Can people hear the difference between lossy MP3 digital music files and lossless ones? Opinions differ strongly, with much obfuscation around audio cables, mastering and hi-fi componentry muddying the waters.
This article was prompted by commentard critiques of Sonos streaming Wi-Fi speaker/player reviews and audiophiles saying I knew less about hi-fi music than a hole in the head.
Okay, granted I knew less than someone devoted to hi-fi with a dedicated listening room, and carefully selected and matched components often costing upwards of a couple of thousand pounds, sometimes much more.
But I did have Sonos speakers and these, for my iTunes digital music collection, were infinitely superior to the commodity-level Sony and Sharp speakers, amplifiers and CD players I'd been used to. In particular, the Sonos/iTunes setup provided much better listening than a Sony CD player/amplifier and speakers.
At one level this is plain peculiar as CDs are encoded to a much greater fidelity to the original music than an MP3 file.
My thinking was that if MP3 encoding strips off data from the digital sound files then a lossless form of audio encoding should provide better listening, and that means something like a FLAC codec. Is it possible to get FLAC-encoded music played through my Sonos system and could I tell the difference between that and the same music encoded with MP3?
Basic starting points
Let's first run through what I understand sound, hearing, CD-quality, MP3, FLAC, ALAC and iTunes import to mean. This is being done from the point of view of a Mac user and system, but the principles should be the same, I hope, for Windows users.
- The human ear can hear sounds, vibrations in the air, between 20 and 20,000 Hz with older people gradually losing the ability to hear higher frequencies in that range, such as 16kHz and above. Sounds are typically conceived of as waves in the air, propagating through it with the air vibrating as the wave travels though it.
- Sound at different frequencies is heard as different pitches, with higher frequencies heard as higher notes and lower frequencies heard in the bass part of the audio spectrum.
- Digitisation of music involves a bit depth and sampling rate. How many bits are used to encode the music when its sampled, and how many times a second it is sampled. Thus 16/44.1 is 16-bits sampled at 44.1kHz (cycles per second) or 44,100 times a second, and this is CD-level sound. Everything between sample points is lost.
- The bit rate says how many kilobits of data there are per second of audio. The higher the bit rate the larger the output file. If the bit rate is too low we can hear sounds in the music that are not in the original music. These are called artefacts and may affect cymbals and guitar sounds.
The stream of audio data is sampled and compressed. Typical musical encoding starts at 128kbit/s with 320kbit/s being the highest rate. There are different MP3 encoders with different quality output and you can drop the bit rate down to 8kbit/s for rather muddy low bandwidth speech purposes.
- A constant bit rate (CBR) can be used throughout a piece of music. Alternatively a variable bit rate (VBR) can be used, with fewer bits used during silences and simple music passages and more bits being used for complex, multi-instrument parts of the music.
- A CD-level recording has a 1,411.2 kbit/s bit rate. A 128kbit/s encoding of this represents an approximate 11:1 compression ratio.
- FLAC - Free Lossless Audio Codec - with codec being coder/decoder, is a lossless form of compression. No incoming audio signals are discarded. It is one of several lossless audio codecs, with WavPack and WMA being others. It provides bit-perfect copies of CD-quality music (16/44.1) or even higher (24/192k - known as studio quality). FLAC can reduce audio sources to 50-60 per cent of their original size. FLAC is quite widely used now with websites like Presto Classic offering FLAC files.
- ALAC is the Apple Lossless Audio Codec. It differs from FLAC and FLAC files can be converted to ALAC files.
- Apple claims that "the default encoding format in recent versions of iTunes is MPEG-4 AAC, a compressed format that rivals the sound quality of audio CDs." Like MP3, Advanced Acoustic Coding (AAC) relies on perceptual encoding techniques but is more efficient.
When it comes to telling the difference between MP3s and live music, I turned to Reg reader James Kilby.
He says: "My understanding of MP3 is that its mathematically modelled to remove sounds that a human can't hear. For example we can't hear a very high note after a very low note and vice versa. However different peoples ears/brains will react slightly differently so some people may be able to tell the difference between an MP3 and a FLAC."
(Who's James Kilby and why should you care what he says? In his own words: "My dad was a professional musician and I have worked in the live music arena before migrating over to IT so I would say I am somewhat of an authority to give an opinion on this.")
So the theory for higher-fidelity playback of stored music through the Sonos system is to get a FLAC copy of the music, convert it to ALAC, import that into iTunes, re-set the Sonos music index, and then play the music. The process, though, is fraught with difficulties and different people disagree on whether you can hear any difference between FLAC and MP3 music anyway.