Converting MP3 to Flac good or bad idea? - Page 2

Lossless, reducing size via non audio affecting means, such as "1 100 times" is shorter than "1111111111111111111111111111111111111111111111..................."

Lossy, using psycho-acoustics to reduce the complexity of the sound with as little audible impact as possible, such as, if a very soft sound is played at the same time as a very loud sound, get rid of the softer sound.

 

Thus, lossy -> lossless, exact same sound played as with the lossy file with a sharp increase of file size. It is technically possible to do so, but in practice, it's less than useless.

not benifits from mp3-> flac

As stated about twenty other times, there are zero benefits from converting files that have already been compressed to an  uncompressed format. You will gain nothing but file size. 

 

Here's a better idea. Take a CD you have, rip a song in 192k MP3 (or whatever bitrate you listen to) and then rip it again in a lossless format like FLAC. Then using a program like Foobar, do a double blind test against the files and see if you can spot the difference. I'd say there's a 99% chance you will not be able to with any consistency. I, for one, can barely distinguish a 96k MP3 from a 128k MP3 (a result that surprised me to no end when I first tried it).

 

There are lots and lots of people out there complaining about how "horrible" 320k MP3's sound compared to FLAC, but I promise you, the vast majority would not be able to distinguish a 128k MP3 from a FLAC in a true double blind test.

 

All my iTunes music is encoded at 192k AAC. When I recently bought a non Apple smartphone I had no qualms about taking those same AAC files and downconverting to 128k MP3's so I could fit most of my 22GB AAC library on a 16GB micro SD. And I tell you: the music sounds full and accurate to me. And let me note further that my hearing extends into very high ranges. I was always able to hear computer monitors (back in the days of CRT) that had been left on in a computer lab that was otherwise shut down (even when others in the room heard no such thing).

 

There is a fetishism of high bitrate sound files on these forums that is wholly unwarranted by the actual audio perceptive abilities of the vast majority of listeners. The "sound science" forum is a great place to disabuse yourself of many of these "golden ears" myths (but you have to be humble enough to accept the results of double blind tests that you cannot pass).

I tried the blind test mp3 vs FLAC with my crappy Nano 3G and NC Denon and i still noticed a slight difference in sound.

As I don't have a High-Res Headphone yet, i would be interested to know whether then difference is more obvious on the more expensive models.

 

+1,000,000

 

The bad rep that lossy gets, even among the more informed communities, is sometimes really surprising.  I'd invite anyone to get a copy of Foobar and start ABXing FLACs against 320kbps or even 256kbps MP3s.   Just use a recent implementation of the (free) LAME encoder and DON'T do a search to find those rare and elusive "killer samples."  My prediction is that you won't be able to pass the tests beyond a coin's flip.  Ever.  The state of the art of compression is that good.  Most naysayers hear someone with authority decry MP3 as a format, or say how 192kbps MP3s are "unacceptable," but without testing this is the same as letting someone else tell you why a pricier wine should taste better, or how that expensive line conditioner will buy you better soundstage.

 

Perorming the same tests with 128kbps files, especially the newest AAC implementations, you'd shock and surprise yourself.

 

The question remains "why would you ever go lossy when space is so cheap?"   This is a good point, but practicality and saving money, even just a little of it, are perfectly reasonable ends.  I have a LOT of music.  Being able to store it ALL on my iPod is of significant value.  I could save FLACs of the original files, but those would only seem practical were I to assume my ears would get better with age, and I'd magically be able to sense minute artifacts I couldn't hear today with a gun to my head.

Ahh, the classic placebo effect.

The cup analogy is actually pretty good, I like it.

As for 128kbps, erm, no. 128k is easily distinguishable. 192 is distinguishable. 256 is distinguishable with effort. 320 is not distinguishable (all IMHO, YMMV, of course.)
Edited by revolink24 - 5/5/11 at 2:45pm

It's not a blind test if you're manually switching between files that you can see on a media player.  Further, I don't even think the Nano will play FLAC.

 

Try foobar and the ABX Comparator plugin if you really want to perform a double blind test: http://www.foobar2000.org/components/view/foo_abx.  I cannot personally discern between FLAC and V0 using any of my equipment.
 

 

1. Yes u are right, it doesn't play flac but it does play aiff.

 

2. I let other people "randomly" decide what to play. But yes, the difference was quite difficult to notice and not always consistent, but it was there.

320 vs 256, yeah thats a big difference...

Just use LAME? I'm not an expert but isn't that the best mp3 encoder? Yeah I'd take that bet in a heartbeat.

"The question remains "why would you ever go lossy when space is so cheap?"   This is a good point, but practicality and saving money, even just a little of it, are perfectly reasonable ends."

You're on a hi-fi headphone forum and you're talking about practicality? You could buy a 1/2 terabyte drive for the price of 3 CDs right now, that's quite frugal.

I have a LOT of music as well. Being able to store it all on my iPod doesn't matter to me as a lot of it I don't listen to frequently and I also have a PC.

"I could save FLACs of the original files, but those would only seem practical were I to assume my ears would get better with age, and I'd magically be able to sense minute artifacts I couldn't hear today with a gun to my head."

Please elucidate your definition of "original files".

No possible way that a lossless file of a lossy file sounds any better. Not at all.

Unless I'm mistaken this thread has gone off-topic into the generic Lossy vs. Lossless debate. I don't believe he was up-converting, I may be wrong though.