Mashal al-shboul
2012-Feb-28 23:47 UTC
[Vorbis-dev] Need for help about using vorbis in embedded system
Hi All, ?? I am a new member to the vorbis-dev mailing list. i hope that u receive the help that i've been searching for. I need to compress audio samples captured by wireless sensor node (16-bit PCM at 8Khz). can i use vorbis i such an embedded system environment that has the following HW/SW specifications: -416 MHz Microprocessor(ARM architecture, Intel Xscale family) -32 MB RAM Beside that,i can use C language for developmnet. Do these specifications work for compressing the captured audio samples so that i get them in smaller size for transmitting them wirelessly? ? Any Help or Advice Is Very Appreciated. Regards. Mash'al Jordan Univeristy Of science and Technology -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/vorbis-dev/attachments/20120228/0046dba6/attachment.htm
Ralph Giles
2012-Feb-29 00:07 UTC
[Vorbis-dev] Need for help about using vorbis in embedded system
On 28 February 2012 15:47, Mashal al-shboul <shboul8989 at yahoo.com> wrote:> I need to compress audio samples captured by wireless sensor node (16-bit > PCM at 8Khz). can i use vorbis i such an embedded system environment that > has the following HW/SW specifications: > > -416 MHz Microprocessor(ARM architecture, Intel Xscale family)Yes, the reference encoder should work in realtime on a system like that. Note that's it's not especially well-tuned for 8 kHz audio; most people use it at 48 or 44.1 kHz. It will work though. You might also be interested in the new Opus codec we're developing. It's somewhat higher complexity, but offers superiour compression and the reference implementation includes a fixed-point encoder, which may make up for the complexity difference on XScale. http://opus-codec.org/ Cheers, -r
Mashal al-shboul
2012-Feb-29 23:37 UTC
[Vorbis-dev] Need for help about using vorbis in embedded system
Hi Gunter, Thank you for your helpful information. i would like ask about what i found on the website(vorbis.com) telling that integer-only vorbis is available as decoder,does this mean that integer-only processor (not supporting floating point or weakly supporting it through co-processor) in an embedded system can't have vorbis operating well ?. kindly where can i find tutorials or examples on using ogg/vorbis (especially for embedded systems)? Thanks again for the help Regards. Mash'al Jordan univerity of science and technology. ________________________________ From: Gunter K?nigsmann <gunter at peterpall.de> To: Mashal al-shboul <shboul8989 at yahoo.com> Sent: Wednesday, February 29, 2012 12:22 PM Subject: Re: [Vorbis-dev] Need for help about using vorbis in embedded system Normally when using vorbis to compress audio data the vorbis bit stream is packaged into an ogg? container format before transferring it to another computer or saving it. How this is done can be seen e.G. in the source code of the oggenc utility. Ogg is definitively organized in packets - and might be especially suited for your project: If the first four ogg packets are intact (besides the metadata and the information how many streams of which type are packaged in your ogg container they contain the encoder setup information that is indispensable for decoding the vorbis stream again) your stream will still work if any number of audio packets gets lost. All you loose is the few milliseconds of audio data that were contained in the lost packets. If the device you are building doesn't know if any receiver is currently running and thus cannot make sure that the first four packages of your stream have ever arrived you could even decide to start a new ogg container every minute: If you concatenate any number of short ogg files the result is a long ogg file that is still working and valid - but this way you will get a new valid ogg header containing all vorbis encoder settings that are necessary to decode the stream every minute. Kind regards, ? ? Gunter. On 29.02.2012 01:46, Mashal al-shboul wrote:> Thank you for replying. > i would like to add another question.what is the output(as data > structure) that is expected from the vorbis?,can i get the output in a > segmented form that is suitable for packetizing(in arrays or buffer data > structures) at a sender-side, and reassembling at a receiver-side? > note: i am using iMote2 sensor node with PXA271/XScale processor of ARM > architecture,TinyOS operating system,and C with nesC programming languages. > > sorry for asking much > > Regards, > Mash'al > Jordan > > ------------------------------------------------------------------------ > *From:* Ralph Giles <giles at thaumas.net> > *To:* Mashal al-shboul <shboul8989 at yahoo.com> > *Cc:* "vorbis-dev at xiph.org" <vorbis-dev at xiph.org> > *Sent:* Wednesday, February 29, 2012 2:07 AM > *Subject:* Re: [Vorbis-dev] Need for help about using vorbis in embedded > system > > On 28 February 2012 15:47, Mashal al-shboul <shboul8989 at yahoo.com > <mailto:shboul8989 at yahoo.com>> wrote: > >? > I need to compress audio samples captured by wireless sensor node (16-bit >? > PCM at 8Khz). can i use vorbis i such an embedded system environment that >? > has the following HW/SW specifications: >? > >? > -416 MHz Microprocessor(ARM architecture, Intel Xscale family) > > Yes, the reference encoder should work in realtime on a system like > that. Note that's it's not especially well-tuned for 8 kHz audio; most > people use it at 48 or 44.1 kHz. It will work though. > > You might also be interested in the new Opus codec we're developing. > It's somewhat higher complexity, but offers superiour compression and > the reference implementation includes a fixed-point encoder, which may > make up for the complexity difference on XScale. > http://opus-codec.org/ > > Cheers, > -r > > > > > _______________________________________________ > Vorbis-dev mailing list > Vorbis-dev at xiph.org > http://lists.xiph.org/mailman/listinfo/vorbis-dev-------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/vorbis-dev/attachments/20120229/3bba3f1c/attachment.htm
Possibly Parallel Threads
- converting unsigned short sample to signed short sample
- Help about error in linking to libspeex.a
- The decoded voice is just noise and totally different from the original !
- converting unsigned short sample to signed short sample
- using unsigned short instead of short for input frame