Reference

Bit Depth of MQA

Some misunderstanding has come from over-literal interpretation of our origami diagrams (which are intended to inform concepts for a general audience), misunderstandings of information theory and general lack of appreciation of the novelty in our advanced lossless data burying methods. Another source of confusion has come from commentators asserting that MQA executes the examples in some of our older patent applications. These factors led to a number of misunderstandings about MQA capability and some unsubstantiated assertions.

Unsubstantiated assertions from various blog posts:

• a) 'MQA is 13bit'

• b) 'The resulting stream has easily-identifiable high-frequency noise occupying 3 LSB bits, thus limiting playback on legacy devices effectively to 13bit.'

• c) 'The compatible portion of the MQA signal is equivalent to about 13 to 15 bits at a sample rate of 44.1kHz or 48kHz. The loss of resolution is due to dowling, dither noise, and pseudo-random noise from the high-frequency compression channel which occupies the lower 8 to 11 bits.'

• d) 'MQA increases the noisefloor of the recording.'

• e) 'When fully decoded, the resolution of MQA is limited to 17 bits at 96kHz.'

Answers: To form answers, we have paraphrased the assertions and then go on to analyse several recordings to illustrate these points.

i) "MQA files have around 13 Bit of "lossless" information and everything below 14 (17) Bit is "lossy"

This is incorrect. With a decoder, in general, the MQA system can reach either:

• In excess of either 23-bit dynamic range capability for 1x 24-bit original, or

• 15.93-bit dynamic range capability for 1x 16-bit original, or

• In any case 3–6 bits below the content noise in the audio band, or

• 24 bits in pure lossless mode, and

• In all cases MQA's noisefloor is stable throughout a song or work.

• See the Table in MQA Hierarchy earlier.

A crucial thing to understand is that in the majority of MQA files, the signalling noise is reversibly and losslessly removable. It is at the 16th bit primarily for backward compatibility (eg, CD, Airplay, automotive and other 16-bit restrictions in the 'last mile.' The real noise in the recording cannot be seen by normal FFT analysis.

ii) "Without a decoder we hear 13 bits, that isn't CD quality."

As described earlier, if you don't have a decoder, the channel capacity appears to be typically >15 bits and this is limited by considerations of compatibility, not coding space. The noise is frequency shaped to minimise audibility, as it is for many well-produced CDs. If you have a decoder then, depending on the authoring parameters, the noisefloor in the recording should not be increased anywhere there is music signal.

Many producers, mastering engineers and reviewers agree that without a decoder, MQA can sound closer to the high-resolution original than to the CD derived from it. There are a number of reasons for this to do with the way the legacy file is set up for a generic DAC but also to do with the stability of the apparent noisefloor.

iii) Paraphrase: 'The Nielsen recording shows that MQA are cheating. They take a 16-bit recording and give us back a 24-bit file with lots of noise in it.'

Wrong. All one had to do was read Morten's notes to guess it might have been remastered to 24 bits.[41]