfordshire, OX20 1XX Fax: 01608 737172