Listening test methodology

Evaluation method and design philosophy

Answers on this topic

• How does Rob run listening tests, which tracks does he use, and how should I run my own?
• What is the WTA filter, how was it tuned, and how does it differ from a conventional windowed sinc filter?

Everything he said on this topic (361)

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

• 2014-06-02Watts says DSD64 and DSD128 HF noise must be filtered digitally, criticizes judging Hugo's decimation in theory, and says he AB tested it against a non-decimating filter. Head-Fi, Jun 2014
• 2014-06-03Watts says blanket claims like "native DSD is best" are usually wrong, that audio is too complex to judge without careful AB listening tests, and that such tests overturned his own assumptions. Audiophile Style, Jun 2014
• 2014-06-19Watts compares Pink Floyd 'Wish You Were Here' on CD and DSD via Hugo; CD had better depth, separation and rhythm, scoring 2-0 to CD. Head-Fi, Jun 2014
• 2014-10-12Watts describes repeated 30-second tests on three tracks, single-blind only for surprises, two months on the new QBD's noise shapers, and avoiding coax transformers. Head-Fi, Oct 2014
• 2014-10-22Watts describes his general and specific listening tests, three test tracks, single-blind checks, variation as the test of transparency, misleading noise bursts, and measuring before serious listening. Head-Fi, Oct 2014
• 2014-10-25Watts gauges an improvement by how long it stays noticeable, from one day to over three months, rates the Olimex isolator a one-week change, and calls Hugo's break-in mental adaptation. Head-Fi, Oct 2014
• 2015-02-02Watts says Hugo was never voiced; he designs for transparency by judging variation in placement, timbre and separation, and recounts how Hugo's musicality after a six-year code upgrade surprised him. Head-Fi, Feb 2015
• 2015-02-02Watts says live acoustic instruments are the only absolute reference, judges transparency by range of variation, notes studios adopting Hugo, and uses electronic music to assess timing. Head-Fi, Feb 2015
• 2015-04-15Watts says he assumes nothing about audibility, tests by listening, and pushes each parameter until he hears no difference, which he calls Hugo's big difference from other devices. article, Apr 2015
• 2015-05-20Watts says he cannot understand why this attribute changes perceived depth, which is easy to hear: on a cathedral organ recording the organ moves backward and forward. needs context HIFICLUBE.NET JVH 7:59
• 2015-11-15Watts says speakers suppress depth but trained listeners hear small depth changes; depth is a small-signal non-linearity problem in electronics, valuable for film soundtracks. Head-Fi, Nov 2015
• 2016-01-29Watts says he had a non-audiophile engineer compare 190 dB and 200 dB noise shapers to convince a company he worked with; the engineer heard the difference easily. article, Jan 2016
• 2016-02-18Watts says his reference loudspeakers are by Velour, a small German maker, his headphones are Noble K10, Audeze and NightHawks, and most of his listening tests were on loudspeakers. Head-Fi, Feb 2016
• 2016-03-24Watts says he will describe the tracks he uses for listening tests in a blog post in the next few days. Head-Fi, Mar 2016
• 2016-03-28Watts's first blog post sets out his listening-test method and nine attributes to judge, argues the brain resolves errors below hearing thresholds, and defines musicality as emotional engagement. Head-Fi, Mar 2016
• 2016-03-29Watts says AB tests struggle to tell transparency from noise floor modulation, that better noise shapers in DAVE improved depth, and that Verilog simulation can quantify very small errors. Head-Fi, Mar 2016
• 2016-04-22Watts preferred a $2 USB cable over a $200 one on DAVE until he separated the headphone cable, then heard no difference; he says AES input injects RF, sounding brighter. Head-Fi, Apr 2016
• 2016-04-26Watts says DAVE adds no artificial depth, since flat recordings stay flat, and attributes compressed depth in DACs to small-signal amplitude errors rather than ringing. Head-Fi, Apr 2016
• 2016-07-09Watts finds a new DAVE slightly brighter and more transparent than his pre-production unit, credits nitrogen production soldering, and says brain break-in outweighs physical break-in. Head-Fi, Jul 2016
• 2016-07-24Watts says WTA aims to align transients to the original recording, was fine-tuned through thousands of listening tests, and a 256-tap WTA beat a 2000-tap conventional filter. Hifistatement 2:28
• 2016-07-24Watts says nobody listens to noise shapers but he always has; FPGA size limited him, and to settle a client company's arguments he built 180 dB and 200 dB versions. Hifistatement 9:05
• 2016-07-24Watts says a non-audiophile engineer found the 200 dB noise shaper easily beat the 180 dB one considered perfect, reported this to his company, and the arguments stopped. Hifistatement 9:49
• 2016-08-01Watts says DAVE's crossfeed is mathematically identical to 1980s analog crossfeed, DAVE adds no reverb, and he recommends hearing it at home since show listening is less sensitive. Head-Fi, Aug 2016
• 2016-09-19Watts outlines Davina test recordings: two prototypes on one mic feed to compare configurations for depth, home and village sounds, then pro engineers, with 768 kHz originals for decimation testing. Head-Fi, Sep 2016
• 2016-10-18Watts says DAC differences are night and day to him, describes a well-regarded DAC sounding awful, and wonders whether others' systems, his own bias or differing sensitivities explain disagreement. Head-Fi, Oct 2016
• 2016-10-28Watts says DAVE sounds great out of the box, otherwise taste or a flawed system is to blame; it highlights headphone differences, and listeners should question their own impressions. Head-Fi, Oct 2016
• 2017-01-15Watts recounts M scaler development and CES demos, discusses judging the scale of changes, lists gains in timbre, bass pitch and depth, and describes improving stop-band rejection. Head-Fi, Jan 2017
• 2017-01-23Watts says Dave's 26,000 to 164,000 tap step disappointed him, algorithm and 16FS mattered more, 0.5M taps was transformational, and explains reducing errors until inaudible with a biscuit-barrel analogy. Head-Fi, Jan 2017
• 2017-02-05Watts says his USB cable tests used headphones only, cable RF may affect pre-amps and power amps, and early differences vanished when he separated the USB and headphone cables. Head-Fi, Feb 2017
• 2017-02-10Watts says both tap count and algorithm matter; in the 1990s a non-aliasing filter became the first WTA, and a 256-tap WTA beat a 2048-tap conventional filter. Head-Fi, Feb 2017
• 2017-03-05Watts cautions one laptop test proves little: Hugo 1 improved enormously with isolation from a Dell PC, while galvanic isolation on production Blu 2 with DAVE gives a substantial improvement. Head-Fi, Mar 2017
• 2017-03-16Watts says his test used the short supplied USB cable and a 10-pound optical cable, and galvanic isolation would not help there, though results vary with source and power amps. Head-Fi, Mar 2017
• 2017-03-26Watts says his laptop having both optical and USB outputs may explain the result by removing source variability, and he will look into it again. needs context Head-Fi, Mar 2017
• 2017-03-26Watts says this was largely already in Mojo, Hugo 2 has better PSU isolation, and he will re-run tests with Mojo, Hugo 2 and DAVE after a surprising result. needs context Head-Fi, Mar 2017
• 2017-04-06Watts reports a USB ground-loop test slightly truncated depth and added brightness; isolation seems unnecessary for headphones, and Hugo 2 is 5 to 10 times less sensitive than Hugo 1. Head-Fi, Apr 2017
• 2017-04-06Watts reports no consistent grounded versus ungrounded difference with a battery laptop, but turning crossfeed on (max blue) more than doubled perceived depth. Head-Fi, Apr 2017
• 2017-05-07Watts notes a phone recording's poor mic, simple ADC and AAC compression, and is amazed the 164,000 versus 1,000,000 tap difference is audible through it. Head-Fi, May 2017
• 2017-05-08Watts says he hears the M scaler in phone clips via bass, warmth and soundstage, and Davina will let him post better clips once its USB works. Head-Fi, May 2017
• 2017-06-16Watts admits -350 dB depth sensitivity seems irrational but says repeated noise-shaper tests, including digital-only truncators, confirm it, and plans a blind Davina test with different noise-shaper settings. Head-Fi, Jun 2017
• 2017-07-19Watts explains how coupling capacitors, reference decoupling and LF distortion add bloom, why Hugo 2 avoids them by tackling noise floor modulation, and why its warmer filter options exist. Head-Fi, Jul 2017
• 2017-10-13Watts says tiny errors interfere with the brain's processing, so whenever he detects, measures or simulates an error he listens and reduces it until he hears no difference. miceblue425 4:35
• 2017-10-13Watts says a windowing function tapers the truncated sinc, and his WTA window is optimized to reconstruct timing, refined over 20 years through perhaps thousands of listening tests. miceblue425 15:15
• 2017-10-13Watts calls the depth change a puzzle and says the most powerful listening test is a null result, since hearing something as better can mislead. miceblue425 34:45
• 2017-10-13Watts says he can decimate the bandlimited file to 48 kHz, rescale it to 768 kHz through the M Scaler, and compare, finally knowing how big the problem is. miceblue425 41:33
• 2017-10-13Watts says listening tests are a major headache. miceblue425 46:53
• 2017-10-13Watts says listening tests require flat emotions, since excitement over months of work can skew results, recalling building and testing his own hi-fi at 15. miceblue425 47:00
• 2017-10-13Watts says an improvement may just suit your system: noise floor modulation brightening a dull system can be indistinguishable from a real gain in transparency. miceblue425 47:43
• 2017-10-13Watts says his listening tests change one variable, such as tap length, with everything else the same, making small differences easy to hear, but care is essential. miceblue425 48:43
• 2017-10-13Watts says there are too many listening tests. needs context miceblue425 49:57
• 2017-10-13Watts says he designs for himself, not for listeners or Chord, aiming at what is best, though others listen once a product is finished. miceblue425 50:04
• 2017-10-13Watts opposes using distortion to shape sound, says Davina will let him control the chain from microphone to loudspeaker, and says every listening test must be backed by theory. miceblue425 51:40
• 2017-10-13Watts says the soundstage change fascinates him because he does not understand it, and finding out why may reveal another way to improve sound quality. miceblue425 52:26
• 2017-10-13Watts says others listen only after something enormous, recalling 1980s solid versus stranded mains cables on a preamp with 220 dB of supply rejection, which he had his girlfriend blind-test. miceblue425 53:06
• 2017-10-13Watts says his girlfriend's blind result, smoother and warmer, matched his, and that the 330 versus 350 dB depth finding seemed too crazy to post. miceblue425 53:52
• 2017-10-13Watts says he could hear 350 versus 330 dB and others he asked could too, then turns to his silicon design career. miceblue425 54:37
• 2017-10-13Watts says a non-audiophile engineer compared 180 dB and 200 dB noise shapers: the 180 dB sounded harder and brighter from noise floor modulation; the 200 dB had better depth. miceblue425 55:42
• 2017-10-13Watts says he has big problems with double-blind testing. miceblue425 1:00:28
• 2017-10-13Watts says once listeners are confused about which source they are hearing they become stressed and their sensitivity falls apart. miceblue425 1:00:34
• 2017-11-04Watts says current ADCs measure poorly; bandwidth limiting removes HF noise that causes brightening noise floor modulation, but Davina's noise rises only above 500 kHz, and tests need careful evaluation. Head-Fi, Nov 2017
• 2017-11-16Watts says only controlled listening tests settle whether CD is good enough, and that Davina will let him publish original versus decimated and M scaled tracks to gauge the loss. Head-Fi, Nov 2017
• 2017-12-05Watts says audibility cannot be assumed and needs careful structured listening tests, and that chip designers work to a marketing specification without listening, calling audiophiles 'audio fools'. RMAF 1:41
• 2017-12-05Watts describes his method of finding errors and listening-testing each, says the best test is hearing no change, and warns brighter can mean transparency or slight noise floor modulation. RMAF 2:55
• 2017-12-05Watts recounts having a chip manufacturer's engineer compare two noise shapers at his house, the first with 180 dB distortion and noise, a hundred times better than standard ASIC chips. RMAF 3:57
• 2017-12-05Watts says the second noise shaper was 200 dB, and the engineer was surprised how easy the difference was to hear; the first had some noise floor modulation. RMAF 4:45
• 2017-12-05Watts says differences at minus 180 dB at 20 kHz and tiny noise floor modulation should not be audible, yet are, so assumptions fail and listening tests are needed. RMAF 5:33
• 2017-12-05Watts says that since the listener cannot know the recording, he judges changes by whether variability increases, which shows improved fundamental transparency. RMAF 7:43
• 2017-12-05Watts says the variability criterion works even with electronic music: more changes in space, rhythm, pace and timbre mean more transparency. RMAF 9:25
• 2017-12-05Watts says the fix took DAVE's distortion below minus 150 dB, being the last mechanism he could think of, and the listening test showed a big change in sound quality. RMAF 14:31
• 2017-12-05Watts says that, playing Shaun the Sheep through the DAVE prototype, his son, unaware of any change, immediately noticed the improved clarity. RMAF 15:45
• 2017-12-05Watts says his son, a guitarist and not an audiophile, heard an unannounced code change, showing tiny differences matter, and sets a target of 144 dB or below per harmonic. RMAF 16:30
• 2017-12-05Watts says he listens only after thorough measurement, and shows DAVE with noise floor modulation eliminated at 2.5 volts and harmonics below minus 150 and minus 155 dB. RMAF 48:25
• 2017-12-05Watts says John Franks opposed Blu 2, but Watts wanted a million taps, expected a small gain, spent six months, and got a tremendous change, which only listening reveals. RMAF 58:11
• 2017-12-09Watts says his DPLL only tracks long-term frequency when locked, so source jitter has no bearing; it took six years of listening comparisons to perfect. Head-Fi, Dec 2017
• 2017-12-31Watts describes his method: identify an aberration, test its audibility by listening, reduce it until inaudible, using simulation to -300 dB, aiming to close the gap to live music. Head-Fi, Dec 2017
• 2018-01-23Watts says simulating the transient test would take enormous time, so hardware is easier; the harmonic-reduction change was not RF noise, and repeated tests are building evidence it is audible. Head-Fi, Jan 2018
• 2018-01-25Watts recalls hearing no difference between 2Qute's supplied PSU and a car battery, notes a supply could affect sensitive amps, and advises comparing with and without the charger before upgrading. Head-Fi, Jan 2018
• 2018-02-16Watts recounts that Qute owners' PSU reports led him to add regulation and RF filters to 2Qute, which then sounded unchanged on a car battery; try a USB battery pack. Head-Fi, Feb 2018
• 2018-02-23Watts calls the filter the most advanced filter technology, redefining CD, and says he listened alone, sometimes with his sons, first hearing 512,000 taps in August or September. MajorHiFi 21:37
• 2018-02-23Watts says he needed others to confirm what he heard; the million taps worked by Christmas Eve, launched at CES, where Jude was the second person to hear it. MajorHiFi 22:30
• 2018-02-23Watts recounts playing a borrowed Wigmore Hall CD at the show, where he and Jude immediately heard the subway rumble, confirming the change was significant and easy to hear. MajorHiFi 23:15
• 2018-02-27Watts says the percentage came from a poster wanting a number; he prefers accuracy to a sinc in bits, but sound quality matters and only careful listening tests reveal it. Head-Fi, Feb 2018
• 2018-03-08Watts says a session comparing mains, cables and servers brought disagreement but nothing technically unexplainable; all agreed his battery MSI laptop had better depth, though others preferred the custom server. Head-Fi, Mar 2018
• 2018-04-12Watts says the HF filter matters for 96k and 192k, the 256FS filter is easiest heard on piano over extended listening, and changing lights show the filters work. Head-Fi, Apr 2018
• 2018-07-21Watts says his WTA algorithm beat a conventional filter with a tenth of the taps, was fine-tuned through thorough listening tests, and needed changes as tap length grew. c w 13:58
• 2018-07-21Watts says three files will be published; if the difference is huge he will do 10 or 16 million taps, though part of him would like the project closed. c w 32:33
• 2018-08-06Watts found BNC cable differences for Hugo 2 vanished once the headphone cable did not touch them; length and exact 75 ohm do not matter thanks to de-glitch circuitry. Head-Fi, Aug 2018
• 2018-09-01Watts recommends neutral single-blind AB tests against placebo, says low-cost FPGAs have stalled since Xilinx 7 series, and nitrogen reflow soldering improves reliability and sound. Head-Fi, Sep 2018
• 2018-09-21Watts says cable placement once upset his AB tests, brain break-in is real, Hugo 2 warms up in 10 minutes, and he has seen no evidence of hardware break-in. Head-Fi, Sep 2018
• 2018-09-27Watts distinguishes perception from hearing, argues brain correlation exposes errors below noise, and recounts a single-blind -180 vs -200 dB noise shaper test an engineer heard easily. Head-Fi, Sep 2018
• 2018-10-27Watts says the M Scaler's musical benefits matter most, whether it is transformational depends on the listener and system, and buyers should listen first with time. Head-Fi, Oct 2018
• 2018-11-18Watts says TT2 needs no balance control with digital volume, a 705/768 indication means the M Scaler works, and suggests a week of listening before comparing pass-through. Head-Fi, Nov 2018
• 2018-11-26Watts says BNC directly couples grounds so it matters most to remove; listeners may hear no difference between inputs, but should disconnect during listening tests. Head-Fi, Nov 2018
• 2018-12-06Watts says toggling pass-through on a flight with Hugo 2 and Aeons was a huge difference, smaller only with non-transparent transducers, and stress kills AB test sensitivity. Head-Fi, Dec 2018
• 2018-12-12Watts says designers who sculpt a signature sound make all recordings sound alike, and that the variation heard at live events is the key to judging transparency. Pursuit Perfect System 5:07
• 2018-12-12Watts says listeners do not know the recording chain, so it is hard to tell whether a change makes a system more transparent or merely fine-tunes its sound. Pursuit Perfect System 5:53
• 2018-12-12Watts says he judges progress by variations heard in live music, such as timbre differences between instruments and convincing spatial distance, like an organ 200 feet away in a cathedral. Pursuit Perfect System 6:40
• 2018-12-12Watts lists further variations to judge by: space, dynamics, pace and rhythm, instrument separation and power without cross modulation, and the starting and stopping of notes and sharp cracks. Pursuit Perfect System 7:30
• 2018-12-12Watts explains shrinking a design by rewriting code or simplifying it, and says a simpler, smaller noise shaper he tried unexpectedly sounded better in a listening test. Pursuit Perfect System 21:26
• 2018-12-12Watts says he built a 220 dB noise shaper, verified it in simulation, heard deeper depth on switching against the 200 dB one, then spent three months improving it. Pursuit Perfect System 23:38
• 2018-12-13Watts says the WTA algorithm was refined from theory and hundreds of listening tests, so that a 256-tap WTA filter sounded better than a 2048-tap conventional filter. Pursuit Perfect System 7:30
• 2018-12-13Watts says he judges improvements by checking timbre, space, detail and separation separately, and that RF noise pumping with the signal sounds bright and etched, faking transparency. Pursuit Perfect System 25:00
• 2018-12-13Watts says if saxophones sound brighter, detail seems better but timbre variation is worse, a red flag, and that his listening tests usually target a known engineering variable. Pursuit Perfect System 25:55
• 2018-12-13Watts says he varies one measured variable, such as noise shaper performance or distortion, and listens to its effect; many in audio wrongly think distortion sounds better. Pursuit Perfect System 26:39
• 2019-04-21Watts says DAVE is connected to mains earth, ground boxes have the whiff of snake oil, and buyers should do careful AB tests before buying. Head-Fi, Apr 2019
• 2019-06-08Watts says TT2's RCA and XLR outputs can feed two systems at once, with probably negligible effect, and suggests checking by unplugging one. Head-Fi, Jun 2019
• 2019-06-15Watts names his depth test track, Mussorgsky's Pictures at an Exhibition, Gnomus, and says he will be available at Chord's booth or after his seminars. Head-Fi, Jun 2019
• 2019-07-05Watts says if it locks it is fine; optical switchers he has used work well, and an AB listening test can confirm no degradation. Head-Fi, Jul 2019
• 2019-07-17Watts says a DX output is dual AES plus volume data, then argues listening judgements and measurements both need great care, saying only the APx555 measures noise floor modulation accurately. Head-Fi, Jul 2019
• 2019-10-18Watts says pass-through is not bit perfect; it carries the M scaler's 2.76 dB insertion loss, so DAVE direct must be turned down 3 dB for comparison. Head-Fi, Oct 2019
• 2019-11-30Watts prefers the max setting as most like a concert, then describes his AB listening tests on Aeon headphones and B&W 803 D3 speakers, Chord's own tests, and Head-Fi feedback. Head-Fi, Nov 2019
• 2019-12-14Watts tells a TT2 owner every DAC upsamples, the M scaler matches sinc to 16 bits for accurate transients, blind tests are insensitive, and he uses Aeon Closed on planes. Head-Fi, Dec 2019
• 2019-12-18Watts says M scaler output is 11th-order noise-shaped to 24 bits at 768k for -301 dB accuracy, ranks Gaussian dither above TPDF, and advises keeping the source bit perfect. Head-Fi, Dec 2019
• 2019-12-18Watts points to his depth test track on YouTube; the link itself is missing from the record. Head-Fi, Dec 2019
• 2019-12-19Watts says required accuracy is found by making filters more sinc-like and listening; his first WTA filter, DAC64 in 1999, had 1,000 taps, the M Scaler one million. Passion for Sound 22:05
• 2019-12-19Watts says each increase improved timbre variation, instrument separation, focus and bass definition, and he will keep going until he can no longer hear a change. needs context Passion for Sound 22:50
• 2019-12-19Watts says WTA was shaped by the maths, removing image aliasing, and hundreds of listening tests tuning the taper; his first test pitted 256-tap WTA against 2,000-tap Kaiser. Passion for Sound 24:25
• 2020-01-20Watts says source-related SQ changes are now small, that low-power laptops and especially his phone are the best sources, and that he will retest sources against optical. Head-Fi, Jan 2020
• 2020-02-22Watts finds TT2 slightly warmer in high gain, says not to exceed -3L in low gain, to match volume numbers when comparing, and low gain is for IEMs. Head-Fi, Feb 2020
• 2020-06-05Watts says his listening is now very consistent, unlike the early 1980s when sound was best at 2 AM before he discovered RF noise causing noise floor modulation. Head-Fi, Jun 2020
• 2020-06-25Watts says he assesses designs by how much variation he hears in timbre, tempo, pitch, placement and depth, since we cannot know how recordings were made. Part-Time Audiophile 13:55
• 2020-06-25Watts says that without knowing the original event, hearing more variation across these factors shows a product is more transparent because it is more expressive. Part-Time Audiophile 14:40
• 2020-06-25Watts says replacing the conventional IR type filter with a WTA filter running to 88 nanoseconds was immediately audible in the starting and stopping of notes, such as piano attacks. Part-Time Audiophile 24:23
• 2020-06-25Watts says a listening test showed a surprising difference, and recalls evaluating prototypes in the evening while watching Curious George with his young children. Part-Time Audiophile 34:49
• 2020-06-25Watts says his son noticed the change unprompted; lower distortion improves clarity and detail, and assuming distortion below the ear's roughly -90 dB sine-wave limit is inaudible is dangerous. Part-Time Audiophile 35:33
• 2020-07-09Watts defines depth as how far away a sound seems and describes his depth tests using a cathedral organ, judging only which version sounds farther away. Head-Fi, Jul 2020
• 2020-07-29Watts proposes testing a 64-bit floating-point PC EQ by setting all bands to -1 dB and raising DAC volume 1 dB; he heard added brightness, worse separation and flatter depth. Head-Fi, Jul 2020
• 2020-08-11Watts advises waiting for Dave, then comparing a desktop PC with an ungrounded battery laptop; if they sound the same, the source need not be a concern. Head-Fi, Aug 2020
• 2020-08-13Watts estimates 2 m over 1 m BNC gives 25-33% of the ferrite gain, explains the ground loop, recommends Wave, and says other audiophile USB and BNC cables match or degrade certified ones. Head-Fi, Aug 2020
• 2020-10-10Watts says R2R DACs sound as bad as they measure, and that subjective listening can be objectively quantified if parameters are evaluated separately, though it is hard. Head-Fi, Oct 2020
• 2020-11-11Watts compares 1M-tap rectangular sinc, Kaiser and WTA filters by measurement and listening, explains how the WTA equation was tuned, and concludes algorithm matters more than tap length. Head-Fi, Nov 2020
• 2021-01-20Watts says XLR pin 2 equals the RCA signal, the outputs handle over 5A RMS, single-ended drives the Abyss 1266 to 120dB, and balanced comparisons must be volume matched. Head-Fi, Jan 2021
• 2021-01-23Watts says reducing NFM and timing errors correlates with musical involvement, warns impressive sound can mislead listening tests, and notes he missed TT2's small DAC-mode loss before production. Head-Fi, Jan 2021
• 2021-02-04Watts says he would never use analog master tapes as a reference; his future tests will compare reproduced sound with the original live sound across the whole chain. Head-Fi, Feb 2021
• 2021-03-29Watts declines to discuss future products, describes a -350dB artifact his son identified blind, and says pulse array residual noise is resistor thermal noise traded against distortion. Head-Fi, Mar 2021
• 2021-05-04Watts says audible changes need a science-based explanation, citing ADC-project sample rate converter tests (aliasing from -200 to below -350 dB) his son heard blind, and M scaler tempo perception. Head-Fi, May 2021
• 2021-05-08Watts says he plans headphone tests of depth, impact and timbre for his ADC, microphones and headphones, plus a 768k-to-48k-to-768k test of decimation and WTA filters. Head-Fi, May 2021
• 2021-06-13Watts says the pro version will have at least 8 channels and a word clock input, preferred Aeon over Ether 1 closed, and needs closed headphones for travel and tests. needs context Head-Fi, Jun 2021
• 2021-10-03Watts says WTA tuning balanced image attenuation against sinc-like transient reconstruction through thousands of listening tests, and that out-of-band images also cause transient timing errors. Head-Fi, Oct 2021
• 2021-12-14Watts says effectively zero errors remain audible; he minimizes placebo and confirms extraordinary results with blind tests with his son. Head-Fi, Dec 2021
• 2022-01-31Watts says a scaler that small is more like 10 years away, so 768 kHz support is sensible; he did his listening tests with the M Scaler as source. Darko.Audio 5:16
• 2022-01-31Watts says EQ cannot be bit perfect, 64-bit floating point on a PC is not good enough for EQ, and proposes a test: all bands +1 dB, volume -1 dB. Darko.Audio 19:43
• 2022-01-31Watts says COVID-era listening tests revealed very small audible effects, so the DSP had to reproduce -301 dB perfectly in phase and amplitude, requiring a different core. Darko.Audio 26:13
• 2022-01-31Watts says ears detect such tiny effects; he ran blind tests with his son, who independently heard better depth and focus, matching Watts's own impression. Darko.Audio 29:21
• 2022-01-31Watts says he now mostly tests on headphones; speakers give better sensitivity and accuracy, but he can still hear it, and closed-back headphones shut out household noise. Darko.Audio 30:28
• 2022-01-31Watts says inner detail and separation are easier on headphones; he tests timbre variation, separation and focus, and depth, likening depth to hearing a distant cathedral organ. Darko.Audio 31:09
• 2022-01-31Watts says he uses three tracks for depth, heard thousands of times, plus one track for separation and focus and another for separation, focus and timbre variation. Darko.Audio 32:21
• 2022-01-31Watts says his depth tracks are organ and electronic music; he also uses a fast guitar piece and a Norah Jones double bass track used for 15 to 20 years. Darko.Audio 32:52
• 2022-01-31Watts says professional tests focus on attributes, not the music; after A/B tests he listens for pleasure, and judging emotional response takes about a week. Darko.Audio 33:59
• 2022-01-31Watts says depth perception is easy to judge: something is deeper or not. Darko.Audio 34:42
• 2022-02-09Watts urges a poster not to listen with their eyes or their wallet. needs context Head-Fi, Feb 2022
• 2022-03-19Watts says he expected to be called a liar over -301 dB, but hundreds of listening tests show depth degrades without it, and science explains evidence. MFG Audio Reviews 11:30
• 2022-03-19Watts says WTA was tuned mostly by listening, down to ten parts per million on the windowing function, and Mojo 2 has over 500 times a chip DAC's processing. MFG Audio Reviews 22:49
• 2022-03-19Watts says lower out-of-band noise makes Mojo 2 slightly warmer, existing DSP is not transparent, and proposes a test: each EQ band up 1 dB, volume down 1 dB. MFG Audio Reviews 29:18
• 2022-03-19Watts describes quickly listening to four settings of a detuned noise shaper without simulation: setting three sounded best, two and four degraded depth, and one was flat. MFG Audio Reviews 31:36
• 2022-03-19Watts says the delays were 2 microseconds, 100 nanoseconds and under 10 nanoseconds, he chose the best unaware, and asks how the brain detects 100 nanoseconds at minus 201 dB. MFG Audio Reviews 33:04
• 2022-03-19Watts says the question always bothered him: he was not at the recording and knows nothing of its acoustic or microphones, so how can sound quality be judged? MFG Audio Reviews 47:12
• 2022-03-19Watts says he has mistaken brighter for more transparent when it was more noise floor modulation; the real test is variation, bright instruments brighter and warm ones warmer. MFG Audio Reviews 48:09
• 2022-03-19Watts says to judge separation, focus, timbre variation and space; more variation is a fundamental improvement whether or not you like the sound, regardless of recording or system. MFG Audio Reviews 48:56
• 2022-03-19Watts says his A-B tests now target a measurable digital-domain error such as -301 dB phase shift, reducing it until no difference is heard, which yields incredibly small numbers. MFG Audio Reviews 49:42
• 2022-03-19Watts says when a change is known to be tangible and no difference is heard, there is no question whether it is a real improvement. needs context MFG Audio Reviews 50:28
• 2022-03-21Watts warns about placebo in listening tests and explains the HF filter reduces noise floor modulation while WTA2 (16FS to 256FS) improves transient timing, neither changing in-band response. Head-Fi, Mar 2022
• 2022-03-22Watts says the relationship is neither proportional nor logarithmic: a further tenfold error reduction still matters, tiny things need careful listening tests, and he wants bigger FPGAs. article, Mar 2022
• 2022-03-27Watts recounts redesigning Qutest's regulation after Qute proved PSU sensitive, admits his 2016 USB preference on Dave was RF brightness, and calls optical the reference input. Head-Fi, Mar 2022
• 2022-04-09Watts says -301 dB is about 600 electrons driving a Stealth, yet many listening tests, some blind, show it matters, as with Dave versus TT2's small difference. Head-Fi, Apr 2022
• 2022-04-13Watts describes his -301 dB Verilog simulation test with an 80 bit 768 kHz source, explains not publishing, and ties poor reproduced depth to small signal accuracy. Head-Fi, Apr 2022
• 2022-04-14Watts explains Verilog module simulation, says improving Dave's noise shapers from 200dB to 350 dB improved depth each step, and treats the -301 dB test as his transparency metric. Head-Fi, Apr 2022
• 2022-04-15Watts says he embeds listening options in the FPGA behind bit switches; simulation suits steady-state errors, not transients, and analog changes need SPICE and slow layouts. Head-Fi, Apr 2022
• 2022-04-26Watts says the original Mojo was voiced warm for bright headphones, so neutral Mojo 2 needed EQ; his test: raise a PC EQ 1 dB, lower volume 1 dB. Audio Force 6:09
• 2022-04-26Watts says a flat but active PC EQ still sounds flatter in soundstage, less detailed, brighter and worse in timbre, so he set out to design a lossless DSP. needs context Audio Force 7:01
• 2022-04-26Watts recommends listening at a dealer or buying online with a return option, since tastes differ, then thanks the hosts and says it is great to be back in Thailand. Audio Force 17:38
• 2022-04-29Watts says the DCA Stealth beat the Aeon 2 in a volume-matched AB test (Stealth +7 dB), with better depth, bass, separation and timbre variation from lower distortion and resonances. Head-Fi, Apr 2022
• 2022-05-03Watts says a brick-wall filter would reconstruct transients perfectly, the WTA filter approximates it via thousands of listening tests, and its optimum is recording-independent because timing error is structural. Head-Fi, May 2022
• 2022-07-07Watts cites the Lipshitz dither paper, says triangular dither leaves signal-dependent errors, and reports his tests: rectangular worst, pseudo-Gaussian closest, but noise shaping still better for depth and detail. Head-Fi, Jul 2022
• 2022-08-15Watts lists six AB test tracks and what each checks, describes switching on short excerpts, prefers 'clarity' to 'transparency', and says musicality is judged by months of normal listening. Head-Fi, Aug 2022
• 2022-08-19Watts says recordings have unknown provenance (acoustic, mics, ADC, post processing), so his listening tests evaluate variation: the more variation in a product's sound, the more accurate it is. AVstore 9:06
• 2022-08-19Watts says he listens for variation in timbre, space, dynamics and instrument separation and focus, aiming for a completely transparent sound that returns to the original performance. AVstore 9:51
• 2022-08-19Watts says his own recordings give an absolute reference, letting him compare a scaler and DAC on headphones with his memory of the event, and calls variability a powerful technique. AVstore 11:20
• 2022-08-19Watts says not all coefficients can be ideal, so his own five-variable windowing function balances gentle tapering against maximizing ideal values, each variable tuned by listening. AVstore 17:43
• 2022-08-19Watts says he tunes each variable listening for separation and focus, soundstage depth, timbre, bass pitch (dependent on transients), and the starting and stopping of notes. AVstore 18:29
• 2022-08-19Watts says digital makes a wood block crack sound softer, that note starts and stops depend on transients, and optimizing five interacting variables took thousands of listening tests. AVstore 19:17
• 2022-08-19Watts says listening tests optimized his windowing function to within ten parts per million, and that while taps are very important, how the filter is designed is absolutely crucial. AVstore 20:01
• 2022-08-19Watts says he uses about nine test tracks, six of them highlighted, with some for depth only, others for bass pitch, clarity and transparency, note starts and stops, or separation and focus. AVstore 21:36
• 2022-08-19Watts says every change must be checked against all the variables, since a change can improve one aspect and worsen another, requiring retweaking. AVstore 22:22
• 2022-08-19Watts says he moves back and forth to home in on the optimum, re-checks earlier work, and after four or five hours of listening loses the will to live. AVstore 22:47
• 2022-08-19Watts says a specific variable is usually audible within five minutes, but near the optimum one track may take fifty comparisons, as 10 ppm shifts are barely audible yet significant. AVstore 23:35
• 2022-09-06Watts says measurements correlate weakly with listening and depend on intent, his published results represent production DAVEs, and buyers should listen at home before purchasing. Head-Fi, Sep 2022
• 2022-09-07Watts says unidentified blind tests kill sensitivity, DAVE has no measurable noise floor modulation, and explains setting up the AP test correctly with -999dB and fixed input levels. Head-Fi, Sep 2022
• 2022-09-18Watts says that in his recording a cello was four meters from the microphone, which he uses to judge reproduced depth. The Source AV 18:49
• 2022-09-18Watts says some things need only a simple, quick test to evaluate. needs context The Source AV 21:53
• 2022-09-18Watts says the WTA algorithm that sets the M Scaler's WTA filter uses his equations with five variables, which are listened to, modified and honed to a sweet spot. The Source AV 22:00
• 2022-09-18Watts says the sweet spot is set to within 10 parts per million by listening for separation, timbre, bass pitch and soundstage, adjusting each variable and rechecking the others. The Source AV 22:30
• 2022-09-18Watts says the WTA algorithm took thousands upon thousands of listening tests with specific tracks. The Source AV 23:15
• 2022-09-18Watts says he uses one track for timbre variation, a couple for depth, and knows what he is evaluating with each specific track. The Source AV 23:23
• 2022-09-18Watts says the tracks he actually uses for testing are pretty exotic. The Source AV 23:43
• 2022-10-12Watts says the M Scaler improves performance, and that improving transient reconstruction accuracy through the filter's design took thousands of listening tests. Upscale Audio 4:25
• 2022-10-12Watts says the filter design has five coefficients, each fine-tuned by literally thousands of listening tests to an accuracy of 10 parts per million. Upscale Audio 4:43
• 2022-10-12Watts describes scoring each coefficient on half a dozen tracks for timbre, pitch, instrument separation and focus, depth and soundstage, then rechecking earlier coefficients. Upscale Audio 5:10
• 2023-01-04Watts reports the Bixpower BP90-12v ran an M Scaler nearly 10 hours and clearly beat the Krisdonia on depth and separation, the Poweradd less clearly. Head-Fi, Jan 2023
• 2023-01-04Watts says design listening tests are his alone, with occasional blind tests with his trained son Daniel, and Chord runs its own evaluations near production. Head-Fi, Jan 2023
• 2023-01-23Watts explains his recording program to test his pulse array ADC and compare live with reproduced sound, says pro ADCs limit performance, and notes listeners differ in sensitivity. Head-Fi, Jan 2023
• 2023-02-04Watts says he thought 200 dB good enough in 2015, but Dave allowed better noise shapers with 20 elements, which he tested by listening. Dusty Audio 22:26
• 2023-02-04Watts says a 220 dB noise shaper audibly improved depth over 200 dB, so over three months of Dave development he kept improving it until the sound stopped improving. Dusty Audio 23:10
• 2023-02-04Watts describes raising Dave's noise shaper through 240, 260, 280, 300 and 330 dB with daily coding, simulation and listening tests, eventually reaching 350 dB. Dusty Audio 23:57
• 2023-02-04Watts says the chip company's engineers disputed his 200 dB noise shaper requirement, costly in gates, for two or three years, until he invited one to his house to listen. Dusty Audio 26:56
• 2023-02-04Watts says he set noise shapers to 180 dB and 200 dB; the engineer was surprised that the 190 to 200 dB difference was so big and easy to hear. Dusty Audio 27:40
• 2023-02-04Watts says the engineer's report ended the arguments, and when a tiny number seems implausible he confirms it with a blind listening test with his son. Dusty Audio 28:28
• 2023-02-04Watts lists variation in timbre, dynamics and note starts and stops, and says listening tests are hard because the listener cannot know what the recording really sounded like. Dusty Audio 30:07
• 2023-02-04Watts says to tell a fundamental improvement from system tuning, ask whether the sound is more varied in depth, timbre, dynamics and note starts and stops. Dusty Audio 30:50
• 2023-02-04Watts says the WTA algorithm, a formula with five functions, was tuned over 25 years by thousands of listening tests scoring pitch, depth, separation and timbre. Dusty Audio 42:48
• 2023-02-05Watts says the unprocessed recording went straight into the ADC; playback matched in frequency response but sounded warmer and softer than live, though he considered the Stealths slightly bright. Soundnews 3:27
• 2023-02-05Watts says the unit's 12 volt input (he says Qutest, later corrected to 2Qute) let him try a car battery after reports favoring linear supplies, and he was astonished. Soundnews 39:13
• 2023-03-03Watts says subjectivism trumps objectivism but listening must be rigorous, part changes never degrade performance, M Scaler gains were clearer with tube amps, and he writes all FPGA code. Head-Fi, Mar 2023
• 2023-03-26Watts says his WTA windowing function has five variables, each with a sweet spot found by listening and adjusting, to an accuracy of about ten parts per million. Eric's Hi-Fi 3:55
• 2023-03-26Watts says thousands of listening tests tuned the five windowing variables to 10 parts per million, rechecking each, using four or five tracks for different sound parameters. Eric's Hi-Fi 9:32
• 2023-03-26Watts says listening matters most but audiophiles may prefer distortion; his goal is closing the gap to live unamplified sound using measurements and careful, controlled listening tests. Eric's Hi-Fi 32:50
• 2023-03-26Watts says confirmation bias and placebo really happen, so a listening tester must be a very careful neutral observer. Eric's Hi-Fi 33:43
• 2023-03-26Watts says it is very easy to fool yourself. Eric's Hi-Fi 34:09
• 2023-03-26Watts says he has made listening test mistakes, redoes tests from years before, and checks impossibly small effects with blind tests in which his son describes the difference. Eric's Hi-Fi 34:16
• 2023-03-26Watts says his son consistently reports better depth and warmer sound, matching his own prior assessment, and repeat tests months later give the same answer. Eric's Hi-Fi 35:03
• 2023-03-26Watts says his son studies Tonmeister sound recording at Surrey and plays guitar, hears pitch errors at once, and is trained to describe separation, depth and detail in Watts's terms. Eric's Hi-Fi 35:38
• 2023-03-26Watts says accurate shared description matters, so he ignores Head-Fi posts that only say something sounds better, giving more weight to detailed descriptions. Eric's Hi-Fi 36:27
• 2023-03-26Watts says training his son Daniel yields objective answers, and that reactions from non-audiophiles are also interesting. Eric's Hi-Fi 37:45
• 2023-03-26Watts says non-audiophiles found music slower with the M Scaler; he agrees, since removing mechanical, disjointed jerkiness restores natural flow that otherwise seems faster. Eric's Hi-Fi 38:17
• 2023-03-26Watts says it is often very good to have a non-audiophile listen to your gear. Eric's Hi-Fi 39:05
• 2023-03-28Watts says the method is to identify an error, test whether it changes the sound, and shrink it until no change is heard, correlating errors to sound quality. The Headphone Show 38:41
• 2023-03-28Watts recounts a company that bought his IP: an engineer in a listening test heard a 180 dB versus 200 dB noise shaper difference he had predicted. The Headphone Show 39:27
• 2023-04-03Watts says every module must pass a -301 dB test within 0.001 dB amplitude accuracy, and every module that failed a listening test had failed the -301 dB test. Amplify Audio 6:28
• 2023-07-15Watts says all M scalers will work with all current DACs, including Mojo 2 and Hugo 2 via dual data, that scaler benefits need long-term listening, and urges forum civility. Head-Fi, Jul 2023
• 2023-07-24Watts says if the ADC works he will share recordings, including test tracks of sounds at different depths and percussion for speed and impact, plus amateur orchestra recordings. Head-Fi, Jul 2023
• 2023-08-22Watts disputes PGGB's pure-sinc claims, says coefficients must be windowed, and suggests listening for bass pitch, separation, focus and timbre variation. Head-Fi, Aug 2023
• 2023-09-20Watts describes tuning WTA variables to about 100ppm with dedicated test tracks, notes Hugo 2's warm settings remove WTA2, and says Mojo 2's UHD DSP adds warmth without timing loss. Head-Fi, Sep 2023
• 2023-09-23Watts describes his cycle of FPGA design, simulation, digital measurement and AB listening tests, and says he has more things to try on the new scaler. Head-Fi, Sep 2023
• 2023-10-01Watts says 64-bit floating-point EQ has noise floor modulation around -300 dB that is audible as brightness, and proposes a +1 dB EQ with -1 dB volume test. The Source AV 10:40
• 2023-10-01Watts says his systematic approach to listening is listening for variation in pitch, dynamics, timbre and space; more variation means the right track. The Source AV 16:21
• 2023-10-01Watts disagrees, recounting how he told a CanJam visitor that headphones he had heard for 30 seconds two months earlier now sounded different. The Source AV 18:44
• 2023-10-01Watts says the visitor changed the cables back and, after 30 seconds of listening, the headphones sounded the same. needs context The Source AV 19:35
• 2023-10-01Watts says that unless you do a carefully constructed objective listening test, you know nothing. The Source AV 23:26
• 2023-10-01Watts describes evaluating an error, reducing it and listening repeatedly, and says other designers either just play with the sound or are technical and do not listen. The Source AV 23:31
• 2023-10-01Watts says knowing the microphones, cables and ADC of his own recordings lets him evaluate DAC performance. The Source AV 25:26
• 2023-10-01Watts says digital simulation shows errors down to -400 dB before a listening test, and when such a tiny change is audible he asks whether it is placebo. The Source AV 34:06
• 2023-10-01Watts asks whether it is confirmation bias, because these things happen. The Source AV 34:43
• 2023-10-01Watts gives the example of telling yourself you have just spent 100 grand on your system. needs context The Source AV 34:47
• 2023-10-01Watts says if you tell yourself it sounds much better several times, it then just sounds better. needs context The Source AV 34:52
• 2023-10-01Watts says in listening tests he must forget six months of struggling with the code and be a neutral observer, and calls in his son when something extraordinary appears. The Source AV 34:57
• 2023-10-01Watts says his son has good ears and is studying to be a sound recording engineer. The Source AV 35:25
• 2023-10-01Watts says his trained son describes differences in his terms without knowing what he hears, making it blind; matching descriptions show the effect is real. The Source AV 35:31
• 2023-10-01Watts says such variation is a clue: find and fix the cause for better sound, with the further benefit that the unit sounds the same from switch-on. The Source AV 48:50
• 2023-11-22Watts says RF via cables and mains matters more than screening, stays humble about unheard ground boxes, and repeats listening tests to guard against errors. Head-Fi, Nov 2023
• 2023-12-22Watts posts updated seminar slides on transients and interpolation filters and explains instrument separation and focus as hearing a quiet instrument in the mix. Head-Fi, Dec 2023
• 2024-03-20Watts says the modeling matched his 1980s mains filter, built by listening into a 12th-order design with hand-wound air-cored PTFE coils, agreeing with what he found subjectively. MFG Audio Reviews 35:41
• 2024-03-20Watts says lithium-ion batteries are ideal for low-power DAC amps, supplies should be judged against battery operation, and a switch mode supply sounding identical to a car battery is perfect. MFG Audio Reviews 39:06
• 2024-03-20Watts says adding a linear supply makes the sound brighter, crystalline and sharper-edged, fooling listeners into thinking it is better; separation, focus and listening fatigue reveal it is worse. MFG Audio Reviews 39:52
• 2024-03-20Watts says listening tests must decide whether something is better, and that DAVE gave him a brand new FPGA with plenty of capacity. MFG Audio Reviews 44:51
• 2024-03-20Watts says even 330 to 350 dB was audible in blind tests, so tiny signals' amplitude and phase must be perfect or depth changes; analog components show similar errors. MFG Audio Reviews 45:45
• 2024-03-20Watts says every metal-to-metal interface is audible, analog circuits should be kept simple, part choice trades RF noise against dielectric absorption, and assumptions must be checked by listening tests. MFG Audio Reviews 53:45
• 2024-03-20Watts says the best listening tests are where changing an error yields no audible difference, showing the bottom is reached; he mentions tomorrow's talk on objective listening and opens questions. MFG Audio Reviews 54:35
• 2024-05-31Watts says he is not under contract with Chord, a scaler needs physical separation from the DAC, DSD sounds unreal, and subjective impressions can be characterized objectively. Head-Fi, May 2024
• 2024-06-11Watts says the Mojo 2 HF filter's IIR phase change is linear and inaudible, unlike interpolation timing errors, and a listening test against an ADC-project FIR filter confirmed this. Head-Fi, Jun 2024
• 2024-06-24Watts says he does not recall calling the Hugo M Scaler a mess, says it remains essential on his flights, and calls Quartet a major advance on it. Head-Fi, Jun 2024
• 2024-06-24Watts defers Quartet's 4M-tap details to launch, says 5.2 uS at 192kHz is still too big, calls DSD a mess for timing, and describes his reference tracks. Head-Fi, Jun 2024
• 2024-07-12Watts says with finite taps the coefficient recipe matters; the WTA algorithm has five variables, each tuned by listening to minimize timing error as the brain perceives it. HIFICLUB 12:16
• 2024-07-12Watts lists what he listens for (separation, timbre variation, soundstage focus and depth, note starts and stops) with many test tracks, adjusting each variable to an optimum. HIFICLUB 13:09
• 2024-07-12Watts says thousands of listening tests found the WTA variables' sweet spot, which suits everyone's brain, and the work has been more taps plus tuning toward 16-bit performance. HIFICLUB 13:56
• 2024-07-12Watts says theory only gives pointers and his filters are set by sound-quality tests, unlike other upsamplers whose coefficients come from maths; he aims for neither warm nor bright. HIFICLUB 15:30
• 2024-07-12Watts says the original recording is unknowable (microphones, cables, desk, EQ, compression, reverb), so he judges by variation: timbre, instrument separation and focus. HIFICLUB 16:20
• 2024-07-12Watts explains focus as hearing a quiet background instrument when a loud trumpet enters, says good focus reduces listening fatigue, and cites following double bass pitch. HIFICLUB 17:20
• 2024-07-12Watts says testing for variation is objective and needs no knowledge of the recording; more variation means a fundamental improvement, not tailoring a DAC to a system. HIFICLUB 18:05
• 2024-07-12Watts says that to guard against confirmation bias he has his son, a trainee recording engineer, toggle switches and describe what he hears. HIFICLUB 21:10
• 2024-07-12Watts says his son's blind descriptions nearly always match his own for errors at minus 350 dB, then starts a story about consulting on digital power amplifiers for semiconductor firms. HIFICLUB 21:57
• 2024-07-12Watts says the non-audiophile engineer found the minus 180 vs minus 200 dB difference night and day, and his report ended the company's arguments. HIFICLUB 24:09
• 2024-07-12Watts says the company set up a listening room; careful listening tests reveal unmeasurable but important effects, and forty years on he is still finding improvements. HIFICLUB 25:01
• 2024-07-12Watts says he began demoing to audiophiles, first a friend in Singapore, and was anxious because a system in someone else's room always sounds wrong to you. HIFICLUB 51:39
• 2024-09-29Watts says listening tests showed he needed far greater filter coefficient accuracy, so he computed at 40 decimal places with aggressive noise shaping but could not verify performance. The Source AV 6:30
• 2024-09-29Watts says all the listening tests he had done on Quartet had to be rerun. The Source AV 9:43
• 2024-09-29Watts says that with new coefficients and a different calculation routine, the subjectively optimized numbers came out the same, a relief confirming his listening was correct. The Source AV 9:50
• 2024-09-29Watts says that with hardware and metalwork done, he has taken Quartet to a few select audiophiles, including a Singapore friend who owns many Chord products. The Source AV 13:01
• 2024-09-29Watts says the friend keeps a separate apartment for hi-fi, with a main system and a mini system, and that he plugged Quartet into the friend's DAVE. The Source AV 13:49
• 2024-09-29Watts says six years of hundreds of small improvements cost him a sense of scale, so trying a new product in someone else's system shows the real improvement. The Source AV 14:19
• 2024-09-29Watts says that on the friend's system the instruments suddenly sounded like reality, a night-and-day change rather than a small one. The Source AV 15:09
• 2024-09-29Watts says he spent the morning talking, gave a seminar, and has another tomorrow on listening tests and how to objectively quantify sound. The Source AV 24:22
• 2024-09-29Watts says that with recordings of unknown provenance, a change can offset one distortion with another, such as sharpening a too warm, soft system, which goes round in circles. The Source AV 24:37
• 2024-09-29Watts says his aim is to reproduce live unamplified sound, so he needs an objective way of assessing sound quality when the recording's problems are unknown. The Source AV 25:18
• 2024-09-29Watts says they make their own recordings, so with his aural memory he can compare playback at home on DAVEs and M Scalers with the live event. The Source AV 25:34
• 2024-09-29Watts says the key test is variation: a more expressive system is a fundamental improvement and more emotional with the music, not one tweaked warmer or brighter. The Source AV 26:22
• 2024-09-29Watts says that if he calls something good many followers buy it, some without listening, which he says is wrong: you must listen first. The Source AV 29:14
• 2024-10-02Watts says a new unlimited-precision FFT showed his old coefficients were wrong; corrected ones kept the same preferred settings, improved sound, and delayed everything 3 months. Head-Fi, Oct 2024
• 2024-11-22Watts corrects himself: USB sounded brighter on 2Qute from noise floor modulation, not better clocking, as a later test found USB and optical identical; listening tests easily mislead. Head-Fi, Nov 2024
• 2024-12-17Watts introduces his seminar on objectively assessing sound quality, saying listeners often prefer superficially impressive changes that damage musicality, which he defines as emotional involvement with music. Head-Fi, Dec 2024
• 2025-02-17Watts says perception can only be judged by listening tests; the WTA algorithm uses his own windowing formula with five variables, and tweaking them changes the sound. audio-2G 7:28
• 2025-02-17Watts says WTA variables were tuned by listening down to 10 parts per million, that many listeners wrongly hear brightness as accuracy, and that he aims for the original sound. audio-2G 8:13
• 2025-02-17Watts says he makes his own recordings and judges improvements by timbre, pitch, separation, focus and soundstage: greater variation, such as bright instruments sounding brighter, signals a fundamental improvement. audio-2G 8:58
• 2025-02-17Watts says more accurate location, especially greater depth variation with close-miked sounds staying close, signals a fundamental improvement, and that tuning took thousands of listening tests. audio-2G 9:43
• 2025-02-17Watts says different recordings do not need different optimal settings because reconstruction error is fundamental; he uses three or four test recordings per parameter, such as depth. audio-2G 10:31
• 2025-02-17Watts says that for instrument separation and focus he listens to several tracks chosen for that attribute. audio-2G 11:17
• 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 his 10-band alternating half-dB EQ sounds transparent against flat, a standard EQ would be audible, and transparent EQ is a nightmare people misjudge because they do not listen. needs context audio-2G 19:23
• 2025-02-18Watts says the HF filter fails a null test but not perceptually, citing a decimation-filter test showing its sound change comes from less out-of-band noise and noise floor modulation. Head-Fi, Feb 2025
• 2025-05-12Watts says WTA's five variables, optimized over 25 years with thousands of listening tests, were tuned by ear to ten parts per million (0.001%); over half the coefficients are sinc. Audiophile Junkie 10:46
• 2025-05-12Watts says judging listening tests is hard without knowing the original recording, and the key is variation: in space, depth, left-right placement and timbre between bright and rich instruments. Audiophile Junkie 11:34
• 2025-05-12Watts lists further test criteria (tonal colors, low-frequency pitch, dynamics, rhythm and flow, separation and focus, note starts and stops, clarity) that he scored before adjusting the filter. Audiophile Junkie 12:30
• 2025-05-12Watts says he adjusted the coefficients to hit a sweet spot that always existed, sometimes going back to redo other variables, calling it an intense process. Audiophile Junkie 13:15
• 2025-09-14Watts says accurate transients remove a mechanical quality and reveal musicians' timing interplay as instruments are better focused; some hear little difference with the M Scaler, others find it essential. The Source AV 16:52
• 2025-09-14Watts says he has demoed Quartet and the M Scaler where you could barely hear a difference because the system was not transparent. The Source AV 17:36
• 2025-09-14Watts says on other systems the difference is profound, night and day, so it depends on the system's transparency and on the listener's sensitivity. needs context The Source AV 17:48
• 2026-03-17Watts says his first improved noise shaper simulated at 220 dB; switching against 200 dB, separation, focus and timbre sounded identical. ecoustics 36:23
• 2026-03-17Watts says bass pitch was also identical but depth was obviously deeper with the 220 dB noise shaper on an organ recording, which made no sense to him. ecoustics 37:08
• 2026-03-17Watts says blind listeners pick the same depth result, and at the time he suspected an analog proxy effect rather than noise shaper accuracy itself. ecoustics 39:09
• 2026-03-17Watts says he immediately heard the solid-core mains cable sound smoother and warmer, and asked his girlfriend to listen to each and say what differed. ecoustics 54:51
• 2026-03-17Watts says a car battery (12 volts, 300 amps, sub one milliohm) is the perfect supply, and his benchmark is that a DAC on mains should sound like on battery. ecoustics 1:08:48
• 2026-03-17Watts says in listening tests the mains supply sounded better than a 24-volt lithium-ion battery, so he repeated the test carefully, focusing on noise floor modulation. ecoustics 1:11:12
• 2026-03-17Watts says warmer sound and better separation indicate less noise floor modulation and RF; mains won, which he traced to the battery's high RF impedance and poorer cables. ecoustics 1:11:58
• 2026-03-17Watts says sensitivity and budgets differ, some hearing night and day and others a small step, so listen for yourself; each step improves transparency and refinement. ecoustics 1:22:15
• 2026-03-24Watts says Quartet began as a better Hugo M Scaler but became a study of interpolation filters; his unique WTA algorithm was tuned through thousands of listening tests. Audio Force 9:47
• 2026-03-24Watts says each of the WTA equation's five factors had to be set by ear to within ten parts per million, and he did not understand why a million taps was so sensitive. Audio Force 10:45
• 2026-03-24Watts says he wants people convinced by their own listening, and that every DAVE owner who has heard Quartet has been very excited, commenting they knew it would sound a lot better. Audio Force 20:38
• 2026-03-24Watts says he used Quartet in listening tests for the various changes and was amazed. needs context Audio Force 28:45
• 2026-05-03Watts says only carefully structured listening tests show whether an error matters; some measurable things have no audible impact, while other errors stay perceptible however small. Audio Force 3:30
• 2026-05-03Watts says FPGA simulation can show an error at say minus 300 dB, then listening reveals whether reducing it matters; listening is far more important than measurements. Audio Force 5:12
• 2026-10-02Watts says he wants a scientific explanation for any difference he hears, such as between mains cables, dismisses other designers' unscientific explanations as marketing, and wants listening tests done objectively. HiFi Buys 6:35
• 2026-10-02Watts says subjective listening tests must be made as objective, rational and careful as possible because mistakes are easy. HiFi Buys 7:20
• 2026-10-02Watts says listening tests need great care, since warmth and bloom from low-frequency distortion masking hardness and brightness can win a simple test while making things worse. HiFi Buys 42:43
• 2026-10-02Watts says the most important test is emotional, whether music is more involving, which cannot be judged on a simple A/B test and takes several weeks. HiFi Buys 44:25
• 2026-10-02Watts says emotional testing is the ultimate test but rarely works A/B, so he uses test tracks for specific factors such as timbre variation and soundstage width and depth. HiFi Buys 45:32
• 2026-10-02Watts says for instrument separation he listens for whether a quiet background instrument gets swamped when a loud instrument plays. HiFi Buys 46:18
• 2026-10-02Watts says focus is hearing subtle transient details such as a guitarist's plucking, related to but distinct from instrument separation, and linked with clarity. HiFi Buys 46:47
• 2026-10-02Watts says he asks not whether something is better but whether it is more varied than before. HiFi Buys 48:37
• 2026-10-02Watts says if the sound is more varied, you know for certain you have made an improvement. HiFi Buys 48:49
• 2026-10-02Watts says real improvement makes bright instruments brighter and dark ones darker, while noise floor modulation mistaken for transparency makes every instrument brighter, revealing a distortion. HiFi Buys 48:54
• 2026-10-02Watts says a saxophone should sound richer and smoother while bright instruments sound brighter, and distortions can create an artificial, analytical sense of transparency. HiFi Buys 49:43
• 2026-10-02Watts says listeners do not know the original recording or their transducers' quality, a recording is an engineered construct, and listening for variation is how to find reality. HiFi Buys 50:15
• 2026-10-02Watts says he tests assumptions by listening; he added integrators for a 220 dB noise shaper and used a toggle switch because each Xilinx place and route then sounded different. HiFi Buys 1:18:06
• 2026-10-02Watts says newer FPGAs and tools are consistent, but he still keeps the same FPGA code and toggles between the two versions so everything else stays the same. HiFi Buys 1:19:04
• 2026-10-02Watts says the toggle kept the FPGA's power consumption, operating conditions and place and route identical. HiFi Buys 1:19:38
• 2026-10-02Watts says moving from 200 to 220 dB left separation, detail and timbre identical but depth clearly better, which he tested with an organ recording heard in a cathedral. HiFi Buys 1:19:47
• 2026-10-02Watts says intellectually it makes no sense, but under blind conditions other listeners independently report that depth sounds better. HiFi Buys 1:22:59
• 2026-10-02Watts says better timing means better phase accuracy, so he quickly added an adjustable small-signal delay with a toggle switch, without simulating first. HiFi Buys 1:28:26
• 2026-10-02Watts says studies claiming no difference between a modern violin and a Stradivarius actually showed split preference, with half preferring the modern violin's bright sound over the Strad's warmth. HiFi Buys 1:34:09
• 2026-10-02Watts says the studies measured preference, not physical difference, and recounts hearing a blind Strad versus modern violin comparison on BBC Radio Four while driving. HiFi Buys 1:34:53
• 2026-10-02Watts says the radio segment was a blind test and he could immediately tell which was the Strad. HiFi Buys 1:35:37
• 2026-10-02Watts says even on FM radio at 80 mph the Strad was obviously warmer and more expressive, and some people do not perceive such differences because of differing tastes. HiFi Buys 1:35:44
• 2026-10-02Watts says others may be listening for the wrong things or not know what quality unamplified music sounds like. needs context HiFi Buys 1:36:39
• 2026-10-11Watts says raising a variable improved sound until it got worse; each of the five variables had a sweet spot within ten parts per million, audible in careful listening tests. Distorsiuni Interioare 17:52
• 2026-10-11Watts says the key issue is the algorithm; nobody discusses transient reconstruction or tuning filters thousands of times, and he gives customers one listening-optimized solution, not many options. Distorsiuni Interioare 38:03
• 2026-10-14Watts distinguishes listening for emotional enjoyment from listening to make progress, saying emotional testing is poor for judging very small changes; he mentions building his own recordings. Part-Time Audiophile 5:58
• 2026-10-14Watts says recordings have unknown provenance, so an apparent improvement may only balance another problem, e.g. distortion making a dull, warm system seem better. Part-Time Audiophile 6:40
• 2026-10-14Watts says he listens for variation: if a change enlarges the difference between a close-miked voice and a distant object, it is a fundamental improvement. Part-Time Audiophile 7:33
• 2026-10-14Watts lists his criteria: greater timbre variation (warmer sax, brighter trumpet), more accurate starting and stopping of notes, consistent instrument separation, each attribute assessed independently. Part-Time Audiophile 8:21
• 2026-10-14Watts says electronics rarely get better in one attribute and worse in another, unlike transducers; usually other aspects stay the same. Part-Time Audiophile 9:07
• 2026-10-14Watts says reproduced sound loses emotional connection through lack of transparency; after tens of small AB-tested improvements, growing emotional involvement over weeks confirms progress. Part-Time Audiophile 11:46
• 2026-10-14Watts says depth perception is his main test of digital module options; with practice, AB switching reveals very subtle changes, one flat and one deeper. Part-Time Audiophile 22:46
• 2026-10-14Watts says the deeper option is better, provided a dry, close-miked recording stays flat in front, since added reverb could artificially enhance depth. Part-Time Audiophile 23:27
• 2026-10-14Watts says his filters were designed from listening tests, thinking and experiments to remove or minimize the transient timing problem. Part-Time Audiophile 34:01
• 2026-10-14Watts says his ADC's aliasing is vanishingly small; comparing standard, high-end and pulse array ADCs on a female vocal from Daniel's simple analog tape, pitch was the surprise. Part-Time Audiophile 57:48
• 2026-10-17Watts says the female vocalist's pitch reproduction differed greatly: she sounded out of tune on the commercial ADC and in tune on the Quartet ADC. article, Oct 2026
• 2026-10-17Watts says the vocalist's voice and pitch are wrong on the first playback and right on the second, from exactly the same tape. article, Oct 2026
• 2026-10-17Watts says Quartet returns lots, but the key question is the loss from decimating 768k to 48k and back, which needs a bank of test recordings through 2027. article, Oct 2026
• 2026-10-17Watts says he listens for pleasure, has Philip Glass's Koyaanisqatsi on now, and listens as an engineer only during subjective testing. article, Oct 2026
• 2026-10-17Watts says Chord compared a turntable and phono stage direct to a preamp against going through Quartet and DAVE; the digital route sounded a lot better, which puzzled him. article, Oct 2026
• 2026-10-17Watts says two high-frequency filters in cascade sounded no different from one, which is how he knows they are transparent. article, Oct 2026
• 2026-10-17Watts says his Mojo 2 4.4 mm connector listening tests used Quartet and it transformed Mojo 2; it works but would not be a sensible pairing. article, Oct 2026
• 2026-10-17Watts says he uses the EQ a lot; a ten-band test alternating plus and minus half dB gave no loss in transparency, detail, depth, separation or focus. article, Oct 2026