MQA, lossy codecs and streaming services

Formats, recordings and ADCs

Answers on this topic

Everything he said on this topic (37)

One-line summaries in our words, oldest first. Each link opens the original.

2014-04-28Watts suggests the hiss may be master tape noise on older recordings, and says AAC streams sound surprisingly good on Hugo, which he calls a format leveler. Head-Fi, Apr 2014
2016-04-22Watts says he listens to BBC Radio 3 at 320k AAC at home and a 4 TB drive away, and that Hugo first made poor recordings musically enjoyable. Head-Fi, Apr 2016
2016-08-07Watts says MQA is still under evaluation, and that DAVE's FPGA and spare expansion connector would let Chord support future developments if they credibly improved sound quality. Head-Fi, Aug 2016
2017-04-11Watts says Radio 3 passes through three lossy sample rate converters, and recommends setting 44.1 kHz for Radio 3 since most music is 44.1 kHz. Head-Fi, Apr 2017
2017-05-13Watts says MQA interpolation filters, with huge timing errors due to aliasing, sounded dreadful against the WTA filter, and urges listeners not to assume MQA is better. Head-Fi, May 2017
2017-05-18Watts says MQA interpolation filters sound dreadful against a standard WTA filter compared on a Mojo, and that MQA's aliasing is unacceptable, so he refuses to implement it. Head-Fi, May 2017
2017-07-25Watts says the filters treat 16-bit and 24-bit data identically and cannot know whether PCM came from MP3, AAC or uncompressed files, though they know the sample rate. Head-Fi, Jul 2017
2017-10-13Watts says AAC sounds soft against 16-bit through a WTA filter because poorer transient timing blurs the starting and stopping of notes. miceblue425 58:36
2017-11-05Watts says how bad MQA is should be easy to hear on Mojo, and more apparent on Hugo 2 because of its transparency and better WTA filters. Head-Fi, Nov 2017
2018-01-18Watts says not to get him started on MQA. Head-Fi, Jan 2018
2018-01-25Watts says BBC Radio 3 on iPlayer, with two SRC stages (44.1>48>44.1) and AAC 320, is far below 44.1/16 and should not be called HD. Head-Fi, Jan 2018
2018-04-14Watts says ADCs may not resolve cable differences; compression tests on poor tracks prove little, but MP3 or AAC 320 sound softer and warmer than redbook, obvious on BluDave. Head-Fi, Apr 2018
2018-05-14Watts cites Archimago's articles, calls MQA a lossy distortion-generating codec with 1% THD and noise at 20 kHz, and worries its rendering creates transient timing errors by avoiding pre-ringing. Head-Fi, May 2018
2018-06-27Watts says aliasing ruins music: Hugo's -130dB DSD decimation was audible, Hugo 2, TT2 and M scaler reach -220dB, MQA manages -40dB, and 768 kHz needs 24 bits. Head-Fi, Jun 2018
2018-06-27Watts says MQA's degradations become obvious with WTA and M scaler; converting the 56-bit WTA output to 24 bits, triangular dither sounded worse than his truncation noise shapers. Head-Fi, Jun 2018
2018-07-01Watts says AAC, short filters, MQA, minimum phase, NOS and DSD sound soft and warm from transient timing uncertainty; turning off the incisive filter similarly sounds warmer. Head-Fi, Jul 2018
2018-08-10Watts says MQA's filter sounds much worse than WTA, and its compression to 44.1/48 adds a notch, 1% aliasing distortion at 20kHz and lossy coding to about 17 bits. Head-Fi, Aug 2018
2018-08-12Watts says the M scaler makes compression losses audible, audiophile BNC cables sound same or worse, TT2 measures identically in Class B, and the units stack. Head-Fi, Aug 2018
2018-08-27Watts says the M scaler draws 13W, asks posters to keep Head-Fi life-affirming, and says MQA's first unfold belongs in the player but the original 44.1k is better. Head-Fi, Aug 2018
2018-08-29Watts says an orange indicator means 48k, so the player converts 44.1k; yellow means 88.2k MQA unfold, which is lossy with a 22.05 kHz notch; original 44.1k is better. Head-Fi, Aug 2018
2018-11-01Watts says 320k MP3/AAC softness is more apparent via TT2 and M Scaler, single-driver speakers are transparent but flawed, and 7W SE suffices for his 803 D3. Head-Fi, Nov 2018
2020-01-02Watts says MQA uses a short minimum-phase type filter that looks only into the past; calling pre-ringing bad misreads sampling theory, since sinc pre-ringing is essential for perfect reconstruction. Passion for Sound 18:00
2021-04-30Watts says 1 ohm below 3kHz trips thermals, the Aeon 2 closed gives the best depth, MQA's -40dB aliasing is unacceptable, and FPGA timing closure rules out 112.896 MHz clocks. Head-Fi, Apr 2021
2021-05-04Watts says MP3 encodes transients poorly and sounds soft, phase or frequency response errors are not damaging, and his ADC work will test 768k to 48k and back. Head-Fi, May 2021
2022-04-26Watts says MQA has 1% aliasing distortion that degrades transients, making bass soft, bloated and fat, which some mistake for better, so Chord does not support MQA. Audio Force 12:40
2022-04-26Watts says his goal is live unamplified sound, MQA can be decoded in the source so Mojo 2 loses nothing, and MQA is fundamentally the wrong way. Audio Force 13:28
2022-07-12Watts says it is unknown whether streaming services send a true stereo track or the player downmixes, and he now prefers streaming to Blu-ray since it allows full M scaling. Head-Fi, Jul 2022
2022-11-23Watts says MQA is not fit for purpose: 1% HF distortion is not hi-fi, let alone high-end audio, and it destroys transient timing. Head-Fi, Nov 2022
2023-02-04Watts says MQA degrades transients, with very limited taps and aliasing, measured at 1% distortion above 20 kHz, and that its makers have no listening tests showing aliasing is harmless. Dusty Audio 55:01
2023-02-04Watts says MQA ignores transient reconstruction; he heard enormous degradation from its interpolation filters, so Chord refused MQA, losing customers at the time. Dusty Audio 55:44
2023-02-04Watts says MQA can be decoded in software, and reports, untested by him, that some prefer playing MQA files undecoded at 44.1 or 48 kHz. Dusty Audio 56:46
2023-02-04Watts advises not using Tidal. Dusty Audio 57:29
2023-02-04Watts says Stereophile would condemn him for 1 percent high-frequency distortion, yet accepts it for Bob Stuart and MQA, calling them idiots. Dusty Audio 57:53
2023-02-04Watts says some like MQA's sound because losing transients impairs pitch perception, making bass fat, flabby and less tuneful with a bloom some enjoy. Dusty Audio 58:14
2025-02-17Watts says his listening tests always use local files on his SSD because with streaming you do not know the source or whether it was changed. audio-2G 11:57
2025-02-17Watts says he uses Qobuz, and when it streams the same recording as his hard disk copy, in practical terms they sound the same. audio-2G 12:31
2025-02-17Watts says streams can be different recordings (two Norah Jones versions differ greatly), and he designs no source because a perfect DAC fed bit-perfect data would make the source irrelevant. audio-2G 12:43