actually, i cannot understand well enough the use of the lossless codecs.

for example, the album i encoded has little difference from the original one in matter of size. to my hearing it had none difference from the one i have in mp3.

also, you cannot back up flac-ed albums easily. you would need a lot of discs to do that.

anyway, i would like to test again some albums. the thing is that the flac installer makes a frontend for the encoder. what i want is to encode to flac through eac or cdex. i copied the flac.exe into eac's folder but i cannot find it among the other encoders. what am i doing wrong?

EAC:
Tell each to use a "User defined codec" and set the file extension to .flac. Then enter in the location of flac.exe and set your command line paramaters (--best for example). Click OK and then proceed to rip like you would MP3.

And one of the main reasons why people use FLAC (or any other lossless encoder) is so they can have a lossless backup of they're CDs. If you were to lose an extremely rare CD, but you were smart enough to make a FLAC backup on the computer, you can simply burn the FLACs. It'll sound exactly the same as your old disc you just lost.

Also, FLAC can be transcoded to other formats. Keep a FLAC rip on your computer, and then you can transcode those FLACs into MP3s for your portable MP3 player whilst still having a perfect lossless backup on the computer.

Since it's a lossless codec, you won't waste a single bit of your music. Think about it this way: when you encode a cd with flac, you get a smaller file which is easier to handle, playable directly by many programs and some portable devices, it's got no loss of quality and if you happen to lose the original cd you still have a 1:1 copy of it on you hd. I'm backing up my 400 cds collection and I would have chosen flac if I had a bigger hd. But because of my small hd, I've chosen MPC which is stil the best lossy codec (in my opinion). So, if you have enough space, go for flac (or the lossless codec you prefer), no doubt about it.

What do you mean by "little difference?" THe amount of compression depends on the music. I typically get close to 50% reduction. I just checked a 2 CD concert and the overall reduction was 2.04:1, with 112 minutes in 492 MB. So 2 CD's fit in about 25 cents worth of HD space.


The advantage of lossless in general over lossy (like mp3) is that it is a compressed version (and expandable to) the original data on the CD.

There is also the cost advantage of free formats (like flac, vorbis, ...) over mp3 which has license fees for players, encoders, streaming, ...

That's why they started to invent things like Blu-Ray
DVD also seem to be quite large for storing mp3!!!

Why use lossless?
1. Simple, because it is lossless
2. Disc space getting larger cheaper
3. People have no idea what to occupy their large storage capacity with something, so they use lossless audio

Maybe the album is very loud. Do you like recently mastered heavy metal music ? Then it probably is. ~80% new file size is quite common as opposed to ~40% for classical music (numbers are very rough).

It can be smart to at least back up Copy Controlled CD's to lossless. They are much more prone to become unreadable do to scratches than regular cd's

I use lossless because I can store it without missing the hard drive space. Disc space is relatively cheap nowadays.

I just backed up 1400 CDs in FLAC to DVD-R. It took about 120 discs. Cost: roughly $60 (man, that's cheap insurance). That's 1400 *titles*. The actual number of discs was much higher.

Considering the cost of storage in 2005, I would change the title of your post to read: "Why use lossy?"

(ok, ok, I have an iPod too. So mp3 isn't totally dead.)

exactely that's why I have flac and mp3 (for the ipod).
and btw you can hear the difference strongly on good equipement.

You'd better be careful with such statements as long as you don't provide ABX results. This rather highly depends on the bitrate, codec used, settings and the listener's effort than on the equipment.

agreed, sorry...

but I thought, that with an bad stereo equipment you might not even hear any difference between a highly compressed mp3 and flac or even the original cd.
before I started to use an higher end soundcard, even lossless sounds were not that great out of the PC.

Well, on my cheap mono clockradio it's indeed very hard to hear any data reduction artefacts or clipping/dynamics compression, while on the hifi system I'm pretty sure to notice mp3 artefacts on FM radio sometimes.
What I'm trying to say is: Usually artefacts are so bad that you don't need good equipment. How bad does it have to be to not notice them ?

good question.....
actually I tested the difference between MP3 Lame VBR and apples lossless format on an iPod and couldn't hear the difference. But I still like the iPod. Might also be that apples AAC is not really a lossless one. But now I'm pulling into the wrong direction on this topic.

AAC = lossy ... ALAC = lossless

Thx Cosmo, so I used the apple lossless encoder.

Then guessing that ALAC "might not really be lossless" is the wrong conclusion. It's more likely that the LAME mp3 samples you compared it to are within your transparency threshold, in the given situation of that particular hardware.

edit:spelling

ok, I did see some discussions on that in the past. So can I conclude than, that ALAC is a lossless format in a common definition and not only by Apple ?

Yeah! If you do a md5 checksum on a wav and then convert it to alac and back to wav, it will be identical.

I like lossless codecs for a few reasons. First of all, they are very fast. If you have a good drive it is possible to get compression about as fast as a fast burst mode rip.
This can turn making a jukebox with 200 CD's into an easier project, assuming you have a big drive. The output might not be bit perfect, but with the right hardware there will be no audible clicks or pops. The other reason is it is possible to transcode from lossless with the same of better results than a fresh rip. This is very useful for those who are members of the codec of the month club, and especially useful for those who dedicate a lot of time to getting a bit perfect rip, but know they will be encoding with more than one lossy codec.

Is the question "why lossless instead of lossy?" or "why lossless instead of uncompressed?"

I think the first has been answered, but as to the second question, I like lossless vs. WAV files because the tagging features make it much easier to transcode, etc.

i guess you're right. this is a way of fulling my 120GB hd. i am going to organize much better my space after the summer-format.

I use FLAC to encode losless concerts that I've tapes or archived on DAT or hard disk, mp3's (and other lossy formats) can get a cascading or swirling effect that clearly audible on higher end stereo systems. Also you can archive 24/96 kHz files in FLAC as well so that's a total bonus.

Spott

Yes FLAC is great for archives. But please do not spread this idea that any lossy format can be distinguished on higher end stereo systems. Be aware, we have TOS #8:

as I did put a statement before in the same direction, and don't want to violate any forum rules: did anybody run blind tests with MP3 and Flac ?

What do you mean? For what purpose?

well, did some reading in parallel also. Seems like CT in Germany did some blind tests also with lossless formats, and the listeners were not able to identify compression versus lossless. My thinking was before that lossless formats are clearly to identify versus comprossed formats. That's why I asked whether there were blind tests executed on that.

transparency is the term we use when an individual can't distinguish between the compressed and the source, or in this case, the lossy and the lossless. most lossy compressors will achieve transparency for most individuals at some point. lame preset standard for instance; most people cant tell the difference between it and the uncompressed source. i am one of those people. i can't tell lame standard from lossless, but lame preset medium, i can easily distinguish (or, more technically, abx). the point of using lossless is not becasue it is the only way to achiece transparency, because if that was your goal, you could use a lossy compressor. i use lossless for peace of mind and so i can convert my music to a new lossy compressor if a better one for my needs comes out, or if my needs change.

Yes, there are number of tests done even here that compare MP3 and the original track (which is the same as FLAC)
http://www.rjamorim.com/test/

You are aware that listening to a lossless encoded track is the same thing as listening to the same track from the CD, right?

Edit: link

@ VCSkier
thx for clarifying... well actually I use flac for my home pc system and than I convert to mp3 lame vbr for usage at the ipod. Guess that I need to more hearing tests at my system between flac and mp3.
But as harddisks are getting cheaper and cheaper, lossless would be on the safe side anyway, as you pointed out. The is the clear advantage in my thinking: the orginal data is uncompressed, taggable there and can be transformed to any other format.

@Jido
yes I am (except that you don't hear the diskdrive as loud with flac :-) ) Thx for the link, I will check it out.

The listeners were not able to hear the difference because the test-tracks were a joke... very simple pop music and techno, and one very silent classical track (if i remenber correctly).
That wasn't a hard work for the encoders.

lossy ranks from like 32 kbit he-aac to 320 kbit lame insane, so a bit more info is required, also everything pretty much depends on ppl, the original abx page clearly states that listeners need prior training, otherwise every cow is a black one:

I think classical music often has less variety of frequencies because of less instruments and also smaller stereo fields for the same reason. That's why with classical it's more difficult to detect lossy compression artifacts.
I also read about this; it was in German Stiftung Warentest (It is a reliable test magazine without advertisement.) in a test of portable MD recorders.

Its not just about hearing a difference, but the fact that I can so easily transcode to whatever format I want without additional loss

let me try to get some more details on that CT test, I was more than surprized also about the results... may be next week, as I have to leave for ride to Frankfurt now.

I'm sure guruboolez would be able to confirm or deny this since he's a big fan of the classical stuff

for the german readers http://www.heise.de/ct/00/06/092/

Classical music has fewer instruments than what?

Quoted from the ct-link: (please don't let me translate this...)


They write about their conclusion that there is no difference in difficulty of detecting lossy artefacts with classical music vs. other.

Pop. If it isn't an orchestra recording, it maybe e.g. only-piano music. Other music styles use electrical instruments or computer programming. It's of course easier to have more 'virtual' instruments. The great artistic merit of classical music lies in the variation over time.

Well I am not into classical, but I would argue that it is more easy to detect lossy compression with classical music becuase anyone who has some experience with classical music tend to know how violin, piano, etc sounds in real life. Pop rock music is full of distorions therefore you may have to fall back to ABX more frequently because you just cannot tell why it sounds bad. On the top of that I am yet to encounter a classical CD that is clipped or badly mastered. One of the selling point in classical music is clarity and good mastering. So again I expect that lossy codec would reveal itself more readily with classical music.

Triza

Fair enough, but a lot of classical music is orchestral/choral (the vast majority of the stuff I listen to is, certainly) and even some chamber music is 7+ instruments.

My first post here...

I have a 400 gb hd and am just about done ripping and flac'ing all of my cd's to it. I have two motives for using a lossless compressor:

1 - I've tried myself out in double blind tests and realized that I'm neurotic enough to hear an mp3. So, even though the artifacts I noticed were not unpleasant (and not noticable except in side-by-side comparisons), my neurosys makes me want it "right".

2 - Ok, the Real Reason, alluded to by others here: This is my backup of my CD's. A compressed file is not exact. What if I later find some music where the artifacts are objectionable? I'd have to dig out the CD and try to re-encode it. This way, I always have an exact copy of the CD at my fingertips.

Why use a compressor at all?

Well, I don't regularly get 50% like one poster here gets - I think I average about 30% compression. But that's roughly 100gb in a 400 gb collection - nothing to sneeze at.

Like another poster, I convert to mp3 "on the fly" to my player.

Incidentally, my primary motivation for converting to hd was to have an easily browsable jukebox at my fingertips. I use a Squeezebox for playback at home and am very happy with it.

Mark

With classical music it's very easy to reach 2:1 lossless compression (usually even more for chamber music), so it's almost a crime not to use lossless.

So if you use the winamp FLAC or EAC w/FLAC, does it make any difference in which program you use to rip?

EAC is secure. Winamp is not. Filesizes will pretty much be exactly the same.

What do you mean by "secure"...I'm a bit of a n00b.

You will be told whether the rip is bit identical (100% exact) to the CD or not.

I see thanks. Any good resources (links) for setting up EAC for FLAC rips?

1.Open 'External Compression' tab.
2.Check 'Use external program for compression'
3.Change 'Parameter passing scheme' to 'User Defined Encoder'
4.Set file extension to .flac
5.Locate 'flac.exe'
6.Use command line -T "artist=%a" -T "title=%t" -T "album=%g" -T "date=%y" -T "tracknumber=%n" -T "genre=%m" %s

-Martin.

Edit:
Check 'Delete WAV after compression'
Uncheck 'Add ID3 tag'

examples here:
http://www.hydrogenaudio.org/forums/index....23867&hl=suffer
http://www.hydrogenaudio.org/forums/index....&hl=miles+davis.
It's a pity.

Hm, OK...
The huge variation in loudness may still be a disadvantage (when detecting artefacts).