Thoughts on the New Crop of Ogg Aware Players?
Steve Andre' asks: " Given the approaching season, I'm wondering if many have used and have opinions about the new Ogg Vorbis capable portable players out there. What I'd like to find is at least a CD/MP3/Ogg capable player which sounds good and doesn't do 'odd' things. What's it like out there? Can I finally roast my Ogg files and take them with me for a walk?"
How about a head unit for my truck's sound system that plays Ogg? I have yet to find one, anyone have a link?
24 beers in a case, 24 hours in a day. Coincidence? I think not!
The decoder is included with winamp (at least the standard and full versions). They might have added it to the lite version too recently thereby making the "plugin" pointless? (Peter's plugin is out there, if you really want it)
Besides, everyone who was using the Winamp "classic" is using foobar2000 now anyhow :-)
Belief is the currency of delusion.
They're kind of a funny company. They're very open source / Linux friendly, and seem to be very into their user community. It's basically like the kind of coolness I'd want to exhibit if I started a tech company.
That being said, I'm a little concerned about their solvency, and their technology is getting old.
BUT, the good news is they're having a clearance sale right now. I just ordered a 128 MB Neuros for $99, and I'm really looking forward to getting it.
From what I can tell, their Ogg support is perfectly usable, and there are two open source programs on Sourceforge for managing its songs. You can even flash its BIOS using a Linux host - you don't need to boot up into Windows to do it.
So my basic take: For $99, it was worth taking a slightly less polished product, so that I could support an OSS / Linux-friendly company.
However (and maybe this is because the Vorbis files emphasize the high frequencies; I'm not sure) the MP3s sound "flat", somehow.
My understanding of the sound quality difference between MP3 and Ogg Vorbis is that MP3 uses a full Fourier transform (sine, cosine and constant) on the audio, while Ogg Vorbis uses wavelets and does a cosine transform only.
Discontinuities between blocks sent to the sine transform would cause the Gibbs effect; these would be heard as a garbled low-amplitude reverberation of the high frequency components and transients in the audio. This is consistent with the effects of low bitrate compression; at higher bitrates, there would presumably be more terms used in both the sine and cosine transforms, so the amplitude of the compression artifacts would become smaller and therefore inaudible.
Cosine transforms, on the other hand, don't have problems with discontinuities, so there'd be an immediate increase in sound quality, at a given bitrate. Transients (attack on cymbals or the rattle of the chain across the membrane of a snare drum, for example) would be handled by wavelet functions - there's probably some sort of mechanism in the code which sees the sharp attack or decay as fast risetimes or falltimes, ignores processing it by cosine transform, and uses wavelets instead.
But I don't know for sure. For one thing, I am *not* a programmer. I can make "Hello, World" and compile my own kernel, but you really don't want me poring over the source for libvorbis.
Fire and Meat. Yummy.
If you own a SonyEricsson P800 or soon P900, try Leif Wilden's excellent Oggplay ( http://www.geocities.com/p800tools/ ) It's also available for Nokia's Series60 phones (7650, 3650, N-Gage) though as far as I can tell it only supports 16kHz playback for now. It's an excellent choice for the upcoming Siemens SX1 though where it supports the full 44.1kHz. And it's open source as well ( http://sourceforge.net/projects/symbianoggplay/ )
Given the recent fuss over the iPod's $200 battery replacement charge, this should be worth checking out before purchase.