QUOTE(ryran @ Jun 27 2006, 11:27)

to ghostrider, the original poster:
I know a few others have said it, but I STRONGLY encourage you to spend a little time (even if it's just a few minutes) doing some ABX testing for yourself. I used to encode all of my stuff in flac.... Honestly though, for a collection the size of yours, I think all this talk of lossless is just silly. mp3 will always be supported--a dozen years from now, there will be a way to play mp3 files. If you can't tell a difference between the mp3s and the originals now, you won't be able to then so wtf--who cares about transcoding?
I'm not sure a
few minutes of ABX should give people any confidence. In 1998, I was naive and thought that 160kbps was good enough when I compared it on a few songs. But weeks later, I noticed that Talking Heads "Burning Down the House" and Swing Out Sister "Breakout" had a very noticeable and annoying swooshing effect at 160kbps. (There are dozens of songs in my library that make life difficult for MP3 encoders but I can't remember them all off the top of my head.)
I'm not an audiophile maniac but I can still tell the difference between 224kbps and the original WAV. Not on all music but definitely some songs with intricate cymbal sounds. And I can hear the difference listening on a laptop with a crappy headphone jack. It doesn't take golden ears and $100,000 speakers to notice artifacts. I haven't done extensive ABX listening tests with 320kbps. It's not a priority because I've already decided to save to WAVs. It's just a different philosophical approach: you can waste days ABX at various kbps rates trying to gain confidence in the right bit rate for your ears----or you can just skip all that and go lossless.
QUOTE
And YES hard drives are always getting cheaper and space is always increasing. If it's unnecessary though, it's just simply wasteful. I don't actually care to fight one way or the other, but to me, some people's reasoning about the whole lossless thing is flawed: essentially, it seems the idea is to take up more space now, in order to take up less space someday in the future (store lossless now so that you can take advantage of better lossy codecs "on the horizon" as someone said). Is that not just silly?? Wow.
You are correct---it is silly!
But you misinterpreted my post. The idea isn't to geek around forever looking for the latest encoder format! The key idea driving the need for lossless is to never touch your CDs ever again regardless of what
new thing comes in the future. The "new thing" could be any thing. Don't restrict your mind into thinking that a "new thing" is simply a new file format. A "new thing" may be a piece of hardware (Squeezebox --> SONOS). Also, a "new thing" could be software utilities that have nothing to do with encoding (Daemon Tools, etc).
I also want to make another key terminology distinction... when I recommend "lossless", I'm really recommending the full CD image single WAV + CUE. This approach gets you an "archive" qualiity backup of your CD. I know others rip "lossless" by taking a CD with 10 songs and creating 10 separate WAV files.
Case study #1: I first encoded individual CD tracks (not whole CD images) in to MP3 in 1998. At the time, the best I could do based on disk costs was 160kbps using AudioActive encoder. I physically loaded/unloaded about 400 CDs to extract my favorite 1000 songs and encode them to MP3. 1998 was the days of Pentium II 450mhz so the encoding times took forever; it was an all around monumental hassle of a project. In 2004, I learned that LAME is much better than AudioActive and I also decide that since disk space is much cheaper, I'll encode to 224kbps. The first depressing thing that hit me was that I would have to physically handle those
same 400 CDs again plus the hundreds of new ones since 1998.
With the 1998 episode still fresh in my mind, I decided to be a little smarter about this. Just save the whole damn CD to an image file! So that's what I did with my entire collection. I then created my 2nd incarnation of MP3s (at 224kpbs) from my CD image files instead of the CDs themselves.
Case study #2: I first had a Squeezebox V3. With that hardware, I was able to play MP3 files gaplessly (because Slim Devides intelligently reads CUE sheets). I later got into a SONOS box. However, the only way for SONOS to play gaplessly was with WAV or FLAC. So my entire MP3 library was basically unusable on the fancy SONOS box. It wasn't a big deal because of the CD lossless archive investment I made in Case Study #1, I simply ran another batch process to covert everything to FLACs. Again, I did not have to
touch my CDs to make this conversion. Sure, all hardware universally understands MP3. Sure, the SONOS also "understands" MP3 but it's not optimal. For SONOS, the FLAC format works best. It's hard to predict the weird idiosyncracies of future devices.
Case study #3: the CD image files lets me use Daemon Tools to mount them as virtual "CDs". I can then see them in itunes or Windows Media Player with metadata like titlles and album cover art. Also, various other music library utilities interface with freedb,CDDB,etc and these will only work with CD images ---- not MP3s. Also, Foobar2000 reads them beautifually and plays them gaplessly. These are examples showing advantages of lossless having nothing to do with future lossy file formats.
Conclusion:
The choice between lossless CD archive vs lossy is not religious. It's simply a matter of balancing the investment in time/diskspace now vs flexibility in the future. I acknowledge that there are very valid reasons for skipping the lossless step and always physically handling CDs whenever you need to manipulate/recreate your digital music library. I also acknowledge that lossless is overkill for lots of people that just use their ipods for the train commute to work. People can just evalulate their lifestyle, and how extensively they use their music, to determine if lossless archiving is a sensible approach. For me, it's a no brainer. Ripping a large CD collection is a major project (whether lossless or lossy) and we can't generalize into a one-size-fits-all approach. That's why some recommendations seem "silly" to some but make perfect sense to others.