Schiit Gungnir Multibit DAC – Meet “Gumby”

Digital audio has always been a quest to capture the “magic” of analogue audio. We’ve come a long way since CDs were first released, but along the way there have been attempts to simplify circuits and reduce costs resulting in some amazing leaps in sound quality and potentially some drift away from the smoothness of analogue.

For the purpose of this review I think it’s important to declare myself as a digital realist rather than an analogue purist. I admire those who enjoy the ritual of turntables and vinyl, but it just doesn’t suit my lifestyle given how good digital audio can be. The slight trade-offs have been worth it for me and I enjoy the chase for the ultimate (relatively affordable) digital setup.

gumby-5-of-11All this preamble leads me to the topic of this review, the Schiit Gungnir Multibit DAC (also known as Gumby). When Philips first created a DAC back when they worked with Sony on the first digital audio standards, they used a multibit design. Multibit designs are often considered to sound better than the more modern single-bit (delta-sigma) technologies, but MB DACs are generally considered to be more expensive and more complicated to build in order to reach the same (or better) quality than a modern delta-sigma DAC.

Multibit DACs have remained in the shadows (often the very expensive shadows) of digital audio as the popularity and proliferation of delta-sigma (DS) DACs increased. Recently though, MB DACs have enjoyed a resurgence led in no small part by Schiit, the pun-demanding US-based brand known for their irreverent approach to marketing and bold approach to reasonably priced amps and DACs designed to provide unreasonable levels of performance.

I had listened to Schiit gear before and hadn’t been blown away, but I was so intrigued by the reviews I read of their multibit DACs that I got my hands on a Gumby having never listened to one. Before I tell you all about it, let’s clarify (slightly) the difference between Multibit and Delta-Sigma processing.

Multibit & Delta-Sigma: What’s the difference?

OK, time for a massive disclaimer. What follows here may be the most dumbed-down explanation of DAC technology you’ll ever read. I’m adapting and further simplifying a great article from the MSB website in an attempt to explain the differences in ways that will make sense to others like me who don’t have a mind for binary data and similarly mathematical explanations of digital audio.

gumby-6-of-11In the simplest of terms, a DAC receives an analogue signal (fluctuating voltages), but those fluctuating voltages represent binary on/off pulses. The DAC needs to look at the fluctuations of the incoming signal and translate that from a series of on/off pulses into something resembling the much more graduated (wave-like) voltages of an analogue audio signal.

In the most basic terms, a delta-sigma DAC reproduces each sample in the analogue audio signal by producing a series of “pulses”. This multiple-pulse approach is why delta-sigma DACs are sometimes referred to as oversampling DACs – they break up a single sample into many samples or pulses of voltage. Due to the speed at which music is sampled (44,100 samples per second for CD audio), the available time to produce all these pulses is limited so the initial stage output from a delta-sigma DAC creates a jagged representation of the actual signal (think about trying to make a curved wall using regular rectangular bricks). Because of this, delta-sigma DACs employ heavy filtering to measure the “jaggedness” or error (which is predictable) and reconstruct the “correct” waveform. The better the filter, the better the sound, but it’s always an estimation of the actual waveform and is never actually 100% accurate.

Multibit DACs work differently by using a series of resistors to output a single variable voltage (not multiple pulses of the same voltage). For this reason, multibit DACs are sometimes referred to as non-oversampling (NOS) DACs because they reproduce each sample just once with a variable voltage. The resulting output from a multibit DAC more closely represents the actual waveform rather than the jagged version produced by a DS DAC (think about building a curved wall with clay instead of bricks). Now that might sound like an easy decision that multibit is clearly better, but there’s a problem. If any of the resistors used to produce the variable voltages are slightly out of calibration then the whole signal can be messed up. Add to that the variation in resistor performance depending on factors such as their build quality and even operating temperature and you quickly see that MB DACs introduce some very specific challenges of their own. This is why they are generally more expensive than DS DACs.

What I took from all my research was that a multibit DAC will generally sound more musical and natural than a delta-sigma DAC because there’s no need for heavy filtering* to remove the noise and errors created by the jagged waveform created by a DS DAC. However, I also took from my research that I’d have to cough up significantly more money to enjoy those benefits and coming from my already excellent Matrix X-Sabre (delta-sigma) DAC, I wasn’t in any hurry… until the Gumby came along.

* Schiit still employ filtering in their multibit DACs, but the type and degree of filtering required in a MB DAC is quite different as I understand it

Meet Gumby

The Gungnir is Schiit’s second highest performing DAC and, like all but the top shelf Yggdrasil, it comes in both standard (delta-sigma) and multibit guise. You pay about $700 (AUD) extra for the pleasure of multibit technology in a Gungnir and the right to call it a Gumby. For reference, the Matrix X-Sabre that I’ll compare this to later costs roughly the same as the delta-sigma Gungnir so they’re very comparable until the $700 multibit upgrade is added. The question of course is whether it’s $700 well spent.

I won’t go into a lot about the specs of the Gungnir, but here they are:

  • DAC chips:  Analog Devices AD5781BRUZ x 4 (2 per channel, hardware balanced configuration)
  • Analog stage frequency response: 20Hz-20Khz, +/-0.1dB, 1Hz-200KHz, -1dB
  • Maximum output: 4.0V RMS (balanced), 2.0V RMS (single-ended)
  • THD: Less than 0.005%, 20Hz-20KHz, at full output
  • SNR: > 115dB, referenced to 2V RMS
  • Inputs: RCA SPDIF, BNC SPDIF, Optical SPDIF, USB (all up to 24-bit / 192kHz)
  • Outputs: One pair XLR balanced, two pairs RCA single-ended
  • Power supply: two transformers (one for digital supplies, one for analog supplies) with 8 stages of regulation, including separate local supplies for critical digital and analog sections
  • Size: 16″ x 8.75” x 2.25” / 40.6cm x 22.2cm x 5.7cm
  • Weight: 11 lbs / 5 kg

Interestingly, the total harmonic distortion produced by the multibit Gungnir is higher than the delta-sigma DAC, but at 0.005% it’s hardly an issue so much as a point of interest. Perhaps the most interesting specification though is the massive attention to detail that has clearly gone towards the power supplies in the Gumby. It’s a big beast – more than twice the size of the X-Sabre, but the extra space has gone to good use by the look of this list with fully balanced operation, multiple power transformers to isolate sensitive circuitry from noise, and plenty of great input and output options for maximum flexibility.

Connections and Compatibility

gumby-11-of-11The Gumby is basically plug and play for all types of connections except USB from a Windows PC (Apple is still plug and play I believe). For Windows users like me, you need to download the drivers first, but my experience is that they install easily and work seamlessly. Other than installing Windows drivers, everything else just works with the Gumby. Connect to SUB, BNC, RCA SPDIF, or optical and deliver any signal up to 192kHz and you’re set. I’ll discuss in the sound quality section how different inputs may vary in terms of sound quality, but there’s no variation in terms of functionality.

There’s a light on the front of the unit at the right hand end that Schiit refer to as the “buy better gear” light. This light will always flash when you first power up the Gumby and I assume it’s to allow the circuits to all initialise because the Gumby won’t create any sound until you hear a little click and the light stops flashing. Given the importance of temperature in multibit DAC operation, this may also be giving the resistors some time to reach an acceptable operating temperature.

gumby-1-of-11In the unlikely event that you have a truly shitty source or cable, the “buy better gear” light (aka the Clock Mode LED) will continue to flash, but the DAC will still work – it’ll just use a different type of reclocking to improve the signal as best it can. The light is more about letting you know that you could be hearing better sound quality if you upgraded the offending gear. Schiit call out the Airport Express as one such culprit that will illuminate the BBG light, but reassure users to rest easy if the gear is not upgrade-able for any reason – just keep on enjoying it, they say.

For outputs, the Gungnir offers one set of XLR outputs for amplifiers with balanced inputs and two sets of RCA outputs for amps with single-ended inputs. All outputs work simultaneously so the Gungnir can happily be a source to multiple amps if you’re like me and have a few different rigs on the go sometimes.

Sample Rates

It’s probably important to address sample rates here because the Schiit DACs don’t do DSD – they only process PCM audio up to 24-bit / 192kHz. Some people will instantly click away at this point and that’s cool. I believe DSD audio can sound really good so I respect people who want DSD-capable DACs. Personally though I have no problems at all letting go of DSD. I have very little DSD audio in my library and can’t honestly say that it is consistently better than well-recorded PCM (WAV / FLAC) audio. Given how beautifully the Gungnir renders PCM audio I have no qualms about it not doing DSD.

To confirm, the Gungnir (Multibit or Delta-Sigma) will handle 44.1kHz, 48kHz, 88.2kHz, 96kHz, 176.4kHz, and 192kHz recordings at either 16-bit or 24-bit. Basically, if it’s a FLAC recording, the Gumby will be fine with it and will sound great!

Sound Quality: This is the Schiit!

Source selection button
Source selection button

When it comes to defining DAC sound quality, the differences in sound tend to come from two places – the ability of the DAC to manage jitter / signal errors and the characteristics of the output stage of the DAC (where it sends the freshly baked analogue signal to the outputs for your amplifier). The sum of the Gumby’s signal conversion and output stage results in a sound that is smooth and perhaps slightly warm. If I use the comparison of the Matrix X-Sabre as a reference point, the Gumby has an almost identical sound signature in that it’s slightly warm and wonderfully musical. Where the Gumby differentiates itself is in the smoothness. Initially, this smoothness was a bit off-putting and had me thinking the Gumby sounded a touch thick and rolled-off, but time and some warm-up soon corrected my initial erroneous impressions and allowed me to hear the Gumby’s true colours.

Sound Impressions

gumby-7-of-11Now that I’m comfortable and used to the Gumby’s sound I can only describe it as the closest thing to true analogue reproduction that I’ve heard to date. The sound is remarkably smooth and nuanced. Details are laid out in a rich landscape of nuance and texture, but there’s none of the glare that many of us are used to in the world of delta-sigma DACs. At first, the lack of glare can be very off-putting if you’re coming from years of DS decoding (like I was), but given a bit of time, the wonders of a well-executed multibit system become abundantly apparent.

When I switch back and forth between the X-Sabre and the Gumby now I am struck by two things:

  1. The X-Sabre is still an outstanding DAC and sounds almost identical to the Gumby in all ways (signature, soundstage size, quality of imaging and spatial cues) but one…
  2. The Gumby puts the cherry and cream on top by stripping away the final sense of digital haze and letting all of the glorious nuances of every recording shine through

Compared to the slightly digital (by comparison) X-Sabre sound, the Gumby leaves space between each sound by virtue of the lack of digital haze filling the gaps. Each sound is more precisely rendered – probably because the multibit decoding process is getting it more accurate from the get-go and doesn’t require a filter to fill in the blanks. Treble-focussed sounds like cymbal hits have a smoothness that’s beguiling, but isn’t a case of viewing through rose-coloured glasses so much as viewing with no glasses at all. There’s the sense that nothing is getting between the music and your enjoyment when you listen to a Gumby with a nice amplifier and your favourite pair of cans. It’s safe to say that it’s the closest thing I’ve heard to my memories of a nice turntable setup – smooth, detailed and rich, just like a live performance.

Incremental Gains

There are two incremental gains I want to mention here because I think it’s very important to understand them before you dig too deep into your wallet.

Is This Schiit Better?

While I believe you will hear a difference between a DAC like the X-Sabre (i.e. DACs performing in the $1,000-$1,400 range), you shouldn’t upgrade to a Gumby expecting it to transform your entire listening experience. There’s no doubt that I am completely happy with the upgrade from the X-Sabre to the Gumby, but it’s hard to measure the economic value of sound quality improvements and I would hate to lead anyone down the path towards expecting a life-changing upgrade. If this were the Olympic Games, the Gumby would be the gold medal favourite despite being matched up against the best of the best. Other DACs like the X-Sabre are deservedly in the field of contenders, but the Gumby is like Michael Phelps – not always blowing away the competition, but always better.

Maximising the Value of a Gumby

gumby-8-of-11I’ve already covered the fact that there are many input options on the Gumby: USB, optical SPDIF, RCA SPDIF, and BNC. I’m currently writing 2 articles about the pros and cons of the various digital audio interconnect formats and the results have been illuminating. When it comes to the Gumby, it will sound great from a good USB setup, but there is no doubt that different inputs can produce better sound quality. I’m not a technical expert and therefore won’t presume to know how much of the difference is better input circuitry on behalf of the Gumby and how much is a reflection of the pros / cons of each transmission method, but I will detail what I hear on each input:

  • The USB input is clearly the least spectacular of the 4 inputs available. It still sounds great, but you’ll hear gains if you can use alternate methods. What holds the USB input back is a degree of jitter / noise that, while not exactly audible, definitely creates a sense of compression in the music. The soundstage seems a bit smaller, there is less sense of space around individual sounds, and the background is less “black”. I can’t stress enough that USB audio with the Gumby still sounds great – it’s just that other inputs sound even better.
  • The optical and RCA SPDIF inputs should both be quite similar in performance, but I don’t have a high quality optical cable to properly compare the two so I will tackle this from the angle of the RCA SPDIF. As an aside though, the optical SPDIF with a cheap, old cable was still very slightly preferable to the USB connection even when using AudioQuest JitterBugs and the top-of-the-line AudioQuest Diamond USB cable. Connecting via RCA SPDIF improved on the already excellent USB sound by removing the sense of compression and creating a greater sense of space in the sound overall. There’s also the benefit of slightly smoother treble.
  • The top of the tree for input options on the Gumby is definitely the BNC input which takes everything good about the SPDIF sound and further improves on it. All remaining sense of jitter and noise is stripped away and the sound reaches peak levels of smoothness, cleanness, richness of tone, sense of space around sounds and overall size of soundstage. Everything becomes almost holographic in nature to the point that I am now constantly removing my headphones to check if sounds I’m hearing are in the music or in my house. (That’s slightly annoying at times, but so worth it!)

The point I hope I’ve made here is that the Schiit Gungnir Multibit is worth buying regardless of whether you’re going to use USB, SPDIF, or BNC, but if you can find a way to get BNC input you should absolutely go for it. If you’re like me and only have a USB source like a laptop computer, check out my post about upgrading your computer audio without breaking your budget.

Conclusion

This is a tricky conclusion to write because flicking between the X-Sabre and the Gumby tells me that they are in the same league, but there’s no way I’d ever go back to the X-Sabre now that I’ve owned the Gumby for a while (and connected it using BNC). This could be one of those cases where the Gumby perfectly suits my listening preferences thanks to its smooth, fatigue-free and highly musical sound, but don’t equate that with it being overly lush or thick sounding. From my experience, the Gumby is a creator of perfect renditions of life-like auditory experiences. It’s sound is nuanced and detailed with a sense of ambience and soul. It will reveal crappy recordings and will reward great recordings, but either way it will present the best possible rendition of any recording in a manner that’s astoundingly reminiscent of a great vinyl setup.

Lachlan Fennen Written by:

Facilitator, training design consultant, blogger / writer and amateur photographer

20 Comments

  1. Peter
    November 20, 2016
    Reply

    Forget about the USB input of Gumby. Buy a decent USB-spdif converter like Yellowtec PUC2 Lite with a Neutrik AES/BNC transformer and you have a killer-combo. Tried to put in a Jitterbug but that ruined the sound!

    • November 20, 2016
      Reply

      Hi Peter, I agree that the USB input isn’t the strongest. I actually took a slightly different path and setup a Roon endpoint connected to the Gumby via BNC – it sounds sublime!

  2. Rile
    July 28, 2017
    Reply

    What is a roon endpoint and how do I connect it to DAC via BNC. Please email me some instructions as I am a new user: rilebru at hotmail dot com thanks in advance!

      • Rile
        August 9, 2017
        Reply

        Thanks so much and get better soon!

  3. Peter
    August 7, 2017
    Reply

    Hi Lachlan!

    Sorry to hear you are not well. Get better soon! I am VERY curious to hear if the Schiit USB Gen. 5 USB to spdif converter are as good as Schiit claims.

  4. PETER JASZ
    November 29, 2017
    Reply

    Lachlan: Although I likely go further back than you (lol), I too (since 2000/2001) have dispensed with vinyl.
    Today, even 128-320 Kb/s sounds easily the equal (much better actually ) than any up-to $1,500 (US$) table/arm/cartridge.

    Those who knew me since CD’s introduction (1983) throughout the remainder of the 80’s and 1990’s wold now chuckle at the 180-degree turn-around.. No laughing matter (lol), from about the turn of the century, digital sound became enjoyable (for me), and slowly my nearly 2k LP’s were sold off -never to regret or look back (although about 100 of my favorites were retained). Digital sound quality was know on par with the best analog. Things improved from then on, to today. Great stuff.

    Annnnnyway, today’s digital can easily outperform virtually any analog set-up (remind oneself of the big round plastic disc spinning at 33.33 RPM). That Tables can sound impressive there is no argument. But so does digital music. AND, the inter-modulation distortion (IMD) present in virtually all vinyl/arms is mercifully absent in digital replay. The sheer transparency/resolution of digital blows away that of the stone-age table (although I have no experienced tens-of-thousands of dollar set-up’s. Nor do I wish to.) Regardless of how fanciful the hardware, the software remains unabashedly inadequate.

    For anyone with digital I would strongly urge an AC power re-generator to power their digital gear (DAC, CDP, Streamer). You will discover how far digital can go.

    Lachlan, you make an interesting observation with M-B vs. D-S. I’d lie to know (based on your experience) which one “tracks” the signal more faithfully ? In other words, which presents each/every CD in its inherent, unique tonal shading ? The one that does, is the more accurate (and hence enjoyable) of the two.

    I’m also intrigued that spdif (BNC of course) is clearly superior to USB (using a quality cable AQ ‘Carbon or ‘Diamond’).
    I think it may be time to upgrade my Schiit Modi-2 Uber into something more appropriate (that offers BNC inputs).

    In the meantime, I may very well order a Gustard U-12 (spdif converter) and experiment vs. USB. Your comment that a generic spdif cable outperformed USB with AQ ‘Diamond’ is truly fascinating.

    Be well,

    peter jasz

    • November 29, 2017
      Reply

      Hi Peter,
      I can’t tell you from a technical, measurable perspective which format better tracks the signal, but I can tell you that I almost always prefer the Multibit (MB) version of the reproduction to the Delta-Sigma (DS) reproduction. That said, some DS DACs will out-perform some MB DACs and vice versa. In my experience so far, when comparing equally matched DACs, the MB approach is more organic, has a better sense of space and is therefore more realistic to listen to. Based on my reading, this is probably due to the extra filtering and noise shaping required for the DS approach.

      In terms of your surprise with the BNC revelation… me too! I was totally blown away at how significant the difference was. It’s not that a good USB setup is bad, but a decent BNC setup is fantastic!! Even SPDIF may provide significant improvements over USB, but it all still depends on the DAC as some DACs excel at USB and have crappy SPDIF integration so some pre-reading on your specific DAC is helpful. That said, I believe a BNC setup will almost always trump the alternatives because of the much more consistent impedance matching (vs coaxial SPDIF) and noise isolation (vs USB). A top notch optical setup may offer many of these benefits too, but I have not been able to come at the high cost of top-end optical cables and found BNC to be noticeably better than my cheap-to-average priced optical cables.

  5. Victor Lee
    December 7, 2017
    Reply

    It’s astonishing how much Lachlan and Peter Jasz love buying into digital cables. Simply put, and it’s fairly well known, that Schiit’s USB integration has been fairly poor which makes their products fairly dependent on work arounds to bypass their USB implementation or adding additional grounds as their implementation was incredibly finicky with it. However, their USB 5 upgrade is supposed to help quite a bit, but in the mean time it allows them to sell quite a few Wyrds and Eitrs before people jump on the USB 5 upgrade. But ultimately it’s fairly funny how quickly one jumps off the “Audiophile USB cables make so much of a difference and are so much better!” and jump straight to the digital coax and optical realm, and now looking to buy “audiophile” quality optical and coax cables. It never ends for those without knowledge as marketing psychology wins out with those who don’t know any better. Never mind AGAIN that if the data streams arrive intact without audible levels of jitter you’re not going to benefit your sound any – but I don’t think that’s ultimately the point with many of these ‘upgrades.’

    • December 7, 2017
      Reply

      Victor, explain something to me… if USB cables make no difference and digital is digital, why does the USB 5 upgrade make a difference??

      • Victor Lee
        December 8, 2017
        Reply

        Lachlan – with all due respect, do you really have an issue comprehending the difference between a USB interface and a USB cable? The interface involves a USB chips, ground isolation circuits among many other attributes. Just look at the USB5 upgrade board for your Gungnir on Schiit’s site for pete’s sake. Schiit’s USB interfaces are full of variability when it comes to connection quality, nothing to do with the cable but they’re reliant upon if they like your computer’s USB interface. More than 50% of the time Schiit’s ones would either not connect or drop packets and many would just introduce audible jitter. But there’s a massive difference between the interface and the USB cable…

        A USB cable is just a cable, 2 signal wires, a ground and a VCC hot wire wrapped in shielding. In other computer terms it’s like saying your system I/O is the same as a SATA cable. That’s what you trying to reach to, again, showing a fundamental lack of understanding and thus showing that those who do not know any better and don’t care to learn will make the same mistakes again and again.

        • December 8, 2017
          Reply

          Victor, your response contains some really good points and I thank you for sharing the information you have. Can I please ask you once again to pull back on the hostility? My question was honest – if digital is digital, then why should a different USB receiver make any difference. Sure if it’s failing to connect then that’s a big issue because the data isn’t even coming through, but if the data IS getting through then I don’t see how this can differ from USB cable arguments. Regardless of what the item consists of, all arguments about USB cables seem to fall back to the fact that if the data is arriving intact then there’s nothing more to it. How are you suggesting the USB receiver is different and immune to this reasoning?
          Please note, if your future responses continue to be disrespectful and hostile I will be deleting them. I hope though that you will continue the conversation in an open and educational manner. I do not claim to know all the science and report specifically on the subjective nature of what I hear so please help people (including me) to understand your point of view.

          • Victor Lee
            December 8, 2017

            Lachlan – First sentence I asked a question that is not hostile. I concluded on you were making the same mistakes as before because you rely on marketing as your information source. Again, not hostile but an opinion based previous conversations and actions by you. Hostility would be calling you names and trying to equate your intelligence, wealth, maturity (all of which we’ve seen before from another source but you allow those) none of which I’ve done. If you consider all dissenting opinions as hostile then anyone who disagrees with you is hostile.

            In order for this to remain ‘non-hostile’ I encourage you to look up the plethora of information regarding USB transmitters and USB receivers and how their circuit boards are built and what considerations must be taken in to build a good one. They’re absolutely and fundamentally different than a USB cable, you can read up all you want about USB cables and what goes into a USB certified (BTW Audioquest, iFi and other “audiophile” USB cables are NOT USB certified) cable. Then come back and discuss as any opinion that doesn’t align to yours is automatically defined as ‘hostile.’ As you take up the position of the ‘nit-picker’ who’s role is to find any tiny hole in the argument and poke at it (never mind that a blog comment section is a horrid format to try and explain anything technical with 100% accuracy) to not prove a point, but only to ‘prove’ there could be a hole in the argument so you can fall back upon your own. It’s just not a role of an individual who seeks information but looks to push their own rhetoric, in this case it’s something along the lines of, ‘digital cables somehow change the bits to make music sound better with no proof, logic or blind tests but just my own opinion.’ That idealism is hostile toward information and knowledge as it places all the emphasis and the absolute fact on a very biased procedure of “I heard it” with no idea as to the why or even how.

            As for your issue, please understand the data chain. HDD or Internet data -> CPU -> Host USB interface -> USB cable -> USB receiver interface -> output to DAC interface -> Analog output stage -> Analog signal into headphone cables -> headphone driver -> Sound. Now, understand that the data can arrive intact at the USB receiving interface but due to a poor interface can introduce audible jitter and with poor isolation can convey even computer noises through the ground due to poor ground isolation. Here’s a thread about the low end Modi but it has the same base USB interface.

            https://audiosciencereview.com/forum/index.php?threads/budget-dac-review-schiit-modi-2-99.1649/

            But again, if you take this post as hostile then it shows to the bubble that many others have built before you in the “audiophile” realm as they begin to believe in these myths and eventually hold them to be absolute truths with the only data is, “I heard it” and no one will tell them or even show them otherwise. That isn’t hostile but an observation of the ‘audiophile’ world and why so much snake oil is perpetuated through this industry.

          • December 8, 2017

            Thanks for the response, Victor. If you look back through this thread you will see that I asked both you AND Peter to be respectful and civil in this discussion. I will equally delete (after fair warning) anyone’s comments if they continue down a path of personalising their criticisms rather than keeping the conversation to the question at hand.

            You are incorrect in assuming that my conclusions are based on a susceptibility to marketing propaganda – let’s stick to the facts at hand and not draw conclusions and assumptions about one another. There are many things in the world that we still cannot fully explain scientifically and yet we experience them (e.g. intuition is only partly understood, auroras are still not fully understood, etc.). The inability to prove why something happens doesn’t make it any less real so discussing it and questioning it is not a biased procedure – it is totally honest and transparent as I have never claimed to fully comprehend the ‘why’ of any of it.

            I understand the arguments made in terms of the poor performance of the Schiit USB receivers and I completely agree with this having heard the differences myself between USB and SPDIF / BNC on the Gumby (not as pronounced as the Modi review I think), but I still don’t feel like you’ve answered my question. If the ability of a USB receiver to isolate noise, etc. can influence the sound, why are you suggesting that a cable’s ability to isolate noise sources (e.g. keeping power and signal wires better insulated within the cable, preventing external noise sources from entering the cable) cannot have a similar (albeit smaller) influence?

          • Victor Lee
            December 8, 2017

            “If the ability of a USB receiver to isolate noise, etc. can influence the sound, why are you suggesting that a cable’s ability to isolate noise sources (e.g. keeping power and signal wires better insulated within the cable, preventing external noise sources from entering the cable) cannot have a similar (albeit smaller) influence?”

            SImply put, that’s not the role of a USB cable. The USB receiver doesn’t even “isolate noise” from any signal. You’re confusing the power circuitry of the device to USB signal. The USB cable does not eliminate poor grounds or noisy circuitry in the host or the receiving USB interfaces. It’s designed to connect them with over a specified impedance line, that’s it. The “external” sources you speak of for the USB cable are not the interfaces on a cable or the devices connecting to each end, it’s RF signals. Again, a digital signal can have 50% noise and still have a clean signal because it’s not about minute voltages that can change the signal, it’s about thresholds. A USB cable’s role in the system is ultimately passive, no switching, no computing, nothing but a wire that connects two devices, with the caveat that the information is CRC checked upon arrival. If you wish to continue to bring up noise from the power line in the USB wire makes an audible difference, put a piece of tape over the VCC pin to eliminate it completely on a cheap USB cable. If it makes an audible difference it’s likely due to a ground loop from a poorly isolated USB interface on either end and not because a $300 USB cable will make the system sound better. I hope it eventually makes it through but minute voltages do not make any difference in a digital signal, that’s only in analog. Along with a USB cable does not share the same role as the USB interface as they’re fundamentally different…

            This is a USB interface card for your DAC or an example of what one looks like.
            http://www.schiit.com/products/gen-5-usb

            I hope it’s obvious the difference between that and a USB cable is. I hope you understand that most of your arguments are directly from the cable makers themselves. So it’s hard to believe your defense of not being affected by propaganda when you have a complete line of Audioquest products and used direct links to their marketing as ‘proof’ to validate your stance. Even this power line inside the USB cable must be isolated by a few additional millimeters is another cable marker argument. So it’s very very hard to believe you haven’t been affected by the marketing as you continue to repeat. Step out of the audio realm and use some of the information provided by the curators of USB themselves or IT based reasoning. You’ve been influenced by the marketing a lot more than you may be willing to admit.

            And lastly, if you don’t wish to delve into the why or how something could affect the end product (sound) then minimally conduct a blind test. It’s not hard and may give some light into how much the psychology and bias we all put into our gear. The confirmation bias of, ‘it’s better because it’s supposed to be better’ is very strong and until you’re willing to be vulnerable and remove that bias from it will you really experience what additions actually make your sound ‘better’ instead of ‘different.’ I’d bet if you ran through all your USB cables, cheap optical lines and a BNC line blind and just giving them a rating of best to worst you’d see that these ‘obvious’ differences may not be there. But it takes a bit of courage to allow your beliefs to be vulnerable to a test you may not pass.

  6. Peter
    December 7, 2017
    Reply

    I have now upgraded from USB 2 to USB 5 in my Gumby. Gen. 5 is far superior to gen 2. In fact I have now sold my Yellowtec PUC2 Lite and Neutrik AES/BNC transformer. USB cables do matter but unless the source is good it is money wasted. A laptop/pc/server etc. is NOT good enough. If you want to know how good USB can be, you must ad a networkplayer (like Sonore Mrendu or SOtM SMs-200) to the chain.

    • Victor Lee
      December 8, 2017
      Reply

      While I agree the USB5 board is better than the USB2 implementation, what in your mind makes any digital signal change over a wire? Digital signals are CRC checked and are either dropped (popping noises) or allowed through after they pass the CRC check. Throwing in stuff inline to a digital signal arbitrarily doesn’t help unless you’re specifically trying to remove a ground loop or something. You can throw in as much marketing jagon as you’d like but ultimately your DAC plays the CRC checked data stream. No cable can improve an already accurate series of bits.

  7. Peter
    December 8, 2017
    Reply

    Victor Lee – You are wrong about digital signals always being perfect OR getting dropped. Signal in USB is small pulses of current getting translated to “0” or “1”. “Timing is key in that translation and sadly it is never 100 % accurate = you get jitter. I have heard a lot of different digital cables and some sounded down right awfull.

  8. Victor Lee
    December 9, 2017
    Reply

    Peter: “You are wrong about digital signals always being perfect OR getting dropped.”
    That’s exactly what USB isochronous transfers are. Hence why each packet is CRC checked and then allowed through or dropped. Never mind that if one bit is off you’ll get insane jumps in volume or frequency as changing one bit in a number can wildly change the output there’s literally no subtlety in binary never mind that the error will need be constant and consistent enough to establish an audible pattern to interject into the subtlety of music… But since the packets are CRC checked you either get a dropped packet that doesn’t pass or ones that pass that are bit perfect or ones that are marked as errors.

    http://www.xmos.com/news/articles/17888

    “Isochronous transfers are used to transfer data in real-time between host and device. When an isochronous endpoint is set up by the host, the host allocates a specific amount of bandwidth to the isochronous endpoint, and it regularly performs an IN- or OUT-transfer on that endpoint. For example, the host may OUT 1 KByte of data every 125µs to the device. Since a fixed and limited amount of bandwidth has been allocated, there is no time to resend data if anything goes wrong. The data has a CRC as normal, but if the receiving side detects an error there is no resend mechanism.”

    http://wiki.osdev.org/Universal_Serial_Bus

    “The USB protocol highlights the following possible method for the host or a device to detect an error in an isochronous stream:

    -High-speed, high-bandwidth isochronous transactions use data PID sequencing (data bit toggling), an isochronous sink can determine that a data packet was missed when it receives an invalid data PID sequence.
    -The host controller and device can both see SOF packets on the bus. If the SOF packet is issued for a (micro)frame that is expected to carry the periodic data of an isochronous endpoint, but the data is not transmitted, then the hardware can determine that a packet was missed.
    -The protocol provides CRC protection to ensure that the data has not been corrupted.
    -If an endpoint sees the token packet but does not see the associated data packet within a bus transaction timeout period, then the data packet failed to transfer.”

    Peter: “Signal in USB is small pulses of current getting translated to “0” or “1”. “Timing is key in that translation and sadly it is never 100 % accurate = you get jitter. I have heard a lot of different digital cables and some sounded down right awfull.”
    Love to see you try and explain how much jitter actually makes it into the audible level. NONE of this is audible at -126dB.

    https://audiosciencereview.com/forum/index.php?threads/do-usb-audio-cables-make-a-difference.1887/

    Again, so much pseudo science and ‘it could happen’ and no real facts. A certified USB cable that is working properly will not have an audible difference when it comes to jitter. As per that actual test cable length matters much more than if the cable is “audiophile” quality. But look at that, backed up with test data and USB specification. But at least you’re not claiming that a cable gives warmth like some others here.

    • December 11, 2017
      Reply

      Victor, as I understand it, digital audio via USB is not subject to any CRC checking due to the timing elements as per Peter’s posts. Everything I have read about this topic would suggest that it is the nature of this unchecked data and the importance of the timing element that result in fluctuations in sound quality despite using a digital coding / decoding method.

Share your thoughts...