to MP3 or not to MP3?

[washline]

Thank you Joda. I think most people can probably hear the differences if they just concentrate a bit more. I was able to make the distinctions quite easily even on my computer speakers of Fremer's samples from the other day. Even then, there was simply a lot more texture on the WAV files over and above the MP3s. It's particularly interesting that younger people settle for it, since they are able to listen much closer to the noted human limit of 20 khz than older people, whose listening range drops to about 15-16 khz as they age.

[Dan G]

...you must use WAV or AIF encoding or FLAC, ALC, or WMA Lossless. Both MP3 and AAC...

Can someone please explain to me, a bit of an idiot when it comes to technology, what FLAC is. I've downloaded some FLAC files - actually RAR, that then decompress to FLAC, and used some piece of software (called, simply, FLAC) to convert them to WAV files that can then be played on a computer. Is this using an extra step that I don't need to be doing - is there something that actually plays FLAC files as is? Maybe I'm just doing the best I can with the limitations of an old PC running WIN2000.

The reason I am asking here, to keep it relevant to this thread, is that if I am right in my understanding of them, and you can't play a FLAC files, comparing it to an MP3 seems odd. You are really comparing the WAV file to MP3, and FLAC is just another compression tool like zip or rar. Maybe I'm completely wrong...

Also, going a step further and burning WAVs onto CD (which is then converting the WAV to something else...AIFF or something like that...) - does that then give normal CD quality? I can obviously just hook my computer into home system to listen to this stuff, but if I want it in my car I need CDs (not that sound quality matters on my stock Toyota system - mp3 to CD is good enough for driving - but just curious).

[jon abbey]

I'll leave most of that to someone slightly more qualified than me, but one thing I do know for sure is that WAVs and AIFFs are identical and interchangeable.

[RFKorp]

Yeah, AIFF and WAV are both raw formats. There are differences in the way the data is organized on a bit by bit level but nothing that changes the way either sounds as long as they've got the same bit depth and sample rate.

And there are in fact programs that will just play .flac files outright.

[kangaroo?]

.flac is an audio format just like .mp3 .mp4 etc...
It just isn't as widely supported as the rest. One basic player i tend to use for most media formats is vlc:

http://www.videolan.org/vlc/

it tends to support most codecs & it's freeware

as jon & RFKorp said, .wav & .aiff are on the same level being the basic uncompessed audio formats. and yes being uncompressed means that using .wavs or .aiffs you'll get as close as it gets to the CD's quality

[Stephen]

So what I want to know is what does MP3 compression actually do to your audio? reduce the sample rate?

and FLAC is called 'lossless' compression but I don't get how you can reduce the size (in MBs) of a file and not lose some audio along the way.

[RFKorp]

and yes being uncompressed means that using .wavs or .aiffs you'll get as close as it gets to the CD's quality

in fact, being uncompressed, it means that there are .wav and .aiff files that are even better than CD quality assuming they were recorded as such and that they have never been burned as audio to and re-ripped from disc.

[snailed]

So what I want to know is what does MP3 compression actually do to your audio? reduce the sample rate?

and FLAC is called 'lossless' compression but I don't get how you can reduce the size (in MBs) of a file and not lose some audio along the way.

AFAIK, it does not reduce the sample rate but reduces data in the areas in which we do not hear as well, or where most music does not reside (ie. in the 15 - 20 khz area). This is probably why EAI and other experimental music suffer more, because it often focuses on the extremes of the audio spectrum.

It helps to think about lossless compression as somewhat like a .zip or .rar file. It uses computer processing power to reduce the filesize, but when it is decoded, you are left with the exact same data that you put into it.

[kp*]

and yes being uncompressed means that using .wavs or .aiffs you'll get as close as it gets to the CD's quality

in fact, being uncompressed, it means that there are .wav and .aiff files that are even better than CD quality assuming they were recorded as such and that they have never been burned as audio to and re-ripped from disc.

a 44.1k 16-bit wav or aiff is exactly the same quality as redbook standard cd audio. Not better, not worse. The same, the data is just in a different format and the redbook audio has the P & Q codes that control your player and it's display. The only difference will be the quality of the dacs in your computer -vs- the quality of the dacs in your cd player. The data is just in a different format. Both aiff and wav are riff file formats and are very similar.

-kp--

[RFKorp]

and yes being uncompressed means that using .wavs or .aiffs you'll get as close as it gets to the CD's quality

in fact, being uncompressed, it means that there are .wav and .aiff files that are even better than CD quality assuming they were recorded as such and that they have never been burned as audio to and re-ripped from disc.

a 44.1k 16-bit wav or aiff is exactly the same quality as redbook standard cd audio. Not better, not worse.

All I meant was that I don't mind listening to a 48k/24-bit wav or aiff once in a while

[kp*]

So what I want to know is what does MP3 compression actually do to your audio? reduce the sample rate?

and FLAC is called 'lossless' compression but I don't get how you can reduce the size (in MBs) of a file and not lose some audio along the way.

AFAIK, it does not reduce the sample rate but reduces data in the areas in which we do not hear as well, or where most music does not reside (ie. in the 15 - 20 khz area). This is probably why EAI and other experimental music suffer more, because it often focuses on the extremes of the audio spectrum.

It helps to think about lossless compression as somewhat like a .zip or .rar file. It uses computer processing power to reduce the filesize, but when it is decoded, you are left with the exact same data that you put into it.

mp3 try to reduce size by looking at what parts of the sound would be masked and removing the stuff the algorithm thinks we wont hear anyway. There are, of course, flaws in the algorithm, and it also does better on harmonic stuff than on inharmonic or noisy stuff, which is why 's' sounds and cymbals get washed out. The more you ask it to compress ( lower the bit rate) the more ruthless it will be in deleting psychoanalytically boarder line data. Kind of like your mom cleaning your room. ... so 320 throws out much less and is not very compressed... 128 is your mom losing her patients and only half looking at shit before she just throws it in the trash.

Mp3s are lossy cause you cant get the data back once it has be compressed so compressing a track and uncompressing it (to edit or further mix.. whatever) and then re compressing is akin to photo copying a photocopy of a photocopy. Shit is just lost.

When Waves and later Apple and Flac came up with lossless codecs i tested it. Run a checksum, compress, decompress, and rerun the md5 and you will see they are the same. The lossless files can reconstitute the original file exactly.

-kp

[kp*]

All I meant was that I don't mind listening to a 48k/24-bit wav or aiff once in a while

oh... i wasn't sure what you meant. in that case.. nevermind...

[kp*]

When Waves and later Apple and Flac came up with lossless codecs i tested it. Run a checksum, compress, decompress, and rerun the md5 and you will see they are the same. The lossless files can reconstitute the original file exactly.

-kp

Sometimes the md5 might not work cause even though the data is the same, the fileheader might be different..

2 strategoies for that:

1. old skool. (use what we know about phase)

take the original file. copy it. invert the phase. now take the reconstituted file and put that in your daw... and sum the mix, or just play it.

If the file is EXACT sample for sample, we now have 2 identical files that are perfectly revese in phase and a mix of that should yield perfect silence since every sample that has a positive value now is combined with a negitive of the same value. (+5 + -5 = 0)... the bounce should be all zeros.

2. Save the original sample values as text... do the same with the 2nd file.. and do a %diff on them... should get nothing....

booya~

[washline]

Can someone please explain to me, a bit of an idiot when it comes to technology, what FLAC is.

Dan G, some more details on FLAC here: http://en.wikipedia.org/wiki/FLAC

As others have noted, it's a lossless compression format, like Apple Lossless with iTunes. WAV and AIFF are uncompressed formats.

So what I want to know is what does MP3 compression actually do to your audio? reduce the sample rate?

and FLAC is called 'lossless' compression but I don't get how you can reduce the size (in MBs) of a file and not lose some audio along the way.

I would encourage you to read the linked essay from the first page where a complete study is done on what MP3 compression does to the music. The claims are supported with detailed measurements.

The only difference will be the quality of the dacs in your computer -vs- the quality of the dacs in your cd player.

On your CD player, the quality of the transport mechanism comes into play as well. Once the files are ripped, the transport becomes the spinning hard drive, which can at times have negative audible effects on the sound quality of the music.

128 is your mom losing her patients and only half looking at shit before she just throws it in the trash.

I didn't know your mother was a nurse, Kevin

[bryan]

+ this begs another question: to what extent do we force ourselves to find something to like in what we have paid money for ???

I noticed that I'm more severe with records I get for free than records I bought. I dismiss free stuff very fast if I don't like it, whereas I give more tries to bought CD's.

yes.

gotta go, but more thoughts on this later (hopefully)..

[jon abbey]

+ this begs another question: to what extent do we force ourselves to find something to like in what we have paid money for ???

I noticed that I'm more severe with records I get for free than records I bought. I dismiss free stuff very fast if I don't like it, whereas I give more tries to bought CD's.

yes.

gotta go, but more thoughts on this later (hopefully)..

potentially interesting topic, but should be in a new thread, not this one.

[Damon_Smith]

I saw the new Pixar Film, Wall E (A fantastic film for what it is, their best yet, IMO), at Pixar because a friend works there. Their Theatre is totally state of the art, everything looks perfect, even the last Star Wars film sucked way less there than other places - if you can believe that.
I still think the film will hold up in the various lesser formats from basic theatres to dvd, tv, ipod etc.

I will concede to the audiophiles that other formats offer a better, more full experience but if the music is not strong enough to translate to mp3, I don't think the format is at fault unless the main materials are exactly what mp3s do not retain.

[bryan]

I don't think the format is at fault unless the main materials are exactly what mp3s do not retain.

Case dismissed! The format has been exonerated.

[Vincent Kargatis]

Case dismissed! The format has been exonerated.

Actually, I kinda think that argument does do that, in that it implies that the sonic compromises are subject to comparative valuation against the gains from degree of portability, and subsequent advantages. And contrary to (implied) claims above, equivalent alternatives don't exist, and won't until storage capacity, storage size (i.e. form factor), and bandwidth are cost-independent. Until then, sonic quality and portability can be subjectively and reasonably balanced by different people in different ways.

[Damon_Smith]

I think the cases like Richard mentioned, where he could pinpoint a specific sound on the cd that was not present on the mp3
are obviously valid.