OK, I've read the FAQ. I've got LAME (and yes I think it's better than Xing, even if it is like 10 times slower). But there's another question:

Is there a really nice front end for Linux, ala Audiograbber? I'm still doing my ripping and encoding in Windows? Linux seems unable to do cdda extraction on my drive faster than 2x while Audiograbber easily handles 10x on the same drive -- mark this down to Windows' better handling of ATAPI since I didn't throw down for SCSI CD-RW (I know, shame on me).

Any suggestions appreciated.

I suggest grip (it comes with helix gnome).

http://www.nostatic.org/grip/

CDDB, integrated ripping. Full support of lame and other encoders.

Linux seems unable to do cdda extraction on my drive faster than 2x while Audiograbber easily handles 10x on the same drive -- mark this down to Windows' better handling of ATAPI

Saying "ripper X is better than ripper Y because it's faster" is pretty much the same as saying "encoder X is better than encoder Y because it's faster". I.e. you are completely ignoring the quality of the CDDA extraction. Do yourself a favour and use cdparanoia on Linux or EAC on Windows, even if they are 10x slower.

Regards,
Borislav

Sorry, but I don't think so. I've done exact compares of the output files and the cdparanoia files are the same (+/- half a dozen bits at start and end) as the Audiograbber output, only Linux can't run my drive at faster than about 2.4x without generating errors. The same drive booted under windows can run at 10x with no errors.

all the readme's and howto's say that Linux CDDA utilities prefer their CD-ROMs to be booted as IDE-SCSI emulation rather than ATAPI, but compiling a custom kernel without ATAPI support would also disable my UDMA/66 support, at least unless I was willing to hack up the source, which I'm not. Hopefully 2.4 will resolve some of these issues and in the meantime....

OK, yes, I agree with this, it just appeared from your original post that you are comparing the two setups purely based on speed.

Wishes,
Borislav

"Sorry, but I don't think so. I've done exact compares of the output files and the cdparanoia files are the same (+/- half a dozen bits at start and end) as the Audiograbber output ..."

But how does Audiograbber fare with scratched or otherwise damaged CDs? (I've never used it, so I'm not being rhetorical.) As the name might imply, cdparanoia tries to get every last bit off the CD exactly as its supposed to be on there to begin with. I've managed to use it to recover a number of what I thought to be hopelessly damaged CDs. It just tries and tries and tries ... until it gets a good read. Files ripped from CDs that sounded awful playing in the CD player sound great once cdparanoia has had its way with them. The Windoze code might be "playing tricks" with the CD device to get speed at the expense of *possible* ripping errors, too. (Following the lead of "only just good enough for the average Joe" set by Microsoft themselves...)

Also, keep in mind that this is one of those places where Windoze 90% market saturation "help" the end-user. There are (at least) 100x more people ripping and encoding under Windoze, so there are a lot more people developing the software for Windoze. Off the top of my head, I can't name half-a-dozen rippers OR encoders for Linux, but I bet there are at least three or four dozen *popular* ones for Windoze. (Of course the tools for Linux are *clearly* better... :)

I use krabber, locate it at freshmeat... a nice KDE frontend for almost everything in the wide field of CD Audio and / or DATA cd's, can use almost any cd ripper and almost any MP3 encoder, I use it with cdrecord and the xing engine on a dual-P3 550 and rip complete CD's in 15 minutes top quality...