What is mean by lossless audio compression?
 

lossless relative to what?

how can we call that lossless if it is only lossless relative to CD?

actually a dont know what lossless means here...can someone tell me what it means?

lossless in the sense that no data is simply discarded for the sake of a smaller file, mp3, ogg and most popular formats that get such impressive compression employ this technique, and are, as such, lossy.

compare this to image formats, imagine an image that is uncompressed and HUGE, compress it as a jpeg which is lossy and compare it to the same image stored in the png format which is completely lossless. the jpeg is significantly smaller, but full of artifacts caused by the discarding of data by the algorithm, whereas the png will be a pixel per pixel perfect copy of the original image, but doesn't compress nearly as well, despite the compression algorithm being very much newer.

the techniques involved between audio, images and video are very different, but the principle is exactly the same, lossless compression produces output identical to the original input.
lossy compression produces output which sounds or looks like the original, but depending on how it was compressed (64 kbit mp3 vs 192 kbit mp3) there may have been great sacrifices in quality, but relatively tiny files.

sorry...but i think you misunderstand my question...

i know lossless mean data is not discard during compreesion...

but my question is, the lossless codec is lossless compare/relative to what? the CD? the real world enviroment? or? just lossless relative to the file it compress?

it is lossless compared to the file it compresses.
this is quite obvious.
if you tar a file and then untar, is the resulting file similar to a) the original tarred file or b) windows kernel...
if b, then i´d say tar would be quite broken. lucky us that is not the way tar works.
lossless means, that when the file is uncompressed, the original file is reproduced.

it most definetly cannot compare to real world analog data, since it must be digital in the first place to be compressed losslessly.
if you take a picture with digital camera, you cannot zoom it to microscopical level because there is not enough resolution in the first place (since it is digital information). only FBI can do that ;)

It is relative to the original, uncompressed, source of the digital work of art.

As a really simple example, if you start with:
the compress and uncompress it and wind up with:
then it was a lossy compression. But if you get back
then it was not lossy.

Think of it as the difference between folding a sheet of paper in half, and ripping it in two and throwing one half away :)

ZIP is lossless. GZIP is lossless. BZIP2 is lossless. If you compress data and uncompress it, you get the EXACT same input. We're not talking about audio quality at all, but the technical definitions. That is, the algorithms have a perfect inverse.

think of your data as a piece of paper. you want to compress it (lets say you need it to fit in a box). You have two options: lossy and lossless.
lossy:
you cut the paper in half and throw one half away. Now it fits in the box.
lossless:
you fold the paper in half. Now it fits in the box
(thanks to click online for that... )

titanium_geek

here is another one ...

for lossless compression :

original uncompress = compressed
[1133366662288881111111] = [113336*4228*41*7]

but when during replaying of the compressed file , those duplicated "bits" will
expand themselves according to the "tags" , in this case they are " *(number of times) " , so in a sense those duplicated "bits" are not lost and the total size is slightly smaller ....

for lossy compression :

they throw away those "bits" that can't be heard or can't be seen with human ears or eyes , they are negligible and insignificant for human organs ...

but when during replaying of the compressed file , those discarded "bits" are lost forever , but the compression ratio can be very impressive ....

many times , lossy compressed are for smaller multimedia devices , in terms of storage space and $10 earphones/speakers that can't reproduce those qualities coming out from good old full-blown multimedia systems ...




PS :: oh boy .... knowing the answer first then cracking up riddles is fun .....

ok.....i did say i know what lossless mean, but seemed no one hear me:D


here is a thing i really dont know ( if someone still have patience to answer me)

Windows said 32-bit color is true color, how true it is?

By "true color" they mean reproduction so accurate that the human eye can not distinguish any more colors. In other words, if they had a palette twice as large (32 MILLION colors) the human eye could not distinguish between neighboring colors.

hmm ... pure ASCII words are terribly lossy , so hearing and sensing discaded ...


32-bit is very good with shades of opacity , representation of color transparencies are somewhat more faithfull , sensational see-through all the way to the background , much more softer and subtle ...

16-bit(actually is 15-bit but with one "extra" bit for alpha i guess) , most of the times you will only notice the shortcomings when seeing subtle transparencies / backgrounds , colors are harder and fake , maybe except for green color i guess ...

actually the term [ true color ] is just a nick-name , it is nothing "technical" , seeing [true color] in a sense similar to when we are seeing see-through transparencies and shadows in real life , much more softer and subtle ....




PS ::
oh boy ... donno how to crack up riddle about alpha colors and hardware ...