Do digital cables (USB, coax, glass or plastic optical) make a sonic difference, and if so through what mechanism?

The mechanism he gives

Watts has said from 2014 that source jitter is not the route by which a digital cable changes the sound on his DACs. In October 2014, on Hugo, he wrote that USB cables can make a difference, "not due to source jitter (as this is completely removed by Hugo), but due to the amount of RF noise that gets injected into the ground plane of Hugo" Head-Fi, Oct 2014. In a podcast recorded in October 2025 (published March 2026) he gave the same account for his DACs generally: RF noise from the cable "then gets fed into the analog components inside the DAC", adding "At least in my DACs, because jitter isn't an issue" ecoustics 1:03:17. In between he said the same of source jitter in 2019 and 2024 Head-Fi, Apr 2019Head-Fi, Feb 2024.

The chain he describes: RF noise intermodulates with the analog electronics, creating noise floor modulation, which "makes it sound harder and brighter" (May 2014, Hugo) Head-Fi, May 2014. In February 2019 he said the noise does its harm through current flowing through the DAC's ground plane: "But it relies on a current flow through the ground plane. No current flow, no problems." Head-Fi, Feb 2019. In November 2017, discussing the Blu 2 to DAVE BNC link, he separated two effects: random RF noise affects warmth, while "Depth changes are due to signal correlated RF noise which gets into Dave's ground plane, then is demodulated by the analogue electronics" Head-Fi, Nov 2017. For USB in 2014 he said the signal-dependent ground currents from the source are something "the cable can't do much to change" Head-Fi, Oct 2014.

He says these effects fall below what can be measured. At a Nintronics open day in November 2018 (video uploaded December 2018) he said this "explains why we have internet cables That have no measurable differences but are audible" Pursuit Perfect System 29:21. At CanJam Singapore in March 2023: "with digital data it's still perfect, but you can hear differences with different cables", which he put down to noise floor modulation "you'll never be able to measure" Eric's Hi-Fi 32:01. He repeated the point in May 2024 Reactcore 0.

USB cables

2014 (Hugo). He wrote that "A lot of so called audiophile cables actually are RF poor and create more noise", and that he used to be very bothered about cables but is "very much more relaxed about it now" Head-Fi, May 2014. He warned that some expensive USB cables may raise RF noise and so sound brighter: "be careful if it sounds brighter!" Head-Fi, Oct 2014.

2016 to 2018 (DAVE, Hugo 2). In April 2016 he compared a $2 and a $200 USB cable on DAVE with headphones. He first preferred the $2 cable, then found the headphone cable was touching the USB cable; with the cables dressed apart there was "no difference whatsoever" Head-Fi, Apr 2016. He concluded "Dave is insensitive to the USB cable, as long as it is bit perfect data, and the RF characteristics do not affect any other part of the system" Head-Fi, Apr 2016. In February 2017, four cables from 3 to 300 USD: he "could not reliably tell any difference in SQ with Dave" Head-Fi, Feb 2017. He noted those tests were headphones only, that pre-amps and power amps can be RF sensitive, and still said "I maintain that the USB cable is not something to worry about with Dave" Head-Fi, Feb 2017. In October 2017 he said he had "never been able to hear any difference in USB cables whatsoever" on DAVE Head-Fi, Oct 2017, and on Hugo 2 "My listening tests have revealed no change at all in SQ with many different USB cables" Head-Fi, Oct 2017.

From November 2017, in the DAVE thread, he framed the result as two outcomes: cables either sound identical to a quality inexpensive one, or, for some audiophile cables, "Significantly deteriorate the SQ as the RF characteristics of the USB cable are poor" Head-Fi, Nov 2017. In December 2017 he wrote "all quality RF USB cables sound the same", and described one tuned cable that sounded different because it acted as an RF noise generator Head-Fi, Dec 2017. In September 2018: "I haven't ever heard a change using certified USB cables" Head-Fi, Sep 2018. In December 2018 he said his USB cable tests had been with Hugo, Hugo 2 and DAVE Head-Fi, Dec 2018, and that "certified class 2 cables sound the same" Head-Fi, Dec 2018.

2020 to 2022 (Hugo 2, Mojo, Blu 2, all his DACs). For Hugo 2 on battery with headphones, March 2020: "If you hear a difference it's either a placebo or the cable has been bodged to create more RF noise giving the impression of more transparency." He added that with a grounded amplifier "you might hear a small but real change", which optical would solve Head-Fi, Mar 2020. In May 2020: "I could not hear a difference between different certified class 2 USB cables" Head-Fi, May 2020. In October 2021: "Many audiophile USB cables are engineered to increase RF noise, making them brighter sounding. Quality audiophile USB cables sound the same as generic certified cables" Head-Fi, Oct 2021, and "USB cables can only change the sound by influencing the random RF noise levels; so a brighter sound is definitely a degradation" Head-Fi, Oct 2021. His December 2022 summary: "don't bother with audiophile USB cables" Head-Fi, Dec 2022.

2024 (CanJam NYC seminar). He said "RF characteristics of the digital cable can change RF noise levels, particularly if they're not impedance matched" MFG Audio Reviews 17:23; that "generally high-end USB cables are actually designed to make the sound brighter by increasing RF noise" MFG Audio Reviews 25:32; and that "the good audiophile USB cables sounds identical to a standard generic USB cable" MFG Audio Reviews 26:16.

Practical points he has made. Buy certified cables rather than make your own, as they must be impedance matched (October and November 2021) Head-Fi, Oct 2021Head-Fi, Nov 2021; "Solid core wiring is vital for analogue, but not so for digital" Head-Fi, Nov 2021. On length, in August 2021 a certified cable "should be length independent", though he had never gone beyond 2 m Head-Fi, Aug 2021; in May 2020 he said he had not experimented with USB lengths and gave only a guess Head-Fi, May 2020. Ferrites on USB cables made no difference in his tests in 2017 to 2019 Head-Fi, Nov 2017Head-Fi, Jan 2018Head-Fi, Apr 2019, though he allowed that noisier sources might differ Head-Fi, Dec 2017Head-Fi, Jan 2018. In October 2021 he said the cable he uses has ferrites on both ends but "sounds the same as a bog standard certified cable" Head-Fi, Oct 2021. Several times he traced inconsistent results to cable placement: keep the headphone or analog cables away from digital cables Head-Fi, Apr 2016Head-Fi, Aug 2018Head-Fi, Sep 2018Head-Fi, Oct 2021.

Optical cables: glass versus plastic

He has reported no audible difference between plastic and glass optical cables across a decade. April 2014 (Hugo): "I have tried glass against plastic using red-book, I could not hear a difference." Head-Fi, Apr 2014. October 2014: "I can't hear any difference between plastic optical and glass optical cables" Head-Fi, Oct 2014. March 2017 (Hugo 2 thread): "So long as it is bit perfect, I have not been able to hear any difference." Head-Fi, Mar 2017. November 2018: "in the past I could hear no change from a cheap plastic to expensive glass fibre" Head-Fi, Nov 2018. June to December 2022, a 5 m KabelDirekt plastic cable sounded identical to his 2 m Lifatec glass cable on the M Scaler Head-Fi, Jun 2022Head-Fi, Dec 2022, and "there is no need to spend money on glass to improve the sound quality - again with my DACs" Head-Fi, Oct 2022. February 2024: stock optical, the 5 m KabelDirekt and expensive glass fiber: "All work perfectly at 192, all sound identical." He added "there is absolutely no technical reason for them to affect the sound quality on my DACs" Head-Fi, Feb 2024. In the March 2024 seminar: optical type "doesn't matter what type, whether you're using glass or using Plastic fibre" so long as it works at 192 kHz, with the caveat "this is with my DACs" MFG Audio Reviews 26:16MFG Audio Reviews 27:01.

His other comments on optical cable quality concern whether the link works at 192 kHz, not how it sounds. In 2016 he said 192 kHz needs about 0.25 m of plastic or, for longer runs, quality glass Head-Fi, Apr 2016Head-Fi, Jun 2016; in 2019 he said plastic fibers longer than 0.25 m sometimes will not work well at 192 kHz Head-Fi, May 2019. In May 2022 he said regular plastic works at 192 kHz to 1 m but not 2 m Head-Fi, May 2022; a month later he found the 5 m KabelDirekt plastic cable worked at 192 kHz Head-Fi, Jun 2022. In 2014 he said DoP DSD may need glass with some transmitters Head-Fi, Apr 2014.

Optical as the reference connection

Separate from cables, he treats optical as the reference input because it blocks ground currents: "Toslink is the reference digital input, as it can't inject RF and correlated noise into the DACs ground plane" (April 2019) Head-Fi, Apr 2019Head-Fi, Feb 2024. His view of the inputs has changed over time. On Hugo in May 2014 he preferred optical Head-Fi, May 2014; in July 2016, writing about DAVE, he said "Before galvanic isolation, optical was so much better than USB" Head-Fi, Jul 2016, so the 2014 preference concerned Hugo's non-isolated USB input. In 2015 he called USB the best input on DAVE, by a tiny margin over optical Head-Fi, Nov 2015, and in July 2016 he wrote that on DAVE USB was slightly better than optical in detail resolution and flow Head-Fi, Jul 2016. In January 2018 he wrote that he had been wrong to think USB had the edge on optical, putting its extra brightness down to RF noise Head-Fi, Jan 2018Head-Fi, Jan 2018. In February 2019, in the Qutest thread, he still wrote "Technically USB is better from the clock POV; but in practice I have had optical sound identical to USB" Head-Fi, Feb 2019. He repeated the 2018 correction in March 2022, saying that on an MSI gaming laptop he "could hear zero difference between USB and optical" Head-Fi, Mar 2022, and in October 2022: "I too used to prefer USB than optical. But I realised later that I was making a mistake.", adding that the laptop on battery gave USB and optical that sounded identical Head-Fi, Oct 2022. That October 2022 statement ("there is no technical benefit to using USB over optical when using my DACs") differs from his February 2019 "technically USB is better from the clock POV"; both are his, and this page does not choose between them. Separately, in April 2018 and, on an estimated date, September 2018 he called USB best against AES or coax on DAVE and TT2 Head-Fi, Apr 2018Head-Fi, Sep 2018; neither post compares USB with optical. In April 2021 a noisy desktop PC made optical sound better than USB: "So for sure USB can be identical to optical, but with nosier sources, optical will win out." Head-Fi, Apr 2021. In March 2022 he said USB sounds identical to optical or worse, "depending upon the cables and the source" Head-Fi, Mar 2022. On coax, in 2014 optical sounded "noticeably warmer, smoother" than coaxial on Hugo Head-Fi, Jun 2014, and in May 2022: "If you have a choice, always go for optical over coax." Head-Fi, May 2022.

Coax, BNC and AES

In May 2014 he wrote "Only ever get problems with coax if you use un-screened coax cable, or cables that are not RF quality cables." Head-Fi, May 2014. On DAVE fed by Blu over BNC, December 2015: "Replacing it with proper RF quality cables solves noise issues." Head-Fi, Dec 2015. In April 2016 he said AES EBU "will inject RF noise into Dave from the source", making it sound brighter Head-Fi, Apr 2016. In August 2018 he wrote that "audiophile BNC cables either sound the same as a quality 75 ohm RF cable or sound worse", adding "don't let your wallet decide what it sounds like" Head-Fi, Aug 2018. In February 2020, in the Qutest thread, he said an owner's SPDIF locking problems were "down to the design of the cable" Head-Fi, Feb 2020.

The M Scaler to DAC BNC link is the one place where he reports that some cables audibly improve on standard ones. In December 2018 he said that link is RF sensitive because of the noise the M scaler creates, "rather than Dave being sensitive" Head-Fi, Dec 2018. In November 2019 he said clip-on GHz ferrites had sounded considerably better before he added RF isolation to the M Scaler outputs; afterwards solid core ferrites still helped, "but I should add that it's not a huge difference", and he does not bother with them when traveling Head-Fi, Nov 2019. In August 2020 he found longer cables, solid core 2 GHz ferrites and Wave cables helped Head-Fi, Aug 2020Head-Fi, Aug 2020, named Wave as the exception among audiophile BNC cables, and called the effect "the surface finish on the icing of the cake" Head-Fi, Aug 2020; he also called it "a significant difference on an AB test, but the scale of the change is forgotten about after a few days" Head-Fi, Aug 2020. In June 2023, for the Hugo M Scaler, he said an isolated power-bank or Wave cables each stop the GHz current flow, that longer standard BNC cables sound better than short ones, and that "this treatment isn't a night and day difference" Head-Fi, Jun 2023. He has said running the M Scaler on battery removes the BNC cable sensitivity Head-Fi, May 2020Head-Fi, Nov 2019, while noting in April 2021 that this depends on the system and source Head-Fi, Apr 2021. In August 2018, on a Hugo M Scaler to Hugo 2 link, he heard no difference between two BNC cables once the headphone cable was kept away from them Head-Fi, Aug 2018.

For Quartet, he said in May 2024 that the aim of its filters was that "if we can stop the circulation of currents, then it doesn't matter about the quality of the BNC cables" ConvinceMeAudio 2:12. In a Soundnews interview recorded in autumn 2026 and published in October 2026 he described one design goal as making it "so that whatever BNC cable you use, it sounds the same, so you don't need expensive BNC cables" article, Oct 2026.

What he has not said

He names no audiophile USB or optical cable that sounded better than a certified generic or inexpensive one; the only better-sounding cables he names are Wave, for the M Scaler BNC link Head-Fi, Aug 2020. He has not reported USB cable lengths over 2 m Head-Fi, Aug 2021, or listening tests of Ethernet cables; he has said Ethernet isolation is "not 100% guaranteed perfect isolation" Head-Fi, Jan 2020. His conclusions are for his own DACs; he says a DAC sensitive to source jitter "could be a completely different scenario" MFG Audio Reviews 27:01.

Quotes

“USB cables can only change the sound by influencing the random RF noise levels; so a brighter sound is definitely a degradation”2021-10-18 Head-Fi, Oct 2021

“Quality audiophile USB cables sound the same as generic certified cables”2021-10-13 Head-Fi, Oct 2021

“All work perfectly at 192, all sound identical.”2024-02-26 Head-Fi, Feb 2024

“I have tried 4 different USB cables from 3 USD to 300 USD, and could not reliably tell any difference in SQ with Dave.”2017-02-05 Head-Fi, Feb 2017

Dated positions

Caveats and open points

• All of his conclusions are stated for his own DACs; he says a jitter-sensitive DAC could behave differently.
• Most of his cable tests were with headphones; he says power amps and pre-amps can react to RF from a cable, so results with loudspeaker systems may differ (YMWV).
• The optical versus USB versus coax choice of input is a separate question from cables. He revised his view in January 2018, said in February 2019 that USB is technically better on clocking though optical sounded identical, and restated the 2018 correction in 2022; see the best-input answer.
• The M Scaler BNC link is covered in more detail in the BNC cables answer.
• The transcript of the November 2018 open-day talk (uploaded December 2018) says "internet cables"; this may be a mistranscription, so check the recording.

Topics: Cables, adaptors and ferrites, RF noise: mechanisms and audibility