Hello, I may be stupid, but I cannot find any data on how "expensive" the Vorbis Decoder is. Looking for an embedded solution I am interested in how many "MIPS" you need to decode a MONO audio stream at reasonable quality (44 kHz, approx. 100 kbps?) Can it be done in a microcontroller, or do I need ultra-expensive DSPs to do it? Thanks for hints Jens --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Hi: Have you looked at Tremor? Tremor is an integer math decoder and sounds like the kind of thing you're looking for. Not sure where the best web resource for Tremor is, but there's a few links on http://www.xiph.org/ogg/vorbis/ You can get a list of hardware that supports Vorbis from http://wiki.xiph.org/VorbisHardware - this may give you some idea what kind of grunt is needed. Sorry, I'm not a developer, just a user, so I can't provide any more specs than this. Geoff. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
On Friday 19 March 2004 16:55, Jens Prüfer wrote:> Hello, > > I may be stupid, but I cannot find any data on how "expensive" the Vorbis > Decoder is. Looking for an embedded solution I am interested in how many > "MIPS" you need to decode a MONO audio stream at reasonable quality (44 > kHz, approx. 100 kbps?) > > Can it be done in a microcontroller, or do I need ultra-expensive DSPs to > do it?A Sharp Zaurus with a 206MHz STRONG ARM SA1100 processor has about 25% CPU load while decoding a vorbis Stereo stream at160kBit/s. And about 22% at 64kBit/s. This has been measured with the default Tremor library (XMMS decoder plugin). Dominik --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> From: Jens Prüfer [mailto:jens.pruefer@gmx.de] > Sent: 19 March 2004 15:55 > To: vorbis-dev@xiph.org > Subject: [vorbis-dev] How many MIPS for Vorbis Decoder? > > Hello, > > I may be stupid, but I cannot find any data on how > "expensive" the Vorbis > Decoder is. Looking for an embedded solution I am interested > in how many > "MIPS" you need to decode a MONO audio stream at reasonable > quality (44 kHz, approx. 100 kbps?) > > Can it be done in a microcontroller, or do I need > ultra-expensive DSPs to do it?Depends what you define as a "microcontroller" and "DSP" :) There's a lot of people calling quite powerful CPUs "microcontrollers" these days just because they have tons of GPIOs, or "DSPs" just because they have a fast multiplier... Most of the ARM based embedded platforms out there will manage "normal" quality Vorbis streams. Anything from a 74MHz ARM7TDMI with cache, MMU and SDRAM will do the job using Tremor with plenty of MIPS to spare. You'll only run into trouble if you have a platform slower than that, or lacking a "flat" cached memory model. What's this intended for anyway - low power, small form factor, or low cost? If it's just a hobby thing, grab an XScale development board. Otherwise, some ARM based CPU with your particular power/form/cost requirement will most likely do the job. - John Ripley. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Jens Prüfer wrote:> Hello, > > I may be stupid, but I cannot find any data on how "expensive" the Vorbis > Decoder is. Looking for an embedded solution I am interested in how many > "MIPS" you need to decode a MONO audio stream at reasonable quality (44 kHz, > approx. 100 kbps?) > > Can it be done in a microcontroller, or do I need ultra-expensive DSPs to do > it? > > Thanks for hintsYou want to browse the 'Tremor' list archives. Tremor is the integerized vorbis decoder that people use for low-power applications. There have been quite a few threads on the list about optimizing for different needs (CPU cycles vs memory usage, etc). See http://www.xiph.org/archives/tremor/200403/0003.html for example. Peter Harris --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> From: Jens Prüfer [mailto:jens.pruefer@gmx.de] > Sent: 19 March 2004 17:31 > To: vorbis-dev@xiph.org > Subject: Re: [vorbis-dev] How many MIPS for Vorbis Decoder? > > > What's this intended for anyway - low power, small form > factor, or low > > cost? > > Ultra-Low-Cost! Definately. Low power would also be nice. I > was thinking of > one of the TI MSP430's maybe? But when I read "200 MHz Strong > ARM" my heart > skips. But more MHz does not necessarily mean more Speed. The > MSP430F169 has > a MAC unit as well, making those nasty multiplications cheap > (1 cycle only). But it only does 8 MIPS :(Yikes. That's really low. I can't actually find any MDCT or FFT references on TI's site... so I don't know how efficient it would be. A brief scan of the user guide reveals that it's also restricted to a 16 bit address space, but if that means 128KB, it's probably going to fit. Sounds like a lot of porting is required, though, if it will actually work. If you're looking for ultra-low-cost that probably rules out anything ARM based. (Grumble - ARM cores are great but they're always too damn expensive). You don't need anything as powerful (or expensive) as a StrongARM, but there's plenty of other microcontrollers around that might do the job a bit less marginally. Don't forget development time when you compare these things for cost :) If you can find a figure for how fast that DSP performs MDCTs (i.e number of clocks for a block size), that would help. - John Ripley. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> From: Jens Prüfer [mailto:jens.pruefer@gmx.de] > Sent: 19 March 2004 18:39 > To: vorbis-dev@xiph.org > Subject: Re: [vorbis-dev] How many MIPS for Vorbis Decoder? >...> > If you're looking for ultra-low-cost that probably rules > out anything ARM > > based. (Grumble - ARM cores are great but they're always too damn > > expensive). You don't need anything as powerful (or expensive) as a > > StrongARM, but there's plenty of other microcontrollers > around that might > > do the job a bit less marginally. Don't forget development > time when you > > compare these things for cost :) > > OK, so which device would you chose for the lowest cost > solution? Since I also > want to implement part of an IP Stack on the same chip (there > is the uIP that > fits in 40k but again I have no idea how "expensive" that > implementation is > concerning CPU cycles), it has to leave a bit of headroom > when crunching on those DCT :)I don't know much about the "ultra low cost" market. The only stuff I've really played with is targetted for MP3 players in the first place, and those are at least several times faster than the MSP430. It's usually just a case of optimisation to get Vorbis to fit and run real time on them. For arcane low power / low cost DSP architectures, it might be a case of never fitting or running real time.> There is the C2000 DSP/Microcontroller fusion series at TI > that has about 40 > MIPS at an affordable price/1000 u. But it does not have to > be TI of course!Sounds a bit more realistic, that one.> I am so totally clueless (but that will change), I don't have > any idea about > memory usage either. You mentioned 128KB could be a hard > constraint? I was > hoping to get away cheap and just use the on board memory of > the Chip. Ok, go ahead, call me an idiot :)There's a fairly large amount of RAM needed for the codebooks and MDCT decoding. I don't know how much Tremor-lowmem is at these days, but I'd say about 64-128KB. If you don't have the internal RAM available, you'll need external RAM. And if it's an arcane DSP architecture, you'll probably need to DMA external RAM on demand. Fun :)> Just imagine a stand-alone IP ready device that has to decode > a mono UDP Vorbis stream as cheap as humanly possible.I hate to say it, but I think a lower profile codec might be a better idea than cramming Vorbis into an ultra-low-cost device. Something like Speex, perhaps? - John Ripley. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> From: Jens Prüfer [mailto:jens.pruefer@gmx.de] > Sent: 19 March 2004 18:39 > To: vorbis-dev@xiph.org > Subject: Re: [vorbis-dev] How many MIPS for Vorbis Decoder?...> Just imagine a stand-alone IP ready device that has to decode > a mono UDP Vorbis stream as cheap as humanly possible.It just occurred to me that a lot of the higher performance embedded CPUs tend to have ethernet, USB or wireless interfaces built in at a cost saving vs separate chips. That might make some of the otherwise expensive higher performance controllers more attractive. - John Ripley. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
Hello My project is to encode an audio signal with a DSP Texas Instruments TMS320DM642 and store the encoded file in a PCMCIA card. I would want to know if that it is possible to implemente the code source of the encoder OGG VORBIs in the DSP. I am beginner, so understand me. I have an other question: In OGG VORBIS Win 32, I want that "stdin" corresponds to my sound card input (Micro) and that the "stdout"corresponds to my sound card output(loudspeakers) because I want to do measures of OGG vOrBIs performances with a MLSSA card(audio analysis card). How can I do that? Many thanks Guillaume <p>>From: Mike S <wickedchicken@plasticpenguins.com>>Reply-To: vorbis-dev@xiph.org >To: vorbis-dev@xiph.org >Subject: Re: [vorbis-dev] How many MIPS for Vorbis Decoder? >Date: Sun, 21 Mar 2004 09:45:05 -0500 > >John Ripley wrote: > >>I don't know much about the "ultra low cost" market. The only stuff I've >>really played with is targetted for MP3 players in the first place, and >>those are at least several times faster than the MSP430. It's usually just >>a >>case of optimisation to get Vorbis to fit and run real time on them. For >>arcane low power / low cost DSP architectures, it might be a case of never >>fitting or running real time. >> > > >I myself am looking around at building something similar. My idea (and I >don't know how far-fetched it is) is to take an unused portable CD player I >have, and 'intercept' the data line that goes into the DSP. Thus, I would >write an ogg stream onto each track of a cd. When the data comes across the >line, the ogg vorbis hardware would decode it and send it to the dsp to be >played. Obviously I would have to deal with CD speed and buffering issues, >but I think it has a chance of working. Since this would be a one-of-a-kind >device (and just a hobby) I was looking for something cheap and easy to >work with. Someone mentioned an XScale dev board, so maybe I'll take a look >at that. > >This doesn't suit my needs, but would something like this, running linux >and with an audio card attached work for the other guy? > >http://www.soekris.com/net4501.htm >-- > >~Mike >wickedchicken@plasticpenguins.com > >--- >8 ---- >List archives: http://www.xiph.org/archives/ >Ogg project homepage: http://www.xiph.org/ogg/ >To unsubscribe from this list, send a message to >'vorbis-dev-request@xiph.org' >containing only the word 'unsubscribe' in the body. No subject is needed. >Unsubscribe messages sent to the list will be ignored/filtered._________________________________________________________________ Trouvez l'âme soeur sur MSN Rencontres ! http://g.msn.fr/FR1000/9551 --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> From: Dan Conti [mailto:dconti@acm.wwu.edu] > Sent: 20 March 2004 02:47 > To: vorbis-dev@xiph.org > Subject: RE: [vorbis-dev] How many MIPS for Vorbis Decoder?...> > From: owner-vorbis-dev@xiph.org [mailto:owner-vorbis-dev@xiph.org]On > > Behalf Of John Ripley > > Sent: Friday, March 19, 2004 4:51 PM > > To: vorbis-dev@xiph.org > > Subject: Re: [vorbis-dev] How many MIPS for Vorbis Decoder? > > > > The 7312 is a fantastic embedded CPU, but it's quite dated, > and 48KB is > > way too little RAM. I believe the internal SRAM on the > Cirrus 7312 is > > also half speed, and very many cycles to access. I ran some > benchmarks > > on it a while back and discovered to my horror that it was about 5 > > cycles random access and 2-3 cycles sequential depending on > core clock > > speed. It wasn't actually worth the programming effort to > put anything > > in internal RAM because cache was way faster, and internal > vs external > > made very little difference to power consumption or performance. > > There are AAC, WMA, and MP3 libraries available that require > less than 24K > of ram; i was able to get each of these running on the > on-chip ram on that > processor. Along with a relatively full fat filesystem, audio > buffering, > simple UI, etc. For that part, SRAM is actually a saving > grace, since random > access to sram is about half as many cycles as random access > to sdram, but still, it's not an ideal setup in many ways.AAC, WMA and MP3 manage that 24KB footprint usually because they keep their codebooks in ROM, or have some overlay system that pages the appropriate book on demand per file, and the advertising doesn't count ROM space. You can't do that with Vorbis because the codebooks aren't fixed. This is even more unfortunate because you can't "compress" the working tables at compile time. On platforms with flash that the CPU can't automatically page (NAND), demand paging of tables can't be done real time, or at the very least not easily.> Do people really put more than ~50K of sram on chip these > days? Guess i'm out of the loop :)50KB ram is so last century, I think even some PICs have more than that now :) ...> I just noticed you work at rio. I thought you guys switched > to the 7312 for > most of your stuff? Maybe that was just a few of the > handhelds. And that rio boombox product that never came to be :)For the flash players (and Nitrus) it's a Sigmatel 3410, which is basically a souped up Motorola 56k with 288KB RAM. The Karma has a PortalPlayer 5003 in it - two ARM7TDMI cores and 16MB SDRAM. The 7312 stopped being useful because it lacks the peripherals we need.> I agree though, in general, if you have a complex UI to > support it's going > to be difficult to fit it in a small memory footprint. Also i > dont really know how much ram tremor wants these days.Both Tremor-lowmem (last I checked) and my own little decoder take in the ball park of 128KB RAM (x86 code and data, similar on ARM).> > I think over the next year we may see a lot of "MP3" player > chips which > > have plenty of internal RAM, e.g 256KB+. Embedded DRAM > (refreshed 1T) > > seems to be the next big thing. Makes my life easier anyway :) > > I really doubt it. The success of the ipod mini throws the > whole flash based > player market into question. Why would chip manufacturers add > more on chip > ram, which results in significantly more expensive parts, > when people are > buying HD based products which have off-chip memory for buffering?I think you're severely underestimating the cost and power consumption of off-chip memory. The flash market does exist for a reason, despite Apple's reality distortion field. - John Ripley. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.