Manufacturer's Comment
Editor: Jim Austin's assessment of the lossy/lossless question is astute. After developing lossless compression (footnote 1), we realized that losslessness in the digital transmission path was not addressing some critical issues. Jim raises the question "lossless compared to what," and this is a key point. Data can be transferred bit-for-bit to a D/A converter, but there is no such thing as a lossless converter to or from analog; and, in reality, internal signals in practical converters are rarely bit-repeatable. So, MQA is stepping beyond the lossless pipe to attack the problem of providing to the human listener, as closely as possible, a "lossless" analog-to-analog path, tackling practical issues in the converters as well as real-world topics such as convenience and flexibility (footnote 2).
Any engineered system involves design decisions and the balancing of attributes. Rather than continuing progress by making a "faster horse," we are taking insights from advanced sampling theory, from auditory neuroscience, and from the fruits of hundreds of hours of listening, to reach an intelligent balance. The authentication shows that the sound signed off on by the label is faithfully reproduced; the total lack of modulation artifacts distances MQA from both "lossy" codecs and from real-world "lossless digital." In the end, there is only one real way to judge, which is to listen.
Finally, a comment about the analysis of 2L's Veslemøy Synsk (2L 2L-078). This album (released in May 2011) was one of the first to be encoded to MQA (in December 2015). DXD is the most revealing and challenging source; successive generations of converters vary a lot in terms of noise floor (see fig.1). Over the last two years we have steadily improved the encapsulation and deblurring stages to best match the various converter designs, and in the case of 2L, several albums have been re-encoded where [2L recording engineer] Morten [Lindberg] and we felt it was advantageous (footnote 3).
Footnote 1: The MQA technical team, which is still led by Peter Craven and Bob Stuart, first demonstrated lossless compression to the music industry back in 1995. The late Michael Gerzon was part of that initiative.
Footnote 2: For example, the hierarchical nature of the MQA system allows a single file to be used in several replay scenarios, including with no decoder.
Footnote 3: Including important examples such as Magnificat (2L-106), Mozart Concertos (2L-038-2016), Piano Improvisations (2L-082), and Quiet Winter Night (2L-087).
Fig.1 Analysis of 2L 2L-078. Background noise is assessed over the 3 seconds between 0:01 and 0:04. The music is captured at 2:50. The original noise (dark blue) and music (magenta) are plotted for the full DXD frequency range. The decoder output is shown for comparison: a recent encoding of noise (brown) and music (orange); the noise floor of the December 2015 encoding (cyan). The plots of the 2015 encoding agree well with Jim Austin's analog measurements of that file (although in the analog domain thermal noise has limited the dynamic range below 48kHz) (24dB/vertical div.).
2L-078 is one of a dozen albums that used the early Sphynx2 converters; many of these were scheduled to be remastered this winter, and should be available from 2L by the time this issue is available. Fig.2 shows that, using our latest tools, there is no longer added noise in the audioband (even though it was already inaudible and sounded great).—Bob Stuart, MQA
Fig.2 Four models of DXD converters, idle noise-floor spectra (24dB/vertical div.).
Footnote 1: The MQA technical team, which is still led by Peter Craven and Bob Stuart, first demonstrated lossless compression to the music industry back in 1995. The late Michael Gerzon was part of that initiative.
