iRiver released news announcement detailing their plans for Vorbis playback.

In summary, most of their players are sure to support Vorbis, with the possible exception of the iFP-1xx line, and iMP-50, 100, and 150 line. Getting Vorbis running on the low-end iMP line may be difficult, and is quite difficult with the iFP-1xx line.

iHP-120 will support Vorbis out of the box, while the other iHP-1xxs will receive firmware upgrades in October. Next in line are iMP-550, 400, iFP-3xx, iMP-350, and finally, iMP-250, which will start getting upgraded firmware in December.

iMP-250, 350, iFP-3xx users will get a choice between MP3+WMA firmware and MP3+Vorbis firmware.

For the products to support Vorbis starting from December, iRiver has got 128kbps playback working, but they're aiming for a December release which supports bitrates up to 360kbps. They've had no such luck with the iFP-1xx line, however, and if no "breakthrough" in Vorbis decoding is found, then no Vorbis for the iFP-1xx line. They say they're still working on it though.

The same in English:
http://iriver.com/company/news_view.asp?idx=355

Quote from iriver site:

Thank you very much.

That should have read "a choice between 'MP3 & Ogg' firmware and 'MP3 & WMA' firmware".

... must
... resist
.... urge
..........


[mindless retort mode]
WELL, Thats a f*cking hard choice, isn't it?!
[/mindless retort mode]

...ahhh

Lev, I at least have to congratulate you for trying to resist....

- M.

*APPLAUDS LOUDLY*



MUST....GET......IRIVER........SOON..........

RoFL!!! Was there even a choice there?

Why do they continue to call it OGG? It boggles my mind!

Thanks for the info!

Why is it so difficult to get a Vorbis decoder running on an iFP-1xx? I mean they have >64 MB available...

great, just deleted all my Vorbis files yesterday ((((((( ... but i'm so happy ... now, we only need vorbis 1.1 (tuned one

Maybe there isn't enough space on the firmware to handle another codec.

Regards

They only have 64MB of storage, not 64MB of ram

For me, now there is. Will I get the Karma or the iHP-120? Let see which makes it out to market first. I'd also like to see a price on the 20GB iRiver player.

Finally...a straightforward announcement from iRiver

One bummer though...

As announced, they will support only 44.1 kHz files. Pull anything off a DVD and the sample rate (usually) will be 48 kHz. Resampling cuts quality (marginally) and undoes any dithering. This is a step in the right direction, though.

xen-uno

I figured they used some kind of RAM in a solid state player . Maybe it's too slow for the enormous buffer needs of Vorbis ?!

What does that mean exactly? Hopefully it doesn't mean that anything over 128kbs won't play? That would be all of my files.

What a great news!

I think we need PEELING now. Does anybody know when VORBIS team going to make support for it?

I just thought about terrible thing. iRiver says their decoder will be capable of playing up to 360 kbs. But as I know Vorbis Ogg uses up to 520 kbs. So what will be iRivers' decoder? CBR? Or ABR? It will not be able to decode files encoded in quality managment mode. Am I right?

Well. One must assume that it can't decode a file with an average bitrate above 360 kbps within the buffer, and I don't know how many seconds the iRiver buffers.

iRiver's buffer works also on Audio CD tracks (more than 1100 kbs). So I don't think the problem lays in bitrate and size.

Hmm, I am confident that when (IF) they do release this Ogg firmware, that:

It will fail for most people.
..
Then, there will be lots of talking.
..
And then an updated firmware release will be released within a week.
..
And all will be well, no matter what bitrate you use
..

Is it safe to assume that the iHP's will also display Ogg Comments, as opposed to the ID3 tags that they use for MP3?

Guess that IMP-150 of mine will be featured on eBay soon

Altough, on the other hand, I don't seem to have much use for Vorbis (now)

I read on iRiver's site that implementing Ogg on iMP-150 isn't impossible and work continues.

Finally, so Ogg on the IMP-350 in 2004 just as they hinted earlier this year. I don't think I'll care by then.

Until yesterday I still don't understand why Xiph call this "The year of the Fish" , almost no process is made,even Vorbis 1.0.1 but now I can feel the "FORCE"

/edit


stupid me . WTF is ACC BTW

Great sig!

They cite the lack of SDRAM as the reason.

iMP players use SDRAM for anti-skip buffering. iFP-3xx players use it for buffering while encoding from the microphone. iFP-1xx players don't need anti-skip, and they don't do encoding either. So they don't have any SDRAM.

But their current Vorbis decoding implementation uses SDRAM, which they cannot do without. Yet.

Actually Vorbis seems to use a very different buffering approach during decoding than CDA or other audio formats. My experience is not as a developer, but I did do pretty extensive tests on the Tremor decoder for the PhatNoise (/Kenwood) Music Keg for Vorbis-encoded files, and I found that (in it's current incarnation) the maximum nominal bitrate it can do consistently is 136kbps (-q 4.25, Post 1.0 CVS). However, it decodes 1411kbps PCM WAVs just fine and ~750-850kbps FLACs fine as well.

But give it, say, a -q 5.00 Vorbis file, and it'll balk if the actual bitrate spikes too far over ~210kbps. I have some Kraftwerk and Fatboy Slim tracks that spike up to ~210kbps even at -q 4.25, and it still doesn't fail, but the decoder is close to it's limit. I measured that anything much over that bitrate would cause playback to skip and distort in pitch. The guy who modded the Tremor code for this platform said it seemed the buffering wasn't keeping up with real-time playback, and the buffer would run out periodically, causing these problems (if I'm remembering his description correctly).

A specific example, Fatboy Slim's Kalifornia would skip and distort severely in the opening 14 seconds if encoded with anything over -q 4.25. You should hear -q 10...~499kbps nominal. Playback sounded like a guitar phrasing pedal was being used to drop the pitch dramatically, and there was also bad distortion and skipping (playback stopping entirely for fraction of a second as the buffer "refilled") every couple of seconds.

So playing back CDA at 1411kbps seems to require much lower memory and/or buffering overhead than decoding Vorbis at considerably lower bitrates (even using the integer-based Tremor libs).

Also, for reference, I used FLAC exclusively on the same device for a month or so with no audible playback errors. My FLACs average between 750-850kbps or so.


Edit: Note that my experience is with a different piece of hardware, but I think the same buffering problems with Vorbis playback may exist between the iRiver device and the PhatBox/Music Keg. Didn't want to mislead anyone.

Doesn't sound very promising for Vorbis, unfortunately.

Nonsense ! Use BeSweet or headAC3he to convert your AC3 stuff to Vorbis and your probs are solved, as both use ssrc.dll from DSPguru in 32 FP resolution for smplaing rate conversion, and you can set SSRC so it will dither the 16 Bit output just fine .....

The reason Ogg has problems at higher bitrates is (I believe) due to the processing power needed to decode the files rather than the physical size of the file.

That is why the physically larger CD audio and FLAC files don't have these problems; CD Audio doesn't need decoded and FLAC is designed from the beginning to be very efficient in the decoding stage.

HJW > Use BeSweet...

I do. In a forum (such as this) that deals with absolutes, "hacks" such as resampling are frowned upon (even if necessary). Worse yet, my mother would kill me if she ever found out I downsampled a tune. She has a history of violence and is a vorbis ultra fanatic. In a scene reminiscent of "Casino" (the blackjack cheats that signaled each other), when I told her that I had transcoded the ogg she was listening to, she had my brother hold my right hand down, yelling "YOU'LL NEVER TRANSCODE ANY SONG AGAIN...NOT WHILE I'M STILL LIVIN!!" She was about to strike my hand with a 5 pound maul before I yelled out "from flac mom...FROM FLAC!!" That cooled her jets. She knew FLAC was a lossless format, and was an acceptable transcode. I had lied though...the source file was a high quality mp3. The day continued without further incident.

xen-uno

LOL!

[OT]
You should test the limits, if you're feeling lucky that is. Transcode her very favorite song from lame --alt-preset insane --> mppenc --braindead --> oggenc -q 6, and see if she notices. (Oh yeah...and wear steel gauntlets that day...)
[/OT]

huh guys, 64Mbit = 8Mbyte