Slashdot Mirror
FLAC Gets First Update In 6 Years
An anonymous reader writes "The Free Lossless Audio Codec, FLAC, loved by audiophiles for its lossless fidelity has been
updated to version 1.3.0. FLAC is an audio format similar to MP3, but 'lossless', meaning that audio compressed in FLAC doesn't suffer any loss in quality. FLAC v1.3.0 is the first update in almost 6 years and it is also the first release from the new Xiph.Org maintainer team."
Big new feature: ReplayGain works for sampling rates up to 192kHz so you can finally control the volume of your obsessively ripped LPs.
The latter.
http://xiph.org/~xiphmont/demo/neil-young.html
While this is mostly accurate, articles like this fail to mention where 192KHz is useful. That is, for certain types of digital post-processing and effects. Doing a digital time or frequency shift (not a re-sample, that's simple and effectively lossless) yields atrociously poor results if using 44.1 or even 48 KHz. With 192KHz, you can't hear the difference, and that is why it is used in the studio. Auto-tune is a decent example of that kind of processing. It works much better at higher bit rates.
None of this matters to the average listener though, or to the DJ who only cares about a simple speed up or slow down (or re-sample).
MP3 compresses audio files so that they have the same playback within the range of human auditory sensation. FLAC is superior because it retains full audio fidelity across the entire frequency spectrum. This will be of the utmost importance if you are a dog.
Opus is ONLY useful for voice and low-bitrate audio. For high-fi stuff, it's no better than anything else.
I sure as hell would never use it for music.
No. Both are methods of compressing audio data for later playback, just with different trade-offs.
With MP3 of course you are losing fidelity, and with FLAC you are using more disk space and limiting the devices on which your audio data can be played back.
So while they are both different horses for different courses, but they both have the same goal - storage of audio, with data compression.
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
Oversampling at the ADC level is NOT post-processing. Post means "after", in case you didn't know. If the desire is to use audio for a multitude of uses, say to play back in a sampler frequency shifted or to "correct" some awful notes (as is commonly done), it is still worthwhile to record raw audio at 192kHz. It is never worthwhile to distribute the finished product at high sample rates, unless the finished product IS in fact sample material intended to be used by studio people.
Apple never sold music below 128kbps, nor as MP3. They currently sell music as 256kbps AAC.
Vinyl may have a nostalgical value. The quality is indeed WAY worse than that of a CD, for example.
With MP3 of course you are losing fidelity, and with FLAC you are using more disk space and limiting the devices on which your audio data can be played back.
My cell phone (which doubles as the portable music player) can play FLAC, as can my computer and my network-connected home theater receiver. I think my smart TV can play it too, but I've never had a reason to check...
While you're definitely sacrificing disk space, the argument about fewer devices being able to play it is certainly not as true as it used to be. I still carry most of my music around in FLAC format, and just buy a bigger SD card for the phone, and choose some albums I don't want to carry around.
I'm a sound designer, I mostly work on feature films; I use FLAC for my remote archives -- uploading to S3 goes a lot faster this way, particularly when the audio media is sparse. A 20 minute FX premix might be 10% the size of an equivalent WAV because of all the silence. The flac(1) tool also has a handy --keep-foreign-metadata option that generally gives byte-for-byte round trip accuracy, even for embedded metadata. I also use Apple Lossless for my local library, mainly because it supports ID3 and Apple clients (like Pro Tools) support it more commonly than FLAC.
Don't blame me, I voted for Baltar.
FLAC also includes error detection - each frame has as 16-bit crc and the file header includes an md5 hash of raw audio data. Doesn't help with repairing corruption but at least you can detect it and avoid playing the corrupt frames as ear-splitting noise unlike wav.
I'd look at the spectrals on those "lossless" files you bought. Plenty of music on bandcamp was quite clearly converted to flac from MPs.
I'd look at the spectrals on those "lossless" files you bought. Plenty of music on bandcamp was quite clearly converted to flac from MPs.
Here is software that makes it pretty easy to check:
http://en.true-audio.com/Tau_Analyzer_-_CD_Authenticity_Detector
It's worth noting that mobile devices often decode popular compressed audio and video formats in dedicated hardware. Modern, powerful devices can play audio and sometimes video reliably in software, but they use a lot more battery power to do so in comparison, so sticking with formats natively supported by your hardware is still usually the best idea.
I think a few chips got Vorbis support and it wouldn't surprise me to find that FLAC made it in to real hardware somewhere, but there's a reason MP3 was basically the only real portable format choice for years.
I used to get high on life, but I developed a tolerance. Now I need something stronger.
Actually, FLAC is technically similar to MP3 in a sense.
It consists of an inherently lossy encoding in the frequency domain (like MP3) plus an encoding of the difference between the lossily encoded audio and the original. The first part is a bit more straightforward than MP3 because it does not do any tricks adapted to the human ear.
"We mustn't be caught by surprise by our own advancing technology" -- Aldous Huxley
It consists of an inherently lossy encoding in the frequency domain (like MP3) plus an encoding of the difference between the lossily encoded audio and the original.
While a few other lossless formats do this (mostly for backward-compatibility), FLAC does not convert the audio into the frequency domain. It either uses a polynomial or linear function: http://xiph.org/flac/documentation_format_overview.html
FLAC is asymmetric; lots of computrons to encode, but not very much to decode. I had an old iPod Video, and the battery lasted longer playing FLAC in Rockbox than it did playing MP3s in the native Apple software (or in Rockbox, for that matter). Despite being done in software, FLAC is just so stupidly easy to decode that it's nearly a moot point.