MQA Responds to GoldenSound (if this makes no sense, you are spending your time wisely) [Updated]

Equal rights. Justice. A few weeks ago, I pointed readers to a relatively new YouTube hifi reviewer, GoldenSound. The subject of his most popular video, by a multiple of around 15x, was MQA. Spoiler Alert: it was not a positive take.

Here’s said video:

In brief, GoldenSound created a number of files that contained test signals with some also having music embedded in the same file. Using MQA’s free encoding tool for Tidal, GoldenSound attempted to have MQA do its thing to these files. According to MQA, 11 of the 14 files failed to process, and the remaining 3 files returned a number of error messages. That’s about that in terms of the actual test. You can watch GoldenSound’s video for his interpretation of the results.

As is the case with nearly everything written or said about MQA, there are people who find the GoldenSound video convincing, and those who don’t. I think it’s safe to say that most people do not have the technical chops to make these determinations, yours truly included, but they do nonetheless.

Bob Stuart, one the founders of MQA, has recently responded to the GoldenSound video on his Bob Talks blog. Here are a few key points raised in that response:

Responses to Specific Claims

  1. MQA did not delete his files; that accusation is false. MQA is not a rights holder nor distributor. We do not issue takedown notices to distributors or DSPs.
  2. MQA has never made false claims about ‘losslessness’. MQA has been clear from the outset that our process operates in a wider frame of reference that includes the whole chain including A/D and D/A converters. [1]
  3. Provenance: MQA files are delivered losslessly and reconstruct exactly the sound that an artist, studio or label approves.
  4. The blogger’s test failed because he submitted signals that do not resemble music to an encoder that was configured only for music works. Nonsense comes out. This is like being disappointed when a F1 car struggles on an off-road race.
  5. He submitted high-rate composite files containing unsafe levels of ultrasonic signals –in places 100 times higher than in music recordings – resulting in 10x encoder overload. (See Appendix 2)
  6. System error messages generated by the MQA encoder were ignored. [2]
  7. MQA provided detailed feedback to the blogger before publication. [3] He ignored it and later dismissed our detailed guidance as ‘marketing’.
  8. MQA does not add distortion and by design, does not introduce detectable aliasing. (See Appendix 4)
  9. MQA is different from regular PCM for important reasons to do with sound quality. Compared to regular PCM, MQA can deliver higher temporal resolution and lower blur while using less data in delivery.
  10. An MQA encoder can encode any signal that fits in a PCM file, but it is highly optimised for files containing information that is meaningful to human listeners. (See Appendices 3 & 4)
  11. Finally, the title of the video is illogical. [4]

General comment:

The blogger displayed a lack of integrity, violating agreements and Terms of Use with multiple parties within the chain. He claimed to be unbiased, yet from the outset, the narrative pursued an agenda – a libellous manifesto that was unscientific, illogical, inaccurate.

He hadn’t researched how MQA works and padded the video with a litany of alternative facts previously debunked many times by MQA and others.

This is not the first time that people have leveraged MQA to gain ‘clicks’, it won’t be the last.

There are 4 appendices attached to the main post that dig into more detail.

I recommend reading Bob Stuart’s entire reply — All that glitters is not gold(en) — if you are interested in getting both sides of this story.

Update 5.26.21:

GoldenSound just posted his rebuttal to MQA’s rebuttal: