Transient timing error from reconstruction

Reconstruction filters and timing

Answers on this topic

• How can upscaling to 705.6/768 kHz improve sound if the recording has nothing above 22 kHz?
• What does Rob mean by 'starting and stopping of notes', timbre variation and 'sense of flow', and what causes them?
• What does Rob think of Class D amplifiers, and how do his pulse array power amps differ from them?
• What is a transient timing error, and why is it audible when a fixed delay or phase shift is not?
• Why does tap length matter, and why a million taps?

Everything he said on this topic (164)

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

• 2014-04-17Watts explains that bandlimited samples are perfectly reconstructed by an infinite-tap filter, including transients starting mid-sample, and timing precision falls off as taps are reduced. Head-Fi, Apr 2014
• 2014-06-21Watts distinguishes pitch resolution from the 4uS inter-aural delay, and argues sample rates matter for transient timing, so a good interpolation filter makes CD sufficient. Head-Fi, Jun 2014
• 2014-07-03Watts explains that an infinite-tap filter perfectly reconstructs bandlimited signals while finite taps cause transient timing errors, hundreds of uS for one tap, so he lengthened taps by listening. Head-Fi, Jul 2014
• 2015-03-01Watts says short filters cause transient timing errors of tens of microseconds; Hugo uses 26,000 taps, 164,000 still beats 100,000, and about a million taps would give 16-bit accuracy. Head-Fi, Mar 2015
• 2015-05-20Watts says digital engineers underestimated the importance of reconstructing the original continuous signal with timing accuracy, and that improving time resolution proved a big benefit. HIFICLUBE.NET JVH 2:12
• 2015-06-18Watts says half-band filters time transients poorly (-6dB at FS/2); 16-bit sinc accuracy needs about 1M taps; Hugo is about 12 bits, Dave about 14. Head-Fi, Jun 2015
• 2015-09-25Watts argues DSD is flawed by its format: limited noise-shaper resolution loses small signals and depth, and level-dependent timing non-linearity softens transients and compresses timbre. Head-Fi, Sep 2015
• 2015-11-10Watts says Dave's key gain is timing: WTA filtering to 256FS reconstructs timing to 88 nS, improving starting and stopping of notes, while depth comes from the noise shapers. Head-Fi, Nov 2015
• 2015-11-22Watts recounts developing Dave: smoothness from reduced noise floor modulation, Hugo-derived timing gains from 256FS WTA (88nS), and noise shapers raised from 200 to 350 dB for depth. Head-Fi, Nov 2015
• 2016-01-29Watts says a single-stage 16FS WTA first stage sounded much better, by more than 26,000 versus 164,000 taps, and 256FS second-stage filtering improved timing accuracy from microseconds to nanoseconds. article, Jan 2016
• 2016-03-07Watts is quoted: developing Hugo in 2013 he thought 1 microsecond accuracy was enough, but developing DAVE he confirmed nanosecond-level accuracy has an effect. article, Mar 2016
• 2016-03-07Watts is quoted that simple filters in DAC ICs give timing errors around 100 microseconds on a transient burst; he says more FIR taps and processing accuracy are needed. article, Mar 2016
• 2016-05-17Watts explains DSD's and PWM's switching, jitter and amplitude-timing errors, how pulse array balances switching to avoid them, and why DAVE needs 350 dB noise shapers for depth. Head-Fi, May 2016
• 2016-07-24Watts says this mattered more than going from 26 thousand to 164 thousand taps, 16x filtering is accurate only to microseconds, and he added a 16x-to-256x WTA stage. Hifistatement 7:47
• 2016-08-19Watts hypothesizes that brain break-in, first noticed with Hugo, is the brain unlearning coping strategies for sampled-data timing errors, which DAVE showed must be accurate to nanoseconds. Head-Fi, Aug 2016
• 2016-08-19Watts separates input jitter, removed by the DPLL and local clock, from transient reconstruction timing, where a conventional filter can misplace a zero crossing by thousands of nanoseconds. Head-Fi, Aug 2016
• 2016-10-16Watts argues musicality comes first and explains why Mojo and DAVE excel on transient timing (NOS errors, tap length, 1M taps), small-signal accuracy and noise floor modulation. Head-Fi, Oct 2016
• 2017-01-09Watts explains Dave's 16FS-to-256FS WTA stage replacing IIR filters, says any timing error is audible, notes Hugo 2 filter settings switch that stage, and contrasts 88 nS with 2.6 uS. Head-Fi, Jan 2017
• 2017-01-10Watts contrasts a one-tap NOS filter's roughly 100 uS worst-case timing error with an ideal infinite filter, plans Davina round-trip files, and links 1M-tap 16FS filters to Fourier uncertainty. Head-Fi, Jan 2017
• 2017-01-23Watts says 16-bit coefficient accuracy has no bearing on recording bit depth: timing errors arise in interpolation, and an ideal sinc filter with 16-bit values needs over 1M taps. Head-Fi, Jan 2017
• 2017-01-28Watts explains peak error, why 16-bit accuracy needs about 1M taps, why doubling sample rate only halves conventional filters' error, and that taps are not the whole story. Head-Fi, Jan 2017
• 2017-02-13Watts defends his Mojo timing illustration as a simplification, targets non-linear timing uncertainty rather than ringing, says doubling taps halves the error, and notes WTA accommodates half-band decimation. Head-Fi, Feb 2017
• 2017-02-13Watts says interpolated transient timing depends on sample timing unless the filter is ideal, and plans a Davina null test with half-sample-delayed white noise to measure the error. Head-Fi, Feb 2017
• 2017-05-21Watts says down-sampling creates transient timing errors through aliasing, but up-sampling creates many times more, as reconstruction is much harder than decimation. Head-Fi, May 2017
• 2017-08-26Watts says audible timing errors are non-linear, not fixed delays; the M scaler's 1,015,808 taps match ideal sinc to 16 bits, Hugo 2 about 13; doubling taps adds a bit. Head-Fi, Aug 2017
• 2017-10-13Watts says nobody hears above 20 kHz and he cannot hear above 15 kHz, but very small errors in reconstructing transient timing profoundly affect music perception. miceblue425 9:42
• 2017-10-13Watts says he once thought transients needed 4 microsecond accuracy but now puts it at 100 nanoseconds or less, and that conventional DAC interpolation filters reconstruct timing poorly. miceblue425 10:26
• 2017-10-13Watts says a DAC should reproduce the analog signal inside the ADC, not bit-perfect data, and introduces an illustration of a sampled 20 kHz sine burst. miceblue425 11:08
• 2017-10-13Watts says a minimum-phase or 'IR' filter, any filter with no pre-ringing, gives 100 microseconds of timing error and loses the transient's start, while an infinite sinc filter reconstructs it perfectly. miceblue425 12:06
• 2017-10-20Watts says sampling theory needs an infinite-tap sinc filter, that limited tap length gets reconstructed timing wrong, and that he calculated getting on a million taps for guaranteed 16-bit performance. Hifistatement 44
• 2017-10-31Watts says non-sinc interpolation reconstructs transient timing non-linearly, so frequency-domain reasoning fails; he likens the error to jitter but says it exceeds 100 nS. Head-Fi, Oct 2017
• 2017-12-11Watts says hi-res files need the same tap length, since doubling the sample rate cuts timing error area fourfold (2 bits) while the M scaler gives about a 16-bit improvement. Head-Fi, Dec 2017
• 2017-12-18Watts calls NOS DACs an unmitigated disaster for transparency, with massive transient timing errors and out-of-band images, and says they sound soft, flat and out of focus. Head-Fi, Dec 2017
• 2017-12-18Watts says sampling images cause transient timing errors and their out-of-band noise overloads analog stages causing noise floor modulation, which none of his DACs from Mojo to DAVE measurably show. Head-Fi, Dec 2017
• 2018-02-23Watts says he thinks in terms of recovering the original bandlimited analog signal rather than data, that the transcript's garbled subject does this poorly, and that the brain is very sensitive to non-linear timing errors. MajorHiFi 3:26
• 2018-02-23Watts explains that a regular DAC places a transient slightly early or late depending on previous data and sample position, whereas recovery should be identical wherever sampling occurred. MajorHiFi 4:11
• 2018-02-23Watts says studying sinc functions and hearing physiology led him to conclude that without a sinc filter timing errors relative to 22 microseconds made digital sound bad, among other reasons. MajorHiFi 13:47
• 2018-03-16Watts attributes reduced loudest-instrument dominance to no noise floor modulation and the WTA filter, says an M scaler enhances it, and calls dominance a major cause of fatigue. Head-Fi, Mar 2018
• 2018-07-21Watts says CD samples are 22 microseconds apart, the brain is very sensitive to timing errors, and digital reconstructs transients slightly early or late depending on sampling and surrounding music. c w 8:27
• 2018-07-21Watts says the interpolation filter recovers the waveform, but conventional DACs calculate it poorly, so transients shift constantly between too early and too late, confusing the brain. c w 9:12
• 2018-07-21Watts illustrates with a sampled 20 kHz sine-wave burst, saying a simple conventional filter such as a NOS, IR or MQA filter produces a different waveform from the original. c w 9:57
• 2018-07-21Watts says a filter allowing no pre-ringing loses the original pulse, an error of about 100 microseconds, while a sinc function filter perfectly reconstructs the bandlimited transient. c w 10:54
• 2018-07-21Watts sums up: high-end DACs use a few hundred taps, the M Scaler has 10,000 times more processing, recovering 16-bit CD to 16-bit accuracy and transient timing to unprecedented accuracy. c w 27:04
• 2018-07-26Watts posts his CanJam presentation: transients cue placement, pitch and timbre; conventional filters mistime them; 1M WTA taps match sinc to 16.6 bits using 528 DSP cores. Head-Fi, Jul 2018
• 2018-08-02Watts says Blu 2 and M scaler filter code is identical, explains that upscaling recovers transient timing rather than adding information, and why pulse array needs DSD converted to PCM. Head-Fi, Aug 2018
• 2018-08-04Watts says the issue is non-linear transient timing, the master clock and source jitter are unimportant thanks to the DPLL, and Hugo 2 gets full 1M taps via 3.5mm-to-BNC adaptors. Head-Fi, Aug 2018
• 2018-08-31Watts says the original is the unsampled analog signal; the M scaler's million taps match ideal sinc to better than 16 bits, recovering transient timing regular filters get wrong. Head-Fi, Aug 2018
• 2018-09-24Watts says any varying transient timing error is audible however small, citing audible gains from half a million to one million taps and 350 dB over 330 dB noise shapers. Head-Fi, Sep 2018
• 2018-10-23Watts explains the M scaler matches sinc to 16.7 bits, fixing transient timing, but cannot repair aliasing from modern half-band ADCs, possibly why 1960s recordings sound amazing. Head-Fi, Oct 2018
• 2018-11-11Watts explains the M scaler's apparently slower tempo as reduced dominance effect, because 1M-tap WTA cuts program-dependent transient timing errors; 164,000 to 1M taps is very audible. Head-Fi, Nov 2018
• 2018-11-28Watts says a random timing error would need 24 bits beyond 1M taps, but it is program-correlated so matters more than dither; video mode uses an asymmetric filter. Head-Fi, Nov 2018
• 2018-12-13Watts says a regular DAC reproduces transients slightly early or late, the uncertainty shifting with past and future signal, and that such timing errors have big subjective impact. Pursuit Perfect System 3:55
• 2018-12-13Watts says modulated transient timing confuses the brain, degrading bass, timbre, pitch, separation, focus and spatial location, and blames limited processing before starting a tone-burst illustration. Pursuit Perfect System 4:38
• 2018-12-13Watts illustrates sampling a 20 kHz tone burst and says simple filters such as NOS, IIR or MQA filters reproduce the digital data without the original transient shape. Pursuit Perfect System 5:23
• 2018-12-13Watts attributes the M Scaler's stronger instrument power without cross-modulation to removing the timing modulation that past and future signal causes in conventional filters, restoring 16-bit accuracy. Pursuit Perfect System 16:26
• 2019-01-29Watts says finite filters trade oversampling rate against data, that 4 uS interaural resolution proved insufficient, and that high rates matter for transient timing accuracy, not hearing 768k. Head-Fi, Jan 2019
• 2019-03-04Watts explains white/green adds a WTA 16FS-to-256FS stage replacing an IIR filter, why sinc accuracy needs long taps, and says preferring orange suggests an overly bright system. Head-Fi, Mar 2019
• 2019-03-29Watts says TT2 filters are interpolation tweaks not EQ, argues NOS DACs are not truly bit perfect, explains transient timing errors, and warns IIR EQ and dithered volume are not transparent. Head-Fi, Mar 2019
• 2019-04-03Watts says his 768k recording resembles M scaled 44.1k, doubling the rate cuts the error area fourfold, and R2R or 2.8 MHz noise-shaped DACs limit accuracy above 768k. Head-Fi, Apr 2019
• 2019-04-10Watts attacks the industry's 'ringing is bad' meme, discusses ultrasonic intermodulation, distinguishes EQ pre-ringing from reconstruction, and calls transient timing the real problem. Head-Fi, Apr 2019
• 2019-05-02Watts gives the M scaler's 65 C thermal trip, says it recovers transient timing to 16-bit accuracy rather than adding information, and attributes optical's darker sound to RF, not jitter. Head-Fi, May 2019
• 2019-06-13Watts says 96 kHz releases are better but not enormously; doubling sample rate only quarters the timing error, and M scaled CD beats 192k; Davina will test this. Head-Fi, Jun 2019
• 2019-06-14Watts says Davina hit hardware and decimation-filter problems, argues audible timing errors are non-linear ones, and cites WTA 2 tests as evidence of sensitivity below the 7-4 uS interaural limit. Head-Fi, Jun 2019
• 2019-06-14Watts says timing errors cannot be recovered once introduced, and in DSD small-amplitude steps are delayed longer through the modulator than large-signal steps. Head-Fi, Jun 2019
• 2019-06-15Watts says 96k beats 48k mainly because cheap half-band ADC decimation filters alias, that SRC to 44.1k adds distortion, and that only signal-dependent timing errors matter. Head-Fi, Jun 2019
• 2019-06-16Watts shows by a zero-crossing example why higher rates improve transient timing, says 705/768 kHz is needed, and says hi-res ADC noise causes noise floor modulation the HF filter reduces. Head-Fi, Jun 2019
• 2019-07-15Watts says NOS DACs do oversample, with a zero-order hold, the crudest filter possible, whose large timing and aliasing errors make music sound soft, flat and out of focus. Head-Fi, Jul 2019
• 2019-07-22Watts says the M scaler converts 48k to 768k with 16-bit sinc accuracy, each doubling of rate quarters the transient error area, and WTA2 timing gains above 768k remain audible. Head-Fi, Jul 2019
• 2019-09-09Watts says M scaler accuracy depends on input, far beyond 24 bits on sine waves, at least 16.7 bits on transients; DAVE 2 is unplanned and DAVE took 2.5 years. Head-Fi, Sep 2019
• 2019-12-19Watts says without a sinc function transients come out slightly early or late, constantly modulating with the filter design and with past and future music. Passion for Sound 9:42
• 2019-12-19Watts says the job is reconstructing between samples; a NOS filter is a crude first-order hold with huge transient errors, e.g. losing the first pulse of a 20 kHz burst. Passion for Sound 17:48
• 2019-12-19Watts says transient errors can be very large in amplitude; a real DAC places a zero crossing slightly early or late, a constantly varying timing error troubling the brain. Passion for Sound 18:32
• 2019-12-19Watts says 16-bit Red Book arguably needs timing reconstructed only to 16 bits, and that 16.7-bit filter accuracy is not DAC accuracy: a steady 24-bit sine wave comes out 24-bit. Passion for Sound 20:18
• 2019-12-19Watts says standard measurements use steady-state signals without transients, so reconstruction accuracy only matters when transients are involved. Passion for Sound 20:54
• 2020-01-02Watts says his earlier three-stage IIR filter measured very like the WTA FIR stage, yet the change is audible, so even 88 nanosecond timing resolution matters. Passion for Sound 4:23
• 2020-01-02Watts says we perceive better than 1.3 microseconds, cannot yet set a required accuracy, and suspects any transient timing uncertainty, even 10 nanoseconds, is audible. Passion for Sound 5:03
• 2020-01-02Watts says this is another reason for a noise shaper running at 9.6 nanoseconds; his outputs can reproduce a 50 MHz sine wave, improving transient reconstruction. Passion for Sound 5:50
• 2020-01-02Watts says that once a filter stops looking into the future, you get huge transient errors. Passion for Sound 18:46
• 2020-03-27Watts says the M scaler gives a natural sense of flow and less fatigue because loud instruments no longer modulate transient timing, about 10 times more accurate than DAVE alone. Head-Fi, Mar 2020
• 2020-06-25Watts says the benefit of 192 kHz is that it reduces the timing problem. Part-Time Audiophile 39:31
• 2020-06-25Watts says doubling the sample rate quarters the timing error, only 2 bits, while the M Scaler matches a sinc function to better than 16 bits to reconstruct timing. Part-Time Audiophile 39:38
• 2021-03-29Watts says Linn-style PWM FPGA DACs have amplitude-dependent transient timing errors and a roughly 3MHz noise shaper, unable to reach the 350dB accuracy of his pulse array. Head-Fi, Mar 2021
• 2021-09-09Watts says WTA benefits shrink at higher frequencies because transient error area falls fourfold per sample-rate doubling; DAC and pre-amp modes perform identically; studio ADCs will have phantom power. Head-Fi, Sep 2021
• 2021-09-09Watts says a 768kHz signal has 256 times less transient error area than 48kHz, why higher rates sound better, and his WTA tuning found one sweet spot across recordings. Head-Fi, Sep 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
• 2021-11-30Watts says correct filter timing errors are amplitude independent but transients intermodulate through non-sinc windows; a WTA simulation matched 0dBFS and -301dB, and coefficient accuracy improved depth. Head-Fi, Nov 2021
• 2022-01-06Watts says the M scaler is 1M taps at 705/768k so only half the taps are used at 352/384, and that tape bias and amplifiers also degrade transient timing. Head-Fi, Jan 2022
• 2022-01-31Watts ties transient timing to the WTA filter: 1 million taps in M Scaler, 40 thousand in Mojo 2 using forty DSP cores per channel; samples are 22 microseconds apart. Darko.Audio 9:24
• 2022-01-31Watts says conventional filters misjudge what happens between samples, so transients arrive early or late, modulated by past and future instruments, blurring pitch, timbre and separation. Darko.Audio 10:12
• 2022-03-19Watts says transient timing is digital audio's Achilles heel: CD samples are 22 microseconds apart and the interpolation filter must locate transients between them, which conventional DACs do poorly. MFG Audio Reviews 16:58
• 2022-03-19Watts says transients come out early or late depending on past and future signal, so instruments modulate each other's timing and the brain cannot separate them. MFG Audio Reviews 17:42
• 2022-03-19Watts illustrates with a sampled sine-wave burst and begins to show what conventional, IIR, NOS, MQA and minimum phase filters do to it. MFG Audio Reviews 18:09
• 2022-03-19Watts says those filters lose the initial transient, a 100 microsecond timing error, so they sound soft, which some like because it hides the DAC's noise floor modulation. needs context MFG Audio Reviews 18:55
• 2022-03-19Watts says each doubling of sample rate cuts the error area fourfold, 256 times less at 705 kHz, the 16FS sweet spot, and WTA2 is less tap-length sensitive than WTA1. MFG Audio Reviews 52:02
• 2022-04-26Watts says typical 64 or 128 tap filters make transients constantly shift early or late; Mojo 2's WTA uses 40 DSP cores per channel and over 40,000 taps. Audio Force 10:44
• 2022-06-22Watts says a NOS filter is the crudest oversampling filter, real filters degrade transient timing, and PWM in Class D has amplitude-dependent timing plus filter phase intermodulation. Head-Fi, Jun 2022
• 2022-07-21Watts says non-supplied PSUs void Chord's warranty, linear supplies lack 100 kHz to 10 GHz RF filtering, and transient timing reconstruction differs from transient response, which only listening tests reveal. Head-Fi, Jul 2022
• 2022-07-25Watts recounts DAVE's nearly 3-year development, questions Stereophile's 16-bit measurements, cites -127dB A-weighted, calls SINAD worthless, and says filters must reconstruct transient timing. Head-Fi, Jul 2022
• 2022-08-19Watts says transient timing is crucial to perceiving pitch, timbre, location and separation, and digital's problem is reconstructing transients with correct timing, not information above 20 kHz. AVstore 2:28
• 2022-08-19Watts says no interpolation filter is perfect, the M Scaler included: past and future music modulates each transient leaving the filter, making it slightly early or late. AVstore 31:45
• 2022-08-19Watts says timbre perception depends on starting transients, citing tests where piano and trumpet are hard to identify without them, so transients modulated by other instruments degrade timbre. AVstore 32:29
• 2022-08-19Watts says this muddling of timbre variation is largely due to inaccurate transient reconstruction, and that the loudest instrument dominates so quieter ones are hard to follow. AVstore 34:09
• 2022-08-19Watts says inside the filter the trumpet's and sax's transients modulate one another, degrading perception of each instrument, so transient timing reconstruction matters for separation. AVstore 35:11
• 2022-08-29Watts says filters make transient timing errors program dependent, which is why 100k and 1M taps sound different, that audibility does not fade with age, and that prototypes are delayed. Head-Fi, Aug 2022
• 2023-02-11Watts says improvements are easiest to hear on poor recordings, first noticed with Hugo, and pre-1968 master tapes lose their realism through ADCs and conventional DACs' transient timing errors. Head-Fi, Feb 2023
• 2023-03-28Watts says R2R DACs can only run at 16 times oversampling, not 104 MHz, limiting transient reconstruction to microseconds rather than the nanoseconds pulse array achieves. The Headphone Show 11:03
• 2023-03-28Watts says PWM has a fatal flaw: pulse edges shift earlier as the signal goes more positive, so transient timing moves with signal level. The Headphone Show 31:11
• 2023-03-28Watts says PWM's timing shift is a major transient reconstruction error that analog filtering cannot fix, and the brain uses transient timing for timbre, separation and bass pitch. The Headphone Show 32:01
• 2023-03-28Watts says 250 kHz distortion is audible not because we hear 250 kHz but because high-frequency distortion changes transient timing. The Headphone Show 42:19
• 2023-03-28Watts says simulating transients at 10 nanoseconds shows no change in transient performance, so the DSD small-signal transient problem is absent. needs context The Headphone Show 48:07
• 2023-11-20Watts says square waves contain no transients, so a perfect DAC reproduces a bandwidth-limited square wave with Gibbs phenomena and no reconstruction timing error. Head-Fi, Nov 2023
• 2024-01-07Watts says 16FS is the interpolation sweet spot (1.3uS vs 2.6uS timing at 8FS for 48kHz), adding more 200T FPGAs is the economic path, and a battery laptop via optical is hard to beat. Head-Fi, Jan 2024
• 2024-05-10Watts says it is not about bandwidth but about timing accuracy: getting transients to be reconstructed. needs context Reactcore 24
• 2024-05-16Watts says every DAC converts samples to a continuous waveform to recover transient timing, and an infinite sinc filter with infinite processing would reconstruct it perfectly. ConvinceMeAudio 6:22
• 2024-05-16Watts says limited taps make transients shift early or late depending on surrounding transients, so the brain loses separation, timbre, location and the starting and stopping of notes. ConvinceMeAudio 6:51
• 2024-05-16Watts says reconstructed transients come slightly early or late, constantly shifting, which confuses the brain, since it uses transient timing for timbre, pitch, separation and placement. The Speaker Shack 1:10
• 2024-05-17Watts says samples are taken 22 microseconds apart and transient timing between them must be reconstructed; otherwise transients arrive early or late, past and future ones disturb the present, confusing the brain. AVstore 2:57
• 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
• 2024-06-18Watts explains how interpolation modulates transient timing, harming separation, timbre, bass and fatigue, says any timing error is perceptible, and lists Quartet's five FPGAs, 2,500 components, 75-ampere supply and two million code lines. article, 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
• 2024-06-26Watts says DSD256 has -20dB noise at 400kHz, all DSD rates show large modulator noise floor modulation, the brain perceives any timing change, and recordings stay limited to 192k. Head-Fi, Jun 2024
• 2024-07-12Watts says any interpolation filter that is not a sinc loses transient timing, his 1980s realization that digital would always be flawed. HIFICLUB 9:04
• 2024-07-12Watts says NOS filters sound soft and warm because their transient errors reach 100 microseconds; better transient reconstruction sounds sharper and more focused, like a camera lens. HIFICLUB 29:50
• 2024-07-12Watts names three main issues for ideal conversion: transients (interpolation filters, running at very high rates), small-signal accuracy (minus 301 dB), and a third. HIFICLUB 31:17
• 2024-08-01Watts says transients also set low-frequency pitch; ordinary interpolation filters shift transient timing back and forth, degrading separation, timbre and soundstage, which FPGA power and his WTA filter improve. Audiophilesdoteu 4:40
• 2024-09-29Watts says Quartet reconstructs transient timing more accurately, aiding timbre and instrument separation; digital mis-times transients, only the M Scaler approach recovers them, and it put DAVE in a new game. The Source AV 15:46
• 2025-01-13Watts compares feed-forward correction (about 20dB) with his second-order analog noise shaper (90dB), showing TT2 into 8 ohms, and links crossover distortion to amplitude-dependent transient delay. Head-Fi, Jan 2025
• 2025-02-17Watts says the brain calculates fundamental pitch from transient timing, and an interpolation filter degrades that timing because CD samples are 22 microseconds apart and reconstruction is never accurate enough. audio-2G 5:28
• 2025-02-17Watts says filter errors make transients early or late, modulated by other instruments, confusing the brain so separation, timbral variation, location, bass pitch and note starts and stops suffer. audio-2G 6:13
• 2025-05-12Watts says digital audio's problem is transient timing: pitch, soundstage, timbre and note starts and stops come from transients, and the DAC must recover the waveform between CD samples. Audiophile Junkie 3:35
• 2025-05-12Watts says very small timing errors substantially affect listening, and conventional DACs reconstruct transient timing poorly, making the sound muddled and confused and causing listening fatigue. Audiophile Junkie 4:18
• 2025-05-12Watts describes a slide of a 20 kHz burst signal whose sampled data builds up from zero, and asks what happens when it passes through an interpolation filter. Audiophile Junkie 4:40
• 2025-05-12Watts says a simple interpolation filter gives a transient error of around 100 microseconds, as do NOS, IR, minimum phase and MQA filters; an infinite sinc filter reconstructs perfectly. Audiophile Junkie 5:21
• 2025-05-12Watts says a pure sinc filter jumps abruptly from zero to full sinc at its start, and that abrupt change itself creates audible transient timing errors. Audiophile Junkie 8:55
• 2026-03-17Watts explains that an interpolation filter places a transient between samples slightly early or late, modulating back and forth and confusing location, timbre and pitch; he guessed 16-bit accuracy sufficed. ecoustics 4:05
• 2026-03-17Watts says 16-bit reconstruction accuracy leaves an error below 16 bits, but he later found that even the tiniest timing reconstruction error has audible consequences. ecoustics 4:49
• 2026-03-17Watts says a regular interpolation filter distorts a CD's information content, while his aim is to preserve it fully; the key information is transient timing embedded in the samples. ecoustics 15:16
• 2026-03-17Watts says NOS filters worsen timing errors so transients blur and the sound turns soft and warm, which can mask a hard, bright DAC with noise floor modulation. ecoustics 22:56
• 2026-03-24Watts says the brain uses transient timing to locate instruments, judge timbre, separate instruments and perceive pitch, and the interpolation filter reconstructs the waveform from samples every 22 microseconds. Audio Force 11:37
• 2026-03-24Watts says conventional interpolation filters in high-end DACs always get transient timing slightly early or late, and the WTA filter reduces that timing uncertainty. Audio Force 12:35
• 2026-03-24Watts says he found two different factors change transient timing, and with a standard interpolation filter improving one degrades the other. Audio Force 13:07
• 2026-03-24Watts says Blackbird WTA optimizes both factors independently, giving a 10 times improvement in sound quality, the primary difference between Quartet and his earlier work. Audio Force 14:13
• 2026-05-03Watts says sampled digital audio, with samples 22 microseconds apart on CD, reproduces transient timing poorly; the interpolation filter reconstructs it, and he has spent 30 years developing WTA. Audio Force 6:49
• 2026-10-02Watts says digital audio needs an infinitely long sinc function, and anything short of it degrades transient timing, which he saw clearly in 1981. HiFi Buys 20:21
• 2026-10-02Watts says interpolation filters make transients slightly early or late, modulated by past and future transients, and this blurring upsets the brain; this is where the WTA comes in. HiFi Buys 26:47
• 2026-10-02Watts calls NOS a first-order hold with the biggest transient timing errors, so it sounds soft and warm, offsetting the hardness of high-frequency distortion. HiFi Buys 42:01
• 2026-10-02Watts says transient timing is one issue: with 768 kilohertz output, transients resolve to only 1.3 microseconds, yet recent biological papers suggest the brain has no limit to timing accuracy. HiFi Buys 1:14:27
• 2026-10-02Watts says these stages aim to reconstruct timing resolution down to within 9.6 nanoseconds. HiFi Buys 1:15:58
• 2026-10-02Watts says they should be clear about what transient timing means. HiFi Buys 1:58:05
• 2026-10-02Watts says he is not bothered about phase, since a linear phase shift is in his experience inaudible. HiFi Buys 1:58:09
• 2026-10-02Watts says what matters is phase or transient timing that shifts with time, amplitude or the music; timing that is consistently wrong would not matter. HiFi Buys 1:58:18
• 2026-10-02Watts compares it to loudspeakers randomly pumping backward and forward, which would look strange and sound awful. HiFi Buys 1:58:59
• 2026-10-11Watts says after a couple of years of research he found two different factors degrading transient reconstruction accuracy. Distorsiuni Interioare 18:37
• 2026-10-11Watts says FFT uncertainty is a frequency resolution limit set by batch size, which translates into a transient timing error, and buffer sizes are not disclosed. Distorsiuni Interioare 35:49
• 2026-10-11Watts says aliasing matters because out-of-band signal-correlated noise degrades transient timing, and Quartet's DSD decimation follows its ADC decimation process. Distorsiuni Interioare 56:10
• 2026-10-11Watts says DSD's distortion near full modulation degrades transients; near-perfect decimation has no downside, but DSD conversion itself creates transient timing errors. Distorsiuni Interioare 57:50
• 2026-10-11Watts says the improvement comes from Quartet's 18 kHz HF filter transparently removing out-of-band noise and signal-correlated distortion, improving transient timing; he had expected digitization to sound worse. Distorsiuni Interioare 1:14:52
• 2026-10-14Watts says low tap count interpolation filters reconstruct transients slightly too early or too late. Part-Time Audiophile 27:59
• 2026-10-14Watts says transient timing modulates with past and future signal and the filter structure, while an infinitely long sinc filter would leave no timing uncertainty. Part-Time Audiophile 28:11
• 2026-10-14Watts says a perfect 768 kHz filter would improve a chip DAC but not match his FPGA DACs, which filter up to 104 MHz, leaving 9.6 nanoseconds of uncertainty. Part-Time Audiophile 35:47
• 2026-10-14Watts says the NOS filter used by R2R DACs is an oversampling first order hold filter, the worst for transient timing, making sound artificially soft, smooth and warm. Part-Time Audiophile 39:00
• 2026-10-17Watts says two contrary effects degrade transient timing in the WTA filter and in all interpolation filters, so improving one worsens the other in a single-stage filter. article, Oct 2026
• 2026-10-17Watts says Quartet's 18 kHz filter in ADC decimation removes everything above 20 kHz, and that out-of-band noise and distortion are audible because they corrupt transient timing. article, Oct 2026
• 2026-10-17Watts says companies claiming more taps use FFT convolution, whose Fourier frequency resolution degrades transient timing, and he is skeptical because FFT sizes are not published. article, Oct 2026