Latency, video and surround

Features, settings and controls

Answers on this topic

Everything he said on this topic (75)

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

2014-06-02Watts calculates Hugo's delay: 26,368 taps over 8x oversampling gives 1,648 samples group delay, 34 mS at 48 kHz, so home theater users may need to reduce video delay. Head-Fi, Jun 2014
2014-06-02Watts says bit depth does not affect delay but doubling sample rate quarters it, since the first stage always upsamples to 8FS. Head-Fi, Jun 2014
2015-03-01Watts says Hugo's WTA filter has 38 ms latency, so interpolation draws on 76 ms of data, and the 164,000-tap filter's latency is much longer. Head-Fi, Mar 2015
2015-06-18Watts says DAVE has an asymmetric WTA filter setting with lower delay for video use, and that DAVE production is scheduled for October 2015. Head-Fi, Jun 2015
2015-10-06Watts says Dave uses one FPGA with ten times Hugo's capacity, drawing 5A, is still a work in progress, and has PCM+ and DSD+ programs, DSD+ offering lower latency. Head-Fi, Oct 2015
2016-08-11Watts says he always uses direct-form FIR filters, folded when fewer multipliers are needed, that FPGAs are SRAM-limited, and that WTA group delay cannot be removed without hurting sound quality. Head-Fi, Aug 2016
2017-01-09Watts says Blu's video mode cuts latency from 0.66 seconds to 100 mS, and a -0.56 second JRiver delay gives him lip sync with the full 1M taps. Head-Fi, Jan 2017
2017-01-09Watts confirms a poster is correct about the M scaler and Dave delay. needs context Head-Fi, Jan 2017
2017-07-10Watts says open loop output impedance governs load response, now +/- 0.1 dB at 8 ohms and +/- 0.05 dB unloaded; measuring must allow for 34 mS WTA latency. Head-Fi, Jul 2017
2017-08-26Watts says Blu 2's low-latency video option cuts delay from 630 mS to 104 mS via an asymmetric filter but degrades sound, so he sets -630 mS in JRiver. Head-Fi, Aug 2017
2017-09-01Watts says AAC is decoded by the source, but Dolby and DTS data sound bad; he leaves them unmuted so users know to set the source to stereo PCM. Head-Fi, Sep 2017
2017-10-11Watts says WTA reconstructs transients regardless of origin; DAVE's 164,000-tap delay is about 0.2 s and 1M taps 1.4 s, and his RT60 explanation for depth is speculative. Head-Fi, Oct 2017
2017-10-13Watts suspects the soundstage opens because the filter now spans 1.4 seconds, about a hall's RT60 reverberation decay time, and reverberation is an important depth cue. miceblue425 34:01
2017-10-13Watts says a million taps gives 16-bit sinc reconstruction; 24-bit would need 256 million taps and minutes of delay, which will never happen and which he doubts is audible. miceblue425 40:07
2017-10-13Watts says the latency is actually 0.63 seconds; the filter looks at 1.4 seconds of data, or 0.7 seconds of latency. miceblue425 57:05
2017-10-13Watts says it runs at 16FS: about 0.7 s latency at 44.1 and 0.63 s at 48 kHz; he rips Blu-rays and sets a negative JRiver delay for sync. miceblue425 57:24
2017-10-17Watts gives the filter delay formula; for the 1,015,808-tap Blu at 48 kHz the delay is 0.661 s while the filter processes 1.323 s of music. Head-Fi, Oct 2017
2017-10-20Watts says the USB video mode, two-thirds of a million taps on the same architecture, collapses the soundstage, though architecture and small-signal work carried over from DAVE also matter. Hifistatement 2:57
2018-03-06Watts says guaranteed 16-bit reconstruction of the original analog signal needs a sinc accurate to 16 bits, meaning a delay over 0.5S, whatever the DAC architecture. Head-Fi, Mar 2018
2018-03-13Watts says Blu 2 already has a video latency mode, hints at new DAC developments in years, and aims to remove digital and mains cable sensitivity; analog cables only partly. Head-Fi, Mar 2018
2018-03-20Watts says complex USB gear adds RF brightness, optical is the benchmark, and gives WTA overload margin, M scaler latency (630 ms, video mode 104 mS) and 16.6-bit accuracy. Head-Fi, Mar 2018
2018-03-22Watts explains the 2.75dB WTA overload margin, delays audio in JRiver to keep 1M taps with video, describes visible noise floor modulation, and gives DAVE and Hugo 2 level limits. Head-Fi, Mar 2018
2018-03-28Watts says his DPLL keeps buffer latency equal across asynchronous units, with maximum 5 uS drift and about 1 uS typical, and all current DACs share the DPLL code. Head-Fi, Mar 2018
2018-04-08Watts says the DPLL minimizes buffer delay while eliminating source jitter, so Hugo 2 and Qutest fed from one SPDIF source show no noticeable delay difference. Head-Fi, Apr 2018
2018-04-17Watts says latency is about 31 mS for 48 kHz input, due to the WTA tap length and 16FS oversampling in WTA1. Head-Fi, Apr 2018
2018-04-24Watts says all displays delay video, so more audio delay is probably needed; his projector has nearly 100mS of video delay. Head-Fi, Apr 2018
2018-05-05Watts advises setting the delay to zero or -10mS, noting a 10mS lip sync error is hard to see. needs context Head-Fi, May 2018
2018-07-21Watts says gain is unchanged for listening tests; the filter's 1.3 seconds of music means 0.63 s delay, so video mode uses an asymmetric filter with 100 ms delay. c w 24:08
2018-07-21Watts describes auto video mode on 48 kHz input, automatic source switching when USB is silent, remote control, and a DX mode for DX power amplifiers still under development. c w 24:49
2018-07-22Watts gives typical 12W draw, manual and auto video modes (1M taps 0.63 S delay, 2/3M taps 0.1S), and says Hugo 2 gets full 1M taps at 768k. Head-Fi, Jul 2018
2018-07-24Watts prefers automatic video mode for convenience; projectors and quality TVs delay enough that 100mS is fine with audio delay set to 0, since audio before video bothers the brain. Head-Fi, Jul 2018
2018-07-26Watts says the M scaler uses 1.42 s of music at 44.1 kHz, giving 0.71 s delay (0.65 s at 48k), hence the low-latency filter option. Head-Fi, Jul 2018
2018-08-06Watts says the M Scaler makes a huge difference to video sound, with stunning Blu-ray depth; he uses JRiver's negative video delay to cover the 'six, three seconds' latency. Pursuit Perfect System 6:37
2018-08-06Watts says video needs the video filter for lip sync, which uses 2/3 M taps but still outputs 705/768 kHz on the dual BNCs. Head-Fi, Aug 2018
2018-08-06Watts says Blu-ray films with a -0.63 s video delay in JRiver sound stupendous with the full 1M taps, with cavernous soundstage and separation. Head-Fi, Aug 2018
2018-08-17Watts says the video filter uses the maximum delay without significant lip-sync issues, 104 ms, since TVs with frame interpolation need an audio delay; longer delay sounds better. Head-Fi, Aug 2018
2018-09-03Watts endorses questioning one's own assumptions, and says he is not a fan of surround, which sounds flat and out of focus to him compared with stereo. Head-Fi, Sep 2018
2018-10-28Watts explains how TT2's volume display shifts with gain mode and where it clips, and gives delays of 34 ms for Hugo 1 and 63 ms for TT2 with video. Head-Fi, Oct 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
2018-12-16Watts says delay is 0.71 s at 44.1k and 0.65 s at 48k, while video mode's asymmetric filter cuts the future part to 0.104 s for lip sync. Head-Fi, Dec 2018
2019-07-02Watts says doubling the sample rate halves the latency, giving 16mS at 96kHz. Head-Fi, Jul 2019
2020-01-20Watts explains isolating output-stage supply currents, shown in TT2's distortion plot, says delay follows tap length and oversampling, describes the M scaler's asymmetric video filter, and heard no USB-optical difference. Head-Fi, Jan 2020
2020-04-11Watts says video mode is set to 104 mS, the longest delay usable with TV frame interpolation, and the delay stays fixed in pass-through in case video is playing. Head-Fi, Apr 2020
2020-04-12Watts confirms a delay of 0.661 s with a 48k source with video mode off, in both pass-through and normal mode, halving at 96k and 0.165 s at 192k. Head-Fi, Apr 2020
2020-04-20Watts says he bought an Nvidia Shield to delay video, with an HDMI switch extracting optical audio for the M scaler; Netflix and Amazon sync but BBC iPlayer does not. Head-Fi, Apr 2020
2020-11-21Watts says latency is unavoidable as reconstruction must see the future; The Mandalorian sounds fantastic at 1M taps via Nvidia Shield; 12 W, just under 1 A at 15 V. Head-Fi, Nov 2020
2021-06-14Watts says he has not investigated it but leans toward agreeing with Bob Lefsetz, never having liked surround sound because it sounds blurry, flat and unfocused. Head-Fi, Jun 2021
2021-10-20Watts uses Dolby off on Shield, says 2go is not his project, recommends a power bank for Hugo 2 desktop use, and describes his DIY solid-core PTFE interconnects. Head-Fi, Oct 2021
2022-01-10Watts says the Nvidia Shield can delay video up to 1 second to match 1M-tap audio latency, raising his appreciation of film, but quality is not all about taps. Head-Fi, Jan 2022
2022-02-01Watts answers Mojo 2 questions: 2G buzz mitigation, dual-data ring contact, 0.35S latency, DSP cores, 9 hours 11 minutes battery on USB, and 99.5dB SPL into HD800. Head-Fi, Feb 2022
2022-05-05Watts says the most accurate transient reconstruction needs symmetric windowing; he uses asymmetric windowing only in video mode, gaining lower latency at a sound-quality price. Head-Fi, May 2022
2022-05-14Watts says WTA taps split between future and past; M scaler video mode uses 98,304 future and 507,904 past taps (606,208 total); ASIO native USB allows DSD512, with fewer pops. Head-Fi, May 2022
2022-07-11Watts says Chord units need stereo PCM downmixes, surround is not on his radar, the Shield's 0.63 sec delay allows full 1M taps, and a Choral prototype supports 1 sec. Head-Fi, Jul 2022
2022-07-12Watts says it is unknown whether streaming services send a true stereo track or the player downmixes, and he now prefers streaming to Blu-ray since it allows full M scaling. Head-Fi, Jul 2022
2022-09-18Watts says an NVIDIA device's negative video delay lets him run the M Scaler at 1 million taps, a bigger M Scaler has one-second delay working, and he dislikes surround. The Source AV 4:38
2022-10-25Watts says a video mode supporting the Nvidia Shield's 1 s delay will be available on the Choral M Scaler, and it makes a tremendous difference. Head-Fi, Oct 2022
2023-03-26Watts says licensing the algorithm is possible but a business decision; software like PGGB cannot run in real time, which matters for streaming and video, where he uses scalers. Eric's Hi-Fi 10:28
2023-03-26Watts says his NVIDIA streamer allows a one second video delay so he can use an M Scaler; original PGGB files sounded nothing like it, and apodizing cannot reconstruct transients. Eric's Hi-Fi 11:13
2023-09-23Watts describes using the Refresh rate app on his Nvidia Shield to match frame rates per streaming app, with sound via an HDMI splitter to optical. Head-Fi, Sep 2023
2023-11-01Watts says sampling theory means shorter latency always degrades transient reconstruction, so low-latency gaming audio is a problem; studio latency also needs attention. Head-Fi, Nov 2023
2023-11-02Watts says an asymmetric FIR with short pre-ringing and long post-ringing beats short symmetric filters, as in the Hugo M Scaler's video mode. Head-Fi, Nov 2023
2025-05-12Watts says there is still a delay: 0.7 of a second with the M Scaler and three seconds with the Quartet M Scaler. Audiophile Junkie 17:12
2025-05-12Watts explains the delay: the more of the audio the filter sees, and the more coefficients are sinc values, the more accurately it reconstructs transient timing. needs context Audiophile Junkie 17:21
2026-08-30Watts says a filter with 28 zeros after the four would have a delay of 900 trillion years, 66,000 lifetimes of the universe, so that level is unreachable. The Source AV 7:27
2026-08-30Watts says a single-stage filter trades off two problems like a seesaw; his Blackbird WTA solves each in turn, a new structure cutting the delay to tens of seconds. The Source AV 8:23
2026-08-30Watts says the trouble with pro audio is that when you tell them three seconds of latency, they want 3 milliseconds. needs context The Source AV 10:07
2026-08-30Watts says latency is an issue for pro audio, and low-latency modes for the DAC sections of the interface project will be a massive problem to solve. The Source AV 10:20
2026-10-02Watts says Quartet has a one-second delay mode because the NVIDIA Shield can set a one-second video delay. HiFi Buys 2:14:45
2026-10-02Watts says perfect sync makes a big difference to the performance, but films lack the expression of ideas found in a good book. needs context HiFi Buys 2:14:58
2026-10-11Watts says the WTA filter handles two conflicting factors together, and that the taps needed came to 4 with twenty-six zeros, with a delay of 900 trillion years. Distorsiuni Interioare 19:19
2026-10-11Watts says the equivalent requirement fell from 900 trillion years to seconds or tens of seconds, so the improvement exceeds an order of magnitude. Distorsiuni Interioare 21:37
2026-10-17Watts says the delay would be 900 trillion years, or 66,000 universe lifetimes, and that this only reaches good reproduction, not the best. needs context article, Oct 2026
2026-10-17Watts says Blackbird WTA improves sinc reconstruction accuracy tenfold, eliminates the other problem and cuts the delay from 900 trillion years to tens of seconds; the algorithm, not four million taps, is the story. article, Oct 2026
2026-10-17Watts clarifies that the age-of-the-universe figure is the time for the signal to get through the filter. needs context article, Oct 2026
2026-10-17Watts says yes, still, because the limit is now tens of seconds, and an Ultima scaler will be a future project. needs context article, Oct 2026