I've read about iRiver's moves to support Vorbis on their players, and being the owner of an IMP-400, await that day with great trepidation...

However, on the thread which mentioned the iRiver Vorbis support announcement, ScorLibran mentioned that Vorbis files above a certain bitrate don't decode very well. He mentions that should actual bitrate spike over ~ 210 kbps, the decoder has problems handling them.

Is this a confirmed issue or what? If its a problem with the decoding process, has anyone taken a further look at it?

Also ( and this might sound a little trollish ), I was considering migrating my music collection to Vorbis 1.01 (when released ) from MP3 to play on both my PC / portable. However, with the above problems, is it really worth it anymore?

I maintain one copy of the files between the PC and the portable, with the music being encoded right now at APS. A similar bitrate in Vorbis would be 5.00, which would (for some samples) run into the above problems. If that should really be the case, is Vorbis support really what its made out to be?

ScorLibran mentioned that he was encoding for his Kenwood MusicKeg, which may have some problems according to what he says.

iRiver is aiming for full support up to 360kbps, so you should be safe.

kjoonlee is correct, I was only comparing my experiences with the Music Keg to the iRiver (and other) portable Vorbis players which may use integer-based decoders such as Tremor (or a variant of Tremor).

However, don't blindly trust that it will do 360kbps just because the manufacturer says it will. PhatNoise stated that their Vorbis decoder worked fine for them for files up to 320kbps, but mine suffered buffer underruns if the bitrate spiked much over ~200kbps. Recent mods to their decoder improved playback performance by enabling the adjustment of certain settings in an init file, mostly relating to buffering.

The manufacturer, for example, may have quickly tested out three easy-to-encode songs that didn't spike bitrates too highly, or do anything that might have caused a buffering problem. In my own test with the PhatNoise version of the Tremor decoder, I was able to get some songs encoded at -q 8 to playback fine, but others at -q 4.5 failed (with both high-accuracy and low-accuracy decoders).

The best bet with the iRiver (or any other of these new portable Vorbis players) is to try music of different styles, each with multiple versions encoded at different -q settings to see if/where you get skips or distortion during playback which may be caused by problems such as I ran into.


Edit: For reference. here are the results of tests I performed on the PhatNoise player (and its Vorbis decoder) to give some idea of where certain tracks would fail while others played fine when encoded with various -q settings.

Don't worry Daybreak,iRiver will release update fireware for iMP400 at the end of 2003.It will support up to 360kps and 44.1 khz.There is no problem in decoding!

And how would you know that? iRiver has said they are working on it, but I hear that's been the case for over a year or more already. Do you have some verified info to back up that claim? I'm about to buy an iFP-3xx player, and Ogg suppprt would be very important for me. Like many people, I am hoping to see this new firmware released, if they can get it working. But at the moment, I'm not holding my breath. I hope it doesn't turn out to be purely vaporware...

The only player I know of (avaliable now) with full support for Ogg is the Rio Karma.

Interesting. Have you (or any other HA member) had the chance to test it yet ?

I'd like to know if it can play Ogg Vorbis files AND MP3 files gaplessly.

Why did I mention Ogg Vorbis ? Because having Ogg Vorbis support does not necessarily mean it will play them gaplessly. Look at the Neuros: back in July/August, it still had no gapless playback support for Ogg Vorbis, and I haven't seen it mentioned on the Neuros forum ever since. I get the impression the firmware developers don't think it's important enough.

Why did I mention MP3 ? Because of the statements one of the Rio developers made on Slashdot. Look here. I think this comment may indicate gapless playback support for MP3s as well. If this is true, the Rio Karma is quite a remarkable portable player.

Well, I think I can partially answer my own question.

I discovered Riovolution, which is a great resource for Rio portable players, including the Karma.

According to this thread, the Karma has gapless playback support for MP3s, but not for Ogg Vorbis with the currently available firmware.



Quite strange, isn't it ?

Anyway, the Rio folks also mention that the Ogg Vorbis issue is now resolved "internally", so the fix should be available with the next firmware update.

Truth to tell, I'm pretty happy with MP3 as it stands. While the level of compression could be higher, using APS, I can't hear a single difference...

Why I'm looking forward to Vorbis is largely due to two things - gapless and proper tagging support..


Hopefully iRiver can get their act together, and put out a firmware that does both MP3 and gapless Vorbis w/o any issues!

Ah...that's a good point. For a brief time with my player as well, I had "gappy" playback even with Vorbis at first until I had access to some of the buffer settings to enable gapless playback (either pre-buffer or EOF_buffer, I can't remember which).

So I was wrong saying that gapless playback can be "automatic" for certain codecs on any player, as there are other considerations. Thanks for the correction.