By Lam Seng Fatt
Soon after the shootout between the Toslink and co-ax during the recent KL International AV Show, I became ‘mad’ because I started to hear things – like differences between file formats. To me, FLAC, AIFF and WAV files sounded different.
I googled around and found out that I was not the only one who was ‘mad’ and could hear funny things. In several forums, debates on the differences in sound quality between file formats have been raging on for years.
On one side were those who were ‘mad’ and could hear things and on the other side were those who proclaimed that “lossless is lossless and nothing more or less”. According to them, FLAC (which is compressed lossless), and AIFF and WAV (which are uncompressed lossless) files all sound the same because the 1s and 0s are all the same. If there were any differences in sound quality, it was due to some phenomena in the digital conversion process.
It can be scientifically proven that all the files of the same song in FLAC, AIFF and WAV are bit perfect. Someone even took a WAV file, converted it to FLAC and reconverted it to WAV and compared the second WAV file with the original WAV file. There was no difference. The second was an exact copy of the first file.
So, why should the FLAC file sound different from the WAV file? I don’t know. I have no explanation for my ‘madness’.
There was some discussion that the AIFF file was invented by Apple and was designed for the Apple OS and the Motorola chip (Macs now use Intel chips) while the WAV was invented by Microsoft/IBM and was designed for Windows and the Intel chip.
There was also discussion on the “endian” properties of files. According to Wikipedia: “The terms endian and endianness refer to the convention used to interpret the bytes making up a data word when those bytes are stored in computer memory. In computing, memory commonly stores binary data by organizing it into 8-bit units called bytes. When reading or writing a data word consisting of multiple such units, the order of the bytes stored in memory determines the interpretation of the data word.
Each memory storage location is associated with an index, called its address, which indicates its position. Bytes of a single data word are generally stored in consecutive memory addresses (e.g. a 32-bit int needs 4 such locations).
Big-endian systems store the most significant byte of a word in the smallest address and the least significant byte is stored in the largest address.
Little-endian systems, in contrast, store the least significant byte in the smallest address.”
WAV files are little-endian byte order whereas the AIFF format was big-endian.
There was also debate over the process of compression and decompression and reconstruction of the data doing something to the music with a FLAC file.
I found out about such’madness’ when I was preparing for another shootout – this time between CD and CAS using a CD player as transport and a laptop playing ripped CD files via USB to the same DAC. I was planning this shootout to silence critics who kept on insisting that CAS could never match CD players.
One audiophile insisted that I use two test tracks from the KL International AV Show Commemorative CD 2014. He requested the first track – Suites 6 in G Major – Prelude (J.S. Bach, Cello Suites) by M.A. Recordings – and the fifth track – Lu Opera; Lady General Mu Takes Command.
Using J. River V 19 and my Lenovo (i5) laptop, I ripped these files into AIFF format and started to compare them with the CD. I used my Roksan Caspian M1 CD player as transport with Van den Hul Digicoupler co-ax to the Wyred4Sound DAC2 and the laptop with Furutech USB cable as the CAS competitor. The DAC was connnected to the Lamm LL2 deluxe preamp with WyWires Platinum interconnects and thence to the Bryston 4BSST also with WyWires Platinum. The speakers were the ATC SCM40 and speaker cables were the Kimber 12TC.
I was quite happy with the comparison with the Lu Opera track which IMHO resulted in the laptop being the winner. However, with the Bach track featuring a cello, the sound via the laptop was clearly leaner and the cello sounded devoid of richness and harmonics. The CD player was the clear winner.
It was at that point in time that I suddenly asked myself whether I was mad to have opened my big mouth to proclaim that CAS could take on CD.
But since google is a friend in need, I found out that there were at least two other ‘mad’ audiophiles in this world who insisted that WAV files sounded harmonically richer than FLAC files. One even said he could hear a difference between a WAV file and an uncompressed FLAC file (which is a format offered by the new version of dBpoweramp).
It so happened that I had previously downloaded WAV and FLAC versions of Jean Frye Sidwell’s songs. I played I Left My Heart in San Franciso and immediately heard that the WAV file sounded fuller and richer than the FLAC file.
Using J River v 19, I converted the Suites 6 in G Major – Prelude from AIFF to a WAV file and played it. This time the CAS could match the CD.
For the heck of it, I ripped You Couldn’t Be Cuter (featuring Diana Krall) from Yo Yo Ma’s Songs of Joy & Peace album. The song was ripped with J River on my laptop into three formats – FLAC, AIFF and WAV. So the issues of using a different encoder, a different laptop with different OS and different chip do not arise.
I requested four audiophiles to share my ‘madness’ and asked them if they could hear any differences between the three tracks. All four heard differences. One preferred the treble of the FLAC file and added that there was more sibilance in the WAV track; another said on the contrary the WAV had a richer sound and better clarity which resulted in the sibilance being more prominent. There was general agreement that the bass sounded better with the WAV file.
So it was proven beyond reasonable doubt that I was not the only ‘mad’ guy around who could hear weird things in music files. There were at least four others in the Kuala Lumpur/Petaling Jaya area who were just as ‘mad’.
All the three files are bit perfect, but they sound different. Why? I do not know.