The benefits of FLAC for audio
Most users of open source software are already familiar with MP3 files for audio. But less of them know about FLAC (Free Lossless Audio Codec). A writer at Open Source explains why he prefers FLAC and why you should consider it too.
Chris Hermansen reports for Opensource.com:
Those of us who are concerned about software freedom should prefer completely free formats like Ogg Vorbis (lossy) and FLAC (lossless, compressed). We should particularly avoid file formats that include options for digital rights management (DRM). In theory, one might think that DRM is just a mechanism to prevent the unauthorized use (theft?) of someone's intellectual property. However, certain vendors use DRM to force their customers to use their software, and sometimes hardware.
Therefore, I buy digital lossless in strong preference to lossy. In fact, if something is only available in a lossy format, I usually don't bother buying it. And not only do I buy lossless, but I buy it at a higher resolution than "CD standard" when available. And for sure my preferred lossless format is FLAC!
And one more reason to buy the high-res stuff: my experience shows me that when music is released in high-res format, it is often well cared for in the production chain and preserves the original dynamic range (loud is LOUD and quiet is ...) and life of the music, without introducing a bunch of artifacts—noise!—into the music.
When I buy digital music downloads, I buy them in FLAC format and try to get 24-bit files and 88.2kHz or 96kHz sample rates. My music files cost money. Why would I be willing to accept poor quality lossy files? And why would I be willing to let the vendor lock me into a particular software and hardware platform?
Chris' post spawned a very large thread on the Linux subreddit, and redditors weren't shy about sharing their opinions about it:
Audioen: ”The article's explanations for the advantages of 24 bit and higher sample rate recordings are pretty much not relevant though. CD quality audio is still good enough to represent the original signal faithfully. During mastering and recording, higher sample rates and bit depths are used for sake of the fidelity of the recording and audio processing chain. As a mastering target, CD is still overspecified for all but the most demanding environments.
If your amplifier has a dB display, you can check how high up you can turn it before it starts to produce hiss in the speakers. I have one which goes from 0 dB (loudest) to -100 dB (quietest), and hissing becomes audible around -30 dB against the background noise. The quietest levels like -99 dB are just completely inaudible. That is, I can turn amplification up by 70 dB from whatever chosen baseline it has, before any further signal is going to be masked by the amplifier's own noise. In other words, I can't hear even the 96 dB worth of SNR in my home environment because my amp can't reproduce it, and in fact I almost never listen to music any louder than -50 dB which is already neighbor-unfriendly. (So, I believe have some headroom before system noise becomes an issue, and I'm pretty 100% sure that there's way more resolution left in the CD-quality audio spec than my fairly good system can reproduce, but unfortunately these numbers are not calibrated, so I do not know how much exactly. I could measure this by making some quiet test signals, and then measuring the amplification level where they become audible, though.)
If you use technology such as FLAC, the amount of entropy (=predictability) determines the compressibility. An argument can be made that in 24-bit audio most of the bottom 8 bits are basically noise, and therefore also completely unpredictable. Somewhere in the higher bits, there is more predictability. Adding resolution is thus an excellent way to increase the portion of signal that can not be predicted by any LPC, and thus the file size for no good reason.
If 24-bit signal controls a speaker cone with 1 cm excursion from one extreme to the other, then the LSB controls 1/16Mth of that, or roughly 0.6 nm. This is comparable to width of a single atom. At least I hope that nobody will ever propose adding any more bits to signal...”
EchoTheRat: ”The new lossy open format to be preferred should be OPUS .”
Plazman: ”Ok, I'm going to be "that guy."
DO NOT waste your money on high resolution music that goes beyond the CD standard of 16/44.1.
Because you can't hear it. This guy talks about using FLAC and Ogg Vorbis. Perhaps he should hear what the guys at Xiph.org, the guys who created those two formats have to say about "hi resolution music."
5225: ”I use FLAC because I can transcode to whatever format I want from it safely. Oh, and I have 2TB of storage space, and I'm going to use my 2TB of storage space. But if it's a download which I can only find an MP3 of, that's fine. I don't really care.”
Redsteakraw: ”24 bit and 88 / 96khz sample rates is overkill for playback. 24 bit is overkill for dynamic range (for playback) in music. There isn't really a quiet enough space for you to realistically hear the difference. And for sample rates larger than 48000hz you can't hear those frequencies. So unless you want frequencies recorded that your dog can only hear there isn't a point to it for a mere consumer. Now it you are recording, mastering, editing or applying filters to the audio you will have the wiggle room needed for your final consumer master at 12bit / 44.1 - 48 khz.
On lossless vs. lossy is that with the lossless copy(FLAC) don't introduce extra artifacts when you transcode. You have a stand in for the original and trans-code for your target device. For example if you have a limited storage and your intended listening is a noisy train you can transcode to a heavily compressed low bit-rate lossy format. If you are listening in your PC you can listen to the flac directly as storage space isn't as much of an issue. You can transcode to the appropriate codec and bitrate for your use-cases as much as you want as your master is a perfect copy of the original. Like a photocopy machine little mistakes are made, transcoding from a lossy format to another lossy format is like copying a copy. You copy the artifacts of the lossy master and add the artifacts that your new codec has.
TLDR. 24 bit / 96khz is overkill for consumer playback. Lossless is good for archiving and as a base to transcode from.”
Darkeox: ”I agree with this article. If I am to buy music, I usually get the CD (because I like to physically have all the booklets and stuff) and then converts it to FLAC, then to AAC when I need to push it to whatever audio device I'm using.”
Spongewardk: ”FLAC is a lossless format while mp3 is a lossey format. I bet you wouldn't be able to hear the difference between the two no matter how good hearing you have, unless you know specifically what to listen for.
The reason to use lossless formats is if you are processing the audio and you need the full signal. Having a lossly format will cause distortions because it is estimating what the sound will be like.
If you need to suddenly use a higher bit rate, or higher frequencies you are out of luck if you have your source as an mp3. The reason why FLACs and other formats are kept is it keeps as much data as it can while compressing it. It is essentially the original audio assuming the recording was done correctly.”
Imakoala: ”If you have your collection stored someone in FLAC, then you can always transcode it to whatever smaller format you want. I've considered transcoding my collection using a relatively low bitrate, actually, so it all fits on my phone. I use cheap earbud headphones anyways, usually while walking around a loud city, so I highly doubt I could tell the difference.”
TheQuietestOne: ”I'm not religious about fidelity, it's just that I've tested and analysed a few audio files in my time and I've found that MP3 is particularly weak when you have source material at or close to 0dBFS (e.g. dance tracks with heavy kicks + bass) or alternatively very quiet passages.”
Camarade: ”Where MP3 really falls short quickly is when there is one "sound" that is more powerful than the others: (almost) only the louder one will be encoded by MP3. IOW it blurs everything and only the main features are kept.
However to notice that, one needs to know the original. Several times I've had MP3 files (not necessarily 320kbps) and then lossless and thought "woah, there are really more things happening at that time in the track"; with properly encoded MP3 it's not so much about the quality but more about the loss of the minor traits.”