[G.Trenchev]

I use lossless,because I respect & adore music,and compressing it just to save hard drive space seems disrespectful.

 

[xnor]

Here's a simple analogy with images:

A BMP image (the original) with 1920x1080 pixels is about 8 MB big.
Save it to PNG (= lossless) with highest compression level and the resulting file is only a bit over 3 MB big.
Save it to JPG with 85% quality (= lossy) and the result is a 1/2 MB small file.

As long as you don't recompress the JPG or zoom in or process it it's fine to look at, you probably won't notice the compression artifacts (though this depends on the image and quality settings).
The PNG on the other hand does not only look identical to the original BMP file, you can also convert it back into a BMP file - it'll be the same as the original.

 

[Gallictic]

Awesome guys, thanks for the responses.
 
Appreciate it.

 

[ZeNmAc]

Many knowledgeable posts in this thread.  Clears a lot up.  I would've replied but everyone put it better than I could've lol.
 
Practically speaking, I rip to FLAC which is what I store on my external hard drive and what I play on my computer.  Then I convert the FLACs to 256Mbps MP3's and put those on my ipod.  That way all my CD's are backed up and I can save space on my ipod since I won't be able to tell the difference listening to the ipod.
 
Quote:

Here's a simple analogy with images:

A BMP image (the original) with 1920x1080 pixels is about 8 MB big.
Save it to PNG (= lossless) with highest compression level and the resulting file is only a bit over 3 MB big.
Save it to JPG with 85% quality (= lossy) and the result is a 1/2 MB small file.

As long as you don't recompress the JPG or zoom in or process it it's fine to look at, you probably won't notice the compression artifacts (though this depends on the image and quality settings).
The PNG on the other hand does not only look identical to the original BMP file, you can also convert it back into a BMP file - it'll be the same as the original.

Great post.  I didn't know that about the image files.  Good to know.
 
 

 

[Sumpfkraut]

Being from the Bay Area, the distinction between earthquake and lifeboat preparedness may be lost on me.

That's why I put the sidenote there.

 

[kalston]

I can't tell the difference between 320/VBR V0 mp3s and FLAC (haven't done that much blind testing though) but I like the idea of not risking to hear any thing that should not be there in the music I truly adore. Even though storage is cheap, I don't have all my music in FLAC, only the music I care the most about, the rest is just VBR V0. I need storage for my blu rays and games too, so I can't just have all my music in lossless for now. I'll probably get myself some home server one day, but that's not possible right now and until then I will keep choosing carefully which files I want to have in lossless.

Heck, I even have hard time telling the difference between LAME's 128s and 320s (again, I haven't done that much ABX so maybe I actually could). Have only done it on some website though, and I have no idea which encoder they used, and I didn't know the songs at all either, which can't help I guess.

 

[danroche]

For all practical purposes, beyond 192kbps my ability to discern between a source file and its compressed counterpart drops into coin-flip territory.  I use LAME VBR -v2 for most of my stuff, giving me somewhere in the low-200s, and to date I have not wanted for more.
 
I do know of (and possess some) "killer" samples where I could catch artifacting even at 320kbps, but these examples are so incredibly rare, and the artifacts so incredibly acute, that I've decided it's not worth worrying about them. If I can play an old and scratched vinyl record and enjoy the music, the same can be said about a compressed music file, provided the encoder and bitrate are strong enough.
 
If, 20 years from now, humankind abandons all lossy media and goes to FLAC or whatever lossless format takes hold, I know that I can always upsample by compressed library to lossless and keep going. (Transcoding only applies to lossy-lossy and not lossy-lossness AFAIK.) Given, I won't be recreating the first generation copy, but as established above I'll still be working with music that's perfectly good for my purposes.
 
In the meantime I'll save hard drive space and maintain battery life on my portable player, which will need to do fewer disc reads to play a song.

 

[Skylab]

Quote:

If, 20 years from now, humankind abandons all lossy media and goes to FLAC or whatever lossless format takes hold, I know that I can always upsample by compressed library to lossless and keep going. (Transcoding only applies to lossy-lossy and not lossy-lossness AFAIK.) Given, I won't be recreating the first generation copy, but as established above I'll still be working with music that's perfectly good for my purposes.
 

 
There would be NO point in that.  You never get back what's lost.  This would only serve to make your files bigger with no sonic benefit.  The only reason would be literally if there was going to be no way to decode the "old" files, which isn't likely.
 
 

 

[danroche]

Quote:

 
There would be NO point in that.  You never get back what's lost.  This would only serve to make your files bigger with no sonic benefit.  The only reason would be literally if there was going to be no way to decode the "old" files, which isn't likely.
 
 

Yes, hence my use of the sentence "Given, I won't be recreating the first generation copy, but as established above I'll still be working with music that's perfectly good for my purposes."  This would only be done in the event, as you say, there was no way to "decode" the old files, hence my use of the words "If, 20 years from now, humankind abandons all lossy media."  
 
This is extremely unlikely, but is the only possible drawback I see to me continuing to use MP3 -v2. If I were a mastering engineer, I might go for the theoretical maximum of quality at all stages, but as a listener, I don't see the need. When EMI wants to secure the Beatles red book recordings for another Christmas compilation, I assume they'll be looking for them somewhere else.
 

 

[Skylab]

The other drawback, again, to lossy coding, sound quality issues aside, is that there WILL be better codecs in the future.  This is certain.  If you rip CD's to lossless, then transcode to lossy, you can always re-transcode to a newer/better codec later, versus the much harder task of re-ripping.

 

[xnor]

Provided that CD drives still exist in 20 years, to be able to re-rip your collection.

 

[Mkubota1]

Quote:

 
There would be NO point in that.  You never get back what's lost.  This would only serve to make your files bigger with no sonic benefit.  The only reason would be literally if there was going to be no way to decode the "old" files, which isn't likely.

Yes- that would be like touching up and old painting.  You could come close, but you'd never get what was once there.
 
For me, 99% of the time I can't hear a difference between LAME -v2 and WAV.  So on my portable devices I keep mostly lossy files.  With the variety of music I listen to and my desire to keep albums intact I simply don't have the space for lossless on everything.  Reasons I can see wanting lossless back-ups even if you can't hear the difference now would be: 
0) The aforementioned cheap storage answer- why not?
1) You are just getting into critical listening and maybe you might hear differently in the future.
2) You are building up your system and future improvements might be more source revealing.
3) Like me you are just a little OCD and even though you can't normally tell them apart you really want those last few bits of data.
4) You're tired of people asking why you listen to lossy files.

 

[plonter]

I personally have all my cd library ripped to flac on my pc at home which i use with my headroom rig for critical listening..and I converted all my library to v0 mp3 so i can put it all on my 32 gb cowon j3.  v0 mp3 are great solution for mobile daps for getting hi fi sound and saving a ton of space..this is why it was invented.   but in my superior home rig i listen only to lossless with my computer or my marantz cd rom.

 

[svyr]

because we got scared off by lower bitrate lossy codecs with bad encoders from the early days of mp3s as well as re-compressed rips.
Storage is and has been very cheap the past few years, letting you keep 1000s of lossless rips on 1 hdd...
Many say they can tell 320kbps from lossless using the ABX plugin and some argue mp3 is oriented towards people with bad/cheap/multimedia equipment (mixing up psychoacoustics and artifacts to some extent )
Finally, if it comes to the point where you need to re-encode your collection because of a new hot format, lossless gives you the option to (and the option to verify it decodes back to check for bugs in the codec). With lossy, you'd be getting artifacts from both the old lossy formats and the new one.

aaand 'Just in case' (better gear in the future, developing golden ears , etc)

 

[Seamless Sounds]

If you want to see what was lost in the process try this. Encode your FLAC file to mp3. In Audacity or any audio editing program, place the original FLAC in the first track, and the mp3 in the second track. Invert the mp3 waveform. Align the tracks so they phase out. All that's left of the audio now is the stuff that was removed. Save the mix and analyze it in a frequency spectrum analyzer. You'll more or less get an idea what parts were removed from the lossless file.
 
I usually have both the original song and the instrumental track in lossless for this purpose (remixing). It's to extract the vocals out of a song. 320kbps mp3s is pretty acceptable (hit or miss depending on encoder used). 256kbps is not.