@richardkrebs thanks for that explanation. So MF confuses the adjustment period unit with threshold of audibility. Would be funny if it weren't so sad. Thanks for the clarification re that source kernel.
Interesting engineering issue going again at the accuracy vs. precision question and error correction. Using binning for data-noise reduction is certainly a tried and true approach. I run into it with microscope cameras.
Would be interesting to know what speed variation was considered audible by the panel. Though it may have more to do with sensor noise being reduced by binning. That begs the question of whether the company became victim of its own over-engineering. Making the measurement periods so short that the measurement error goes through the roof. Fascinating trade-offs! Run into that with electron microscopy reducing probe current for better resolution, but that decreases signal and increases noise.
@antinn @pryso Re trustworthiness of MF, when I notice some of his BS, I wonder how much other BS he is spouting that I don't spot. So I actively ignore him. To make things worse, I pointed out to the editor of TAS last year that this is nonsense. I am surprised that the editor is not reining in his authors. That is the function of an editor, have had a few of those appointments with some journals myself. That lack of editorial fortitude is a significant reason why I cancelled my TAS subscription again. So there is real harm done by spouting/publishing BS.
I know, it is just a hobby, for fun and all that. But that sort of thing turns fun & entertainment into annoyance. Don't need that.