Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×

iPod May Not Have The Horsepower For Ogg [updated] 399

Posted by timothy from the maybe-it's-a-lack-of-a-gumption-bit dept.
An anonymous reader writes "Gizmodo has an interview with a Rio engineer who speculates that current iPods may not have enough CPU power and/or memory to decode Ogg. He concludes that the Minis might be able to do it, and the next generation iPods will certainly be able to. Of course, just because Apple can doesn't mean it will." Update: 06/06 04:44 GMT by T : csm writes with this rebuttal: "According to Monty from Xiph.org (author of the Tremor codec and OGG itself), it should very well be possible to run Ogg on older generation iPods."
This discussion has been archived. No new comments can be posted.

iPod May Not Have The Horsepower For Ogg [updated] More

iPod May Not Have The Horsepower For Ogg [updated]

Comments Filter:
  • I mean really, Apple, what do you have to lose?

    • Re:What is the downside of adding OGG support? (Score:4, Insightful)

      by Anonymous Coward on Saturday June 05, 2004 @06:09PM (#9346460)
      I mean really, Apple, what do you have to lose?


      Developer time and support time, mainly.


      The more important question: What do they have to gain?

      • -I mean really, Apple, what do you have to lose?

        Developer time and support time, mainly.

        The more important question: What do they have to gain?

        this pretty much sums it up from Apple's perspective, but let me expound upon this -- Vorbis is dead for noncommercial use.

        Don't get me wrong, I think Vorbis is an admirable project for a variety of technical and nontechnical reasons. I released music* exclusively in Ogg Vorbis for a while. But most people who are using digital music services are encountering i

        • Re:What is the downside of adding OGG support? (Score:5, Insightful)

          by Durandal64 ( 658649 ) on Saturday June 05, 2004 @11:31PM (#9348194)
          Somewhat off topic, but anyone wanna bet Apple's 'lossless' codec is just their DRM wrapped around FLAC? And yet it was a 30MB+ download!...
          I'll bet you $1000 that it's not.

          The codec is independent of the DRM, and the files generated by Apple's lossless encoder are AAC lossless files with no DRM. Thank you for demonstrating that you have no idea what you're talking about.

          Please send a me $1000 dollar check.

      • Free music (Score:3, Interesting)

        by iamacat ( 583406 )
        mp3 consortium might charge a fee per user*MB or some such for commercial use of the encoder. Or even if they don't charge it now, they could easily add such fees for new licenses in future. If you think that's counter common sense for them, look at SCO.

        Now let's say I offer my own music for free download, and sell some extra tracks to subsidize bandwidth, making it nominally commercial. If I get 100 people downloading 10 songs each daily, this will cost me 30 bucks per day if the license fee is one cent p
    • Credibility in AAC, mostly.

      Y'know, the "better than mp3" codec Apple's trying to push?

    • Re:What is the downside of adding OGG support? (Score:3, Informative)

      by Anonymous Coward
      Apparently it would require more memory and a faster CPU on the iPod. I mean, that's what the article was about.

    • Re:What is the downside of adding OGG support? (Score:4, Insightful)

      by rattler14 ( 459782 ) on Saturday June 05, 2004 @09:09PM (#9347632)
      I mean really, Apple, what do you have to lose?

      probably been said before. But if the current iPod doesn't have enough oomph, then it can be argued that playing an ogg file probably consumes more power. How much more to decode? I don't rightly know. This may be a trivial arguement, but what if playing ogg files shaves an hour of battery lifetime. Then, you have people bitching about the battery life sucking.

      just a thought. I'm sure there are better ones.
  • a Beowulf cluster of ipods then?

    • Since when is a Beowulf cluster joke offtopic? (Score:2, Funny)

      by Anonymous Coward
      Has /. been taken over by pod people? What happend to you guys? Where would we be today without beowulf cluster jokes and Soviet Russia jokes?

  • Vorbis (Score:2, Informative)

    by Anonymous Coward
    It's Vorbis. Not Ogg. Damnit.

  • The name is wrong (Score:5, Informative)

    by iriefrank ( 41550 ) * on Saturday June 05, 2004 @06:08PM (#9346451) Homepage
    Not to be a smartass, but Ogg is just the name of the larger project. The actual audio compression format is called "Vorbis." [angrycoffee.com]

  • Huh? (Score:5, Insightful)

    by Wyatt Earp ( 1029 ) on Saturday June 05, 2004 @06:10PM (#9346464)
    An engineer for a company in direct competition with Apple rips on Apple's hardware. Oh, he's speculating on it.

    "Engineer Hugo Fiennes took a break from his day job as a hardware and firmware designer at Rio Audio (maker of the iPod competitor Karma player, among other things)"

    That's news?

    What's next, someone at Microsoft doesn't like Aqua? Ford engineer says Corvette "not as good as new Mustang"? Fiat engineers doesn't care for Ford Focus?

    • Re:Huh? (Score:3, Insightful)

      by MarcQuadra ( 129430 ) *
      Yeah, this guy seems full of shit, the iPod can ENCODE MP3 in realtime if it has to, it's got a nice beefy ARM CPU, I'm sure it can play Vorbis files if it had a codec.

      And as for memory, the thing has 32MB last I heard, it usually buffers the next two or three entire tracks, so it's got plenty for decoding Vorbis formats.

      • Re:Huh? (Score:3, Informative)

        by jeffehobbs ( 419930 )

        How exactly can the iPod encode mp3 in realtime? The audio it captures via the various 3rd party add-ons is low bitrate .wav files, not .mp3s. All of the various compression and encoding is done by iTunes, then the already encoded files are shuttled to the iPod.

        ~jeff

      • Re:Huh? (Score:5, Informative)

        by Theaetetus ( 590071 ) <theaetetus@slashdot.gmail@com> on Saturday June 05, 2004 @06:38PM (#9346610) Homepage Journal
        Yeah, this guy seems full of shit, the iPod can ENCODE MP3 in realtime if it has to, it's got a nice beefy ARM CPU

        I believe you're referring to the Belkin Voice Recorder when you refer to 'encoding MP3s in realtime'. Not so - the voice recorder stores audio as a mono 16-bit WAV with an 8 kHz sampling rate. Is not encoding MP3s in realtime.

        -T

        • no, it CAN. (Score:4, Informative)

          by gotr00t ( 563828 ) on Saturday June 05, 2004 @09:57PM (#9347906) Journal
          Both the iPod's processors are embedded within a single IC, produced by PortalPlayer. This unit _is capable_ of encoding MP3 audio in realtime, but its just that Apple did not implement this into their software.

          Even though Apple themselves may not support Vorbis audio, ever, the community will implement it if it is possible. Go check out iPodLinux [sourceforge.net]. It has much promise in delivering the things that the Apple stock firmware fails at so miserably.

      • Re:Huh? (Score:4, Interesting)

        by limited ( 17574 ) * on Saturday June 05, 2004 @07:03PM (#9346748) Homepage
        Just because the ipod may have 32MB of buffer for data, doesn't mean it could decode Ogg Vorbis files. As the engineer states in the article, the PP5002 has a relatively small cache, with a relatively expensive penalty if the instruction is not found in the cache but rather in the permanent memory(ROM) of the ipod. Wheter or not the Ogg codec can be optimized to fit in the cache is a different story, and definitely something to be explored before claiming that the gen1-3 iPod's can't handle the job.
        Technical details aside, this guy works for a competing business, and would bad mouth it regardless. Its like posting a story that Rush Limbaugh doesn't like John Kerry- big surprise.

    • Re:Huh? (Score:2, Insightful)

      by TheRealMindChild ( 743925 )
      If this were true, which is more likely:

      A) Manufacturer of the hardware will reveal its inabilities
      B) Competitor of hardware manufacturer will point it out

      Right... that's what I thought.

    • Re:Huh? (Score:5, Informative)

      by in7ane ( 678796 ) on Saturday June 05, 2004 @06:39PM (#9346617)
      Did anybody (rio guy, poster?) at least bother to google [google.com] this? iPods CAN play ogg: (ok under Linux) and not far out of alpha [sourceforge.net] it seems.

      So if somebody managed to get ogg to decode after loading up linux on an iPod, which is not exactly well documented hardware, Apple would not be?

      • Re:Huh? (Score:4, Insightful)

        by tedu ( 647286 ) on Saturday June 05, 2004 @07:04PM (#9346751)
        not sure about you, but i like my music playing back at realtime, not 80% realtime. linux can decode vorbis sure, but i don't count that as playing.

        • Re:Huh? (Score:5, Informative)

          by busonerd ( 534486 ) on Saturday June 05, 2004 @09:40PM (#9347826)
          Speaking as a member of the ipodlinux project, there is a lot of room for optimization. 80% of realtime refers to running a decoder and linux on only 1 of the two cores ( the other is only running a dma-engine) currently. One of the next big tasks is to put both cores to work (for technical reasons, smp is most likely impossible).

          As well, the story is wrong about storing their code in flash. Only the bootloader is stored in flash, which bootstraps the os from the harddrive into sdram, so flash or not, its a non-issue.

          --David Carne
      • You'll see that this is a substandard ogg vorbis playback with only fixed point arithmetic. I guess it technically can play OV files, just not correctly decoded with full floating point quality.
        • This is the trouble with getting ogg vorbis playback on any player that was optimized for wma/mp3 playback. I have a Rio SP250 mp3/wma player hacked to run iRiver firmware. iRiver has been working on ogg vorbis playback for all of its players and it seems they were just barely able to make it work. There older players have somewhat limited bitrates that they are capable of playing (up to about 160 kbps). Given that a lot of these players have special hardware for decoding mp3, it wouldn't surprise me that a
    • Golly, I dunno. Had you RTFA, you'd see that the Rio Karma uses a different model from the same line of processsors from Portal Player, so he might know something about their processor's capabilities. I'm just speculating on that. ;)

  • Comment removed (Score:5, Insightful)

    by account_deleted ( 4530225 ) on Saturday June 05, 2004 @06:10PM (#9346465)
    Comment removed based on user account deletion

  • My Opinion (Score:4, Insightful)

    by luigi22_ ( 733738 ) on Saturday June 05, 2004 @06:11PM (#9346470)
    Adding OGG support would be more than enough to convince me to buy an iPod. I can't really see the downside except for increased strain on the system memory, if what the article claims is true.

    • Re:My Opinion (Score:2, Interesting)

      by Anonymous Coward
      the Rio Karma HAS .ogg support, is $100 cheapter (for the 20 gig version), twice the battery life, tis shorter (a bit wider), and has a very nice UI...why wait for the ipod to get ogg support when there's a good player(s!) out there right now with it and more - you wanna pay $100 for that silver apple on the back?
    • Adding OGG support would be more than enough to convince me to buy an iPod.

      With all due respect, I don't think Apple designs its hardware thinking personally about you (unless you're Steve Jobs, of course). Like most commercial institutions, Apple thinks in terms of "target group(s)". The target group of the people who actually know what OGG is, is too small to be relevant. Sorry guys, get used to it.

      I can't really see the downside except for increased strain on the system memory, if what the article

  • Vorbis Support not Widely Needed (Score:3, Insightful)

    by caffeinefiend ( 681092 ) on Saturday June 05, 2004 @06:11PM (#9346471)
    Apple may offend certain groups, such as Linux Users, by not supporting the Ogg Vorbis system, however the majority of computer users will never even consider using this codec. I submit this for consideration: What Operating system has the largest desktop user market share? Windows, obviously, Apple does not need to support Vorbis because Windows users, in general, have no need for this.

  • This has always been known ... (Score:5, Interesting)

    by phoxix ( 161744 ) on Saturday June 05, 2004 @06:13PM (#9346480)
    http://www.designchain.com/coverstory.asp?issue=su mmer02 [designchain.com]

    Nothing new folks ... that article is almost 2 years old now?

    Sunny Dubey

  • Apple will not (Score:2, Interesting)

    by TwistedSpring ( 594284 ) *
    Apple will probably not support ogg. Ogg has no DRM, and iTunes etc. is based around buying stuff before playing it. I don't own an iPod but I assume I know it plays MP3, I just doubt strongly that apple will add ogg support to it when they probably want to push more people towards iTunes and thus earn more revenue. Ogg doesn't really match up with "revenue", so Apple will probably not support it.

    That doesn't mean to say that 3rd party hackers won't find a way to put ogg on an iPod, of course.
    • As an anonymous poster says below, AAC doesn't have DRM either. It is a layer added by Apple so that they control the specs of the DRM. DRM could easily be added to Ogg Vorbis if someone wanted to do that but I wouldn't count on Apple doing that.

    • Re:Apple will not (Score:5, Informative)

      by mindstrm ( 20013 ) on Saturday June 05, 2004 @06:31PM (#9346581)
      AAC and MP3 do not have DRM either.. the DRM layer was added for the iTunes music STORE, which is a recent addition.... people bought ipods LONG before the iTMS existed.....

      Why do people insist on thinking that ipods and itunes are all just about the store? The majority of ipod owners DONT use the store.

    • What's in it for Apple? (Score:4, Insightful)

      by StrawberryFrog ( 67065 ) on Saturday June 05, 2004 @06:43PM (#9346640) Homepage Journal
      Exactly. Mod parent up. Etc.

      Why would Apple want to support the Ogg Vorbis format? Call me cynical, and I've said this before, but what's in it for Apple?

      Apple support MP3 because it's vital to their business model to get people with MP3 collections on board. Apple supports their own DRM-encumbered format so that they can sell you tunes via iTunes that you can't then share for free.

      What's in it for Apple to support a new format that has no DRM? DRM where they want you to go. MP3 is just the bait.

  • source ? (Score:3, Funny)

    by kayen_telva ( 676872 ) on Saturday June 05, 2004 @06:15PM (#9346491)
    in other news, an ipod competitor says ipod not up to snuff. news at 11

    • Did you RTFA? (Score:2, Informative)

      by genixia ( 220387 )
      If you had you would have seen that the discussion was a technical one explaining why the current iPods can't play Ogg Vorbis and speculated that the next generation iPod probably would have the horsepower to do so.

      It was hardly a "My Karma is better than your iPod" article.

      Face it - It took a Rio engineer to answer the question that most of Slashdot have been asking for years. It's not like Apple have been forthcoming with it.

  • Why OGG? (Score:5, Insightful)

    by Anonymous Coward on Saturday June 05, 2004 @06:16PM (#9346498)
    This article indicates precisely why OGG Vorbis probably isn't a good idea on your ipod or mp3 player... namely, you get 25% LESS battery life. In a non portable, that's fine, but for a portable player with limited battery life... why in the world would anyone choose to get 75% performance with a negligable increase in sound quality (from headphones)?

    • Re:Why OGG? (Score:2, Interesting)

      by nagora ( 177841 )
      why in the world would anyone choose to get 75% performance with a negligable increase in sound quality (from headphones)?

      For one thing my entire CD collection is now in OGG format on my main machine and I'm not about to re-encode for the benefit of Apple's decision to add yet another music format. So, until OGG is an option I'm not interested in an iPod. With it, on the other hand, I can live with 25% less battery life.

      TWW

    • Re:Why OGG? (Score:5, Insightful)

      by mcgroarty ( 633843 ) <brian@mcgroarty.gmail@com> on Saturday June 05, 2004 @06:49PM (#9346671) Homepage
      why in the world would anyone choose to get 75% performance with a negligable increase in sound quality (from headphones)?
      It's not really a linear scale from worse to better. Setting aside the dogmatic choices a lot of the free software people make, the compression artifacts and failures of the MPEG layer 3 and vorbis CODECs is fairly different.

      Even at higher bitrates, mp3 (or its encoders) tend to have a lot of difficulty producing tuned white noise, especially in harmony with better-formed sounds. A breathy voice or a flute can be murder to reproduce. There's also a kind of "glistening" that happens when it tries to represent overtones near the high end of the encoding frequency.

      On the other hand, vorbis seems to more often fail with balances of the frequency range, making some components of sounds louder and others softer than the original, especially with the earlier encoders. Sometimes this merely gives you a too-tuned and prounounced bass range while bands in higher frequencies become too soft. At other times, more complex instruments can lose their character altogether. Steel guitar strings lose the harsher-defined overtones and sound more like nylon, for example.

      Personal preference determines which kind of loss people will choose. Some even pick specific formats to best represent specific styles of music.

    • This article indicates precisely why OGG Vorbis probably isn't a good idea on your ipod or mp3 player... namely, you get 25% LESS battery life.

      and a big part of the reason for that is (next sentence in the article):

      We didn't do a lot of optimisation, so it's running the Vorbis-supplied tremor decoder with only a few tweaks.

      Everyone knows how well-optimized the reference Vorbis implementation is ... all those fools with their alternate 'optimized' codecs are just nutcases, right?

    • ogg can actually increase battery life, in a way (Score:5, Informative)

      by David Jao ( 2759 ) * <djao@dominia.org> on Saturday June 05, 2004 @10:28PM (#9348050) Homepage
      for a portable player with limited battery life... why in the world would anyone choose to get 75% performance with a negligable increase in sound quality (from headphones)?

      The sentence in the article about ogg's battery life is very misleading. Yes, it is true that "you get about 25% less battery life" on ogg vs. mp3. However this comparison is done at the same bitrate -- that is to say, 128 kbps ogg will only have 75% the battery life of 128 kbps mp3.

      But, what the quote doesn't take into account is that nobody uses oggs and mp3s at the same bitrate. I for one find that ogg can match mp3 in sound quality at about 60% of the bitrate. When you use a smaller bitrate, battery life goes up, because your hard drive activity is less. My firsthand experience is that you can get 15 hrs of continuous ogg playback on the karma, if you use a lower bitrate like 64 or 72 kbps. Also, you will note that even if we hypothetically penalized this real-world measurement of 15 hours by a theoretical 25%, it would still be better battery life than an iPod.

      As to your dismissal of headphone sound quality, there are a great [headphone.com] many [headphone.com] headphones [headphone.com] that are good enough to tell the difference. Even without good headphones, 72 kbps mp3 is so bad that anyone who is running out of disk space on their portable can easily justify the switch to vorbis.

  • Can't Linux on iPod Do This? (Score:4, Interesting)

    by Anonymous Coward on Saturday June 05, 2004 @06:18PM (#9346508)
    I thought that the Linux on iPod [sourceforge.net] project managed to get Ogg playback working ?

    Sure - it may not be at 100 percent realtime, but I bet Apple engineers (vs the noble folks who had to reverse engineer the iPod) could manage.
    • You bet they could manage?

      They are engineers, not miracle workers. There are finite limits to the technology here. Even assuming that it is doable, would they make enough profit by adding it to offset the development and support costs involved?

  • I would be impressed, but surprised (Score:3, Interesting)

    by Ars-Fartsica ( 166957 ) on Saturday June 05, 2004 @06:21PM (#9346524)
    Apple is not known for choices - its known for making one approach very easy to use.

    That said, if they build the engine, we will hack it. I look forward to the linux-on-ipod folks dissecting the next gen player and making it play nice with linux as a desktop OS.

  • Technical nitpicking (Score:5, Interesting)

    by mike260 ( 224212 ) on Saturday June 05, 2004 @06:26PM (#9346548)
    [...] This means that running code that doesn't fit in the internal 96kbyte SRAM of the player is very inefficient, both in terms of CPU cycles and power. MP3 and AAC just about squeeze into the internal memory (one at a time, obviously!), but anything that didn't would result in a big power hit - my guess is 30-40%+.

    Surely only code in external RAM would incur this hit. Vorbis decoders spend most of their time doing discrete cosine transforms, which would easily fit into 96K. As would a lot of other performance-critical routines, I'd imagine. So we're talking about a 40% hit on 5% of execution time, which seems pretty trivial, right? Or am I missing something?

  • title should read: (Score:3, Funny)

    by Anonymous Coward on Saturday June 05, 2004 @06:35PM (#9346599)
    "ipod may not care about ogg's cpu-hungry obscure geek-only format"

    • Re:title should read: (Score:4, Insightful)

      by fr0dicus ( 641320 ) on Saturday June 05, 2004 @07:01PM (#9346738) Journal
      Yeah, or 'trade storage space for battery life? No thanks!'

      Or 'seriously, 99% of people don't care',

      or 'it's still compressed you fools, so what if it supposedly sounds 1% better',

      or 'Rio thinks using ogg will make them cooler than Apple'.

      I feel a bit better now. Seriously now though, portable devices are mainly designed to be portable and easy to use. Musical fidelity, albeit important, is really not going to shine through with the crappy little in-ear headphones that people will invariably choose to use. The fidelity is irrelevant and this claim by the Rio chap is more of a drawback of Ogg Vorbis than the iPod in my eyes.

  • The situation in a nutshell (Score:4, Informative)

    by mcgroarty ( 633843 ) <brian@mcgroarty.gmail@com> on Saturday June 05, 2004 @06:38PM (#9346612) Homepage
    The iPod doesn't rely on its CPU to do the decoding for its mpeg formats. The bulk of that is done by a special coprocessor. Whether this is to cut power use or because the slower clock and coprocessor are cheaper than a faster general purpose CPU, I don't know.

    Memory isn't a problem. The full of the iPod's memory is directly addressable, and there are even projects (including iPod Linux) which do Ogg (vorbis, really) decoding, however only at low bitrates. The CPU speed is the strangling factor here. If someone wants to do some hard work, they might be able to raise the bitrate a bit, but owing to people generally relying on VBR encodes, it's going to be difficult to fully enable people's libraries, even when they think they have mostly low-bitrate tunes.

  • Ogg (Score:4, Interesting)

    by Err ( 21062 ) on Saturday June 05, 2004 @06:54PM (#9346695)
    iPod support for Vorbis would be cool.

    What I would truly love would be iPod support for Ogg Speex. I download quite a few audio lectures/interviews, and if the iPod supported Speex, I'd buy one ASAP and go on a campaign to get a few organizations I deal with to put their stuff out in Speex, not just mp3 and wma. For that matter, I'd love to be able to encode my audio books in Speex and have then on the go.

  • Rio Karma is WONDERFUL (from a GNU/Linux user) (Score:4, Interesting)

    by donfede ( 6215 ) * on Saturday June 05, 2004 @06:55PM (#9346710) Homepage
    Forget the ipod, the Rio Karma is wonderful from a GNU/Linux users perspective!

    It plays all my ogg files without problems (a friends iriver could only handle lower bitrate ogg files).

    I could upload music to it quickly and easily from my linux desktop using their java gui and connecting to the rio karma across my lan.

    As I use this player to drive my car speakers (I only have an amp, no head unit), it was very important that the interface be user friendly. This is where I had seen the ipod shine, and where I was doubtful about getting the rio karma (as I knew no owners of one and had not seen a showroom model). However I (and several passengers) found the rio karma interface to be as friendly, if not more so, than the ipod.

    The rio karma was cheaper than the ipod, has more features, and is more cross platform. I have no regrets and strongly recomend it to music fans.

    donfede

  • *Why?* (Score:5, Insightful)

    by adun ( 127187 ) on Saturday June 05, 2004 @06:57PM (#9346714)
    Seriously, who cares about Vorbis outside the faction of *nix users with +1 Amulets of OSS Awe?

    Apple's primary market are the throngs of not-quite-but-almost-technologically-literate end users out there who see gadgets as tools, not lifestyles. Does this afforementioned throng care about Vorbis? No. Should Apple therefore care about Vorbis? No.

    Get the fuck over it, already.

    • Re:*Why?* (Score:3, Insightful)

      by FrostedWheat ( 172733 )
      Seriously, who cares about Vorbis outside the faction of *nix users with +1 Amulets of OSS Awe?

      The same was said about MP3. Who cares about MP3, few computers can decode it in real time anyway...

      I don't understand why everytime there is an article about Ogg here loads of people rush to write about how pointless it is and there's no point using it.

      Get over it!!!

      It's a great format! There's no question about that. People who use it would love to see it better supported. It makes sense! So where's the

  • mild Gizmodo rebuttal from Monty (Score:5, Informative)

    by xiphmont ( 80732 ) on Saturday June 05, 2004 @07:07PM (#9346760) Homepage
    I sent a mild rebuttal to Gizmodo about Hugo's article back when Gizmodo first ran it. As it hasn't appeared on Gizmodo, here's the text:

    Hi there; it's interesting to see one's software being talked about by others as if it fell from the sky :-)

    First off, the original iPod does indeed have the horsepower for Ogg; the original Tremor codec written in C used approximately 40MHz of a Cirrus Maverick (720TDMI), an ARM7 chip somewhat more underpowered than the chip in the iPod. What the Maverick often has that many other DSPs don't have is access to alot of memory. This was indeed a stumbling block for a while.

    Since then, we've made three mostly seperate branches of the Tremor codec line (used in the Rio), each tailored to specific CPU and memory structural differences found in different DSP architectures. Hugo Fiennes didn't mention which he was using... or if he was aware there are multiple different branches today (although I expect he is aware, it's worth mentioning).

    From the story:

    "The 5002 has a "broken" cache (1 wait state per access for program or data, meaning you effectively have half the effective clock rate when running code from external memory). This means that running code that doesn't fit in the internal 96kbyte SRAM of the player is very inefficient, both in terms of CPU cycles and power."

    He didn't say if he meant code, data or both, but modern Tremor can fit both comfortably into this space. This is still substantially faster than the ARM7 DRAM-based access Tremor was originally designed for (using SRAM as a random-replacement cache with 7-14 wait-states on a cache miss).

    Also, he says it uses more power but also says they didn't optimize much and so, they're mostly using the stock ANSI C Tremor decoder, written by a single engineer (me) in a month as a 'starting point' to help other engineers write a Vorbis decoder for their own platforms. The mp3 playback is likely handcoded assembly written by a professional team focused on only that task. This is in fact astounding! It's also a testement to the power of good modern compilers. I smile every time GCC soundly beats me at optimizing.

    I agree that the newer iPods are more likely to decode Ogg and Vorbis with ease. I do, however, strongly believe the original iPods can also do so with room and cycles to spare.

    Monty
    TD, Xiph.Org

  • vapours (Score:3, Funny)

    by crackshoe ( 751995 ) on Saturday June 05, 2004 @07:14PM (#9346785)
    Heaven forbid that a device can't necesarily do something it wasn't designed to do. lordy lordy, i may catch a vapor.

  • This points to an issue for lossy codec design... (Score:5, Informative)

    by Entropius ( 188861 ) on Saturday June 05, 2004 @07:16PM (#9346799)
    Most codecs are designed to simply give the best audio per kilobyte, and recent tests point to (a retuning of) Ogg Vorbis as near the top of the heap. It's certainly achieved what it set out to do: a Free audio codec with fairly wide acceptance that sounds better than mp3/wma under most circumstances.

    Vorbis, of course, takes much more CPU muscle to decode than mp3. The difference may be between 0.1% and 1% of my Athlon XP[1], but obviously on an iPod it matters.

    Maybe it's time for some group to look at weak-CPU audio codecs? You've got to balance audio-quality-per-bitrate with expense (and power consumption!) of CPU required to decode in realtime.

    There's got to be something out there that sounds better than mp3 but can still be decoded with a cheap processor using an amount of power that's not really significant compared to the amplification/transmission circuitry required to get the signal out of the device.

    Ideally this could be done on the decode side: write a codec that produces Vorbis-quality results when decoded by a fast CPU, but that could be decoded by a slow processor to produce a good-enough signal. This would solve the current dilemma: do I encode in vorbis to save disk space/get better quality, or mp3 to play stuff on portables?

    [1]What a stupid name for a processor.
    • That was the original purpose of the MP1 and MP2 formats (MPEG Layers 1 and 2, respectively). They pretty much disappeared after most computers became fast enough to decode MP3 easily. I don't think there's really enough demand for modern low-complexity codecs. After all, LAME-encoded MP3 still sounds better than many "modern" formats, and it's fairly easy to decode. Apple seems to have done a decent job implementing AAC in hardware, and Monty thinks [slashdot.org] that Vorbis on iPod is quite doable. Since hardware is on

    • Re:This points to an issue for lossy codec design. (Score:5, Interesting)

      by adolf ( 21054 ) <flodadolf@gmail.com> on Saturday June 05, 2004 @09:50PM (#9347862) Journal
      Let us assume that it really is a CPU horsepower vs. file size vs. sound quality issue.

      Looking at two extremes:

      Realaudio files used to play justfine, Back In The Day, on a slowish 486. It sounded like shit, but it worked fine. Of course, this same 486 was incapable of playing MP3s. For that, you generally needed a Pentium, and preferably a fast(ish) one.

      And of course, these days, it just doesn't matter. MP3 playing consumes so little CPU time that nobody gives a thought to it running in the background. In other words, the hardware finally caught up (some time ago, really).

      Fast forward, and things are the same, only portable. MP3 files play justfine, on just about everything. My old Riovolt SP-250, after a lot of effort from the Xiph folks and iRiver, is able to play some Vorbis without a hiccup.

      Newer units play all Vorbis justfine, though. They use even less power doing it, and cost less than my SP-250 did. In other words, the hardware is already caught up.

      Sufficient CPU power to play such new-ish formats as Vorbis will eventually creep into more products as the cost of CPU power decreases (eg. Moore's Law).

      I'd like to forecast that it'll be easier, cheaper, faster, and better to simply wait for CPU power to catch up across the board, than to go ahead and invent a scalable codec. By the time you're done making the thing, no matter how brilliant it is, CPUs and DSPs will have advanced the price/performance ratio sufficiently that your efforts will fade into obscurity, just like intel's indeo video format[1] of more than a decade ago.

      Meanwhile, any foolish manufacturers or software developers who jumped on your scalable codec-bandwagon will watch their efforts fizzle and die, as people regroup to support formats that Don't Suck, like our existing OGG Vorbis.

      That said, if you must tinker with software, do feel free to help improve Vorbis. Make it faster, make it smaller. Make it shit golden eggs, whatever. But don't reinvent the wheel without first examining where the rest of the world will be by the time you get done.

      [1]: indeo was created as a high-ish quality, high-bitrate video format, designed to be encoded once and played anywhere. Framerate and quality would drop on low-end devices, while things would be more pristine on faster machines, all from the same source file. It died a quiet death when inevitable increases CPU speed made it a non-issue. Subsequently, better and more-intensive codecs like MPEG1 took over. The near-universal playability, and use, of the previously-hideously-intensive DivX family of codecs drive this point home.

  • Rio Karma (Score:2, Interesting)

    by exigentsky ( 771810 )
    I currently think that the Rio player is a much ebtter bang for the buck than the iPod. * It's smaller and just as stylish * It's much cheaper * It supports more formats * It doesn't lock you down like Apple does and works on mroe platforms well * It has a longer battery life * etc. What exactly is the reason to buy an iPod if you are not into the whole "online music store" thing?
  • Amiga programmers (yes, I know Amiga's dead (long live Amiga!)) can get Ogg Vorbis playback in real time on a 50 MHz m68060. They can also get >10 fps out of the Quake 2 engine on an '060, too. Considering that the ARM is quite a bit faster, it can certainly be done if someone puts in the effort.

    John Klos
    Running Amigas for more than a decade.

  • Neuros Audio (Score:3, Funny)

    by daserver ( 524964 ) on Saturday June 05, 2004 @08:42PM (#9347471) Homepage
    Thank god I bought a Neuros [neurosaudio.com] so I don't have to care about this :)

  • hwardware hackers... (Score:3)

    by Richthofen80 ( 412488 ) on Sunday June 06, 2004 @12:35AM (#9348467) Homepage
    you have just been issued a challenge.

    geeks are a group that hate to be told they 'can't'.

  • You down with OGG? (Score:4, Insightful)

    by telstar ( 236404 ) on Sunday June 06, 2004 @01:10AM (#9348601)
    Yeah, you know me!

    I think OGG's major hurdle is that it's trying to solve a problem that most people aren't aware exists. Storage is cheap, and getting cheaper. For the vast majority of listeners ... they're happy with the sound quality. MP3 is huge 'cause it was first and it was free. ACC is huge because it's tied to iTunes and iTunes is the first pay-per-use DRM system. Unless OGG can offer something new, it'll have a hard time gaining support.

Slashdot Top Deals

BLISS is ignorance.

Close