Schiit Byygy Dac Just Landed

I think we forget a few things. First off, hindsight is 20/20. We have to recognize that Schiit is a growing company and over time things change. The market changes. They came in with a value product and also did something only one other company I know of did (Metrum), in making an industrial DAC into audio output with also their custom filter. The DACs from Bifrost on up were designed with upgradeability in mind. Kind of like a computer motherboard and just upgrade your graphics card. This was fine, until the ante with remote controls, smaller and more affordable chassis, connectivity with you phone (which is an option), and also chasing all the audio stuff.

I also think we forget it is a product with the same name. People still like the older model. If it was a whole new model name, the other one would drop in value like a stone. I think most top end products are still great performers, no matter what. They just might have a few more tradeoffs than more current models, however more current models may offer something less as well. So these are all things to consider. (You are also talking to a guy who still uses an MJ1 as his main amp and for a small amp a Vali 1.)

This happens to a bunch of companies. How many love old JBL or old Crown amps? However, in the commercial audio market, it would hard to find a Crown D150 still in service somewhere. But people still love them. And they love old cars too. Every few years Corvette makes a new model, but many claim that whatever year production is the best. Only stuff like computers and smartphones these days do people NOT want to go back to. I mean, how many want to go back to PS2 graphics or an iPhone 4?

All that is to say that each product has its own hype cycle and over time makes its place in the market. But that takes time. There will always be a hype to the new product and thus FOMO.

 

Some people are just more vocal about their preferences and may be self conscious about how SBAF has had a years long reputation of being pretty biased towards Schiit (... and Sennheiser, and ZMF, and soekris, and ECP, and Eddie Current, and Donald North Audio, and Fostex, and Grado, and Shanling, and more recently iBasso...) even though there have been instances when yeah, certain things have been smacked down as being less than nice e.g. the Heresy/Heretic and Magnius, never mind how contentious other things from the above-named brands have been. I feel like I'm the only person here who actually really liked the Klipsch HP-3 enough to try and work with its idiosyncrasies lmao.

I like to think of it as them showing care for the forum at large in some way, trying to keep it from devolving into an echo chamber of sorts. It should be no surprise by now that I vehemently disagree with a lot of the folks here on some topics, but that doesn't change the fact that this is still a weird sort of cosy online space where folks are generally nodding acquaintes with one another, so that sort of nebulous concern for overt appearances may not be entirely unwarranted or even uncharitable (depending on how it's expressed), even if they might be forgetting that the people behind that company are just... real people that seem to be actually passionate nerds who happen to hang out here too.

But yeah I think people do forget that if you COINCIDENTALLY happen to like how Schiit stuff tends to sound then there's really not a lot of competition at the price points they target. I remember at least a couple regulars who STRONGLY dislike how the Magni Piety sounds, and despite me really enjoying it overall for the price I can respect that and comprehend the rationale behind the lack of enthusiasm.

All that to say that eh, everyone's got preferences and if we all agreed on things then the world would be less fun. I'd have gotten a Pietus or a TA-66 myself if I were just following what were popular on here but given my preferences and want to run lower-impedance loads the Lyr+ made more sense; yes that just happens to be another Schiit product but seriously: compared to similarly-priced things I've demoed locally it's the most pleasing sounding thing I've gotten ears on (note: it was still a risk on account of being a blind buy, but I think I have a general idea of most folks here's preferences and touchstones so I read it as being a relatively safe bet).

Don't get me wrong I AM still worried about it randomly dying either due to lightning strikes or production issues cuz unlike a lot of the folks on here the cost of shipping the thing back to the mothership for repairs should the need arise (I'm KICKING wood right now) would be painfully expensive. Even with that though, what sounds like it that's at a similar price point even accounting for another $100+ spent on shipping? No seriously, I'd love to know.

 

I mean this is going to be good; it's a yggdrasil, which has been technically very good since it got dither. Schiit is just finally moving to discrete delta-sigma to try to lower the noise floor more but does it really matter? What really matters are the amplifiers and analog electronics, filters (which are trivial about 44.1 and 44.8 where there are a few approaches), and clocking.

Proper DA filtering is simply a brickwall filter to kill the mirror images when upsampling then a first order low pass filter somewhere in the mhz to get rid of all that shaped noise. That's all it is. That's why mHz was adopted. Yes delta sigma has less phase shift than many old school DAs that used 6th order filters or worse. First they used second order and then 1st order when the speed of the modulators increased. There's nothing special about it.

Some tout non linear phase filters for the brickwall portion but these are all colored and inferior in real world use. They're used in many things because they have lower latency for most of the signal at the expensive of scrambling the phase of audio within the audible band.

 

I respect where you are coming from, but I agree with Schiit on this one. Digital audio is not a mature technology: It still sounds like ass.

Why Soldiers Don’t March on Bridges | Resonance Explained in 2 Minutes
Engineers today model everything statically. This is akin to assuming infinite time-smearing, as if it doesn't matter when exactly the forces are applied to the system being modelled. In practice, and particularly in audio, timing is everything. Music comprises of impulses. Audio engineering tries to tackle this by quantifying things in frequency space. We do this because simply studying the time-domain behaviour only gets us so far, absent the appropriate tools. The limitations of frequency domain analysis are well known on this forum; It assumes linearity and only reveals first-order distortions since the transformation into frequency space is a first-order function. Second order effects (compression or ringing) are not caught. Time resolution/smearing is not revealed. Etc. The number of possible ways to distort a waveform is infinite, and the more complex a circuit is, the more distortions are introduced.

So I suppose my perspective is that the scientific/mathematical zeitgeist is misguided. We model everything statically. Schiits oversampling filter at least tackles a problem of this kind. Therefore I respect them, their mathematical intuition, and their products. I also have high hopes for their future designs, since they are at least barking up the right tree instead of adding moar taps, or any of the disgusting ideas from that slide deck. It's not that they are ineffective, it's that they are half baked. You may have heard the adage "If you like your sausage, don't find out how they are made". Think of me as a sausage enthusiast who sadly *does*know how the sausage is made: I still love eating them, but sometimes I cry about it.

In general people still think in "for-loops" and "arrays". These are simply special cases of recursion and graphs, and are comparatively limited in their power. You cannot code any algorithm that reduced to a graph-colouring algorithm using for-loops. You need a higher degree of expressivity for that. What makes that slide deck so ugly is the total lack of awareness of or understanding of these facts. So, they crudely dedicate a whole bunch of silicon on the chip to an array of looping structures to mitigate first-order distortions (non-linearities) only to incur higher-order ones. What they need is a general solution such as described in the paper I linked. Such a technique mitigates all distortions in general, and could provide a basis for modelling DAC behaviour and dynamical systems in general. @earnmyturns is right that I am not making a merely aesthetic judgement. It is deeper than that. Rather, I believe we don't understand these things deeply at all yet. His comparison with other fields such as AI is perfectly appropriate. I do not think we understand AI either. I predicted here that we there would be far better ways of doing training transformer models, and described a number of obvious ideas.

One day this technology will be mature. It will look like: perfectly optimal use of underlying manufacturing substrate or parts, based on provably optimal math/algorithms optimizing for the constraints determined by a complex dynamical model of the mechanisms of the ear. However, we are nowhere near that. It's okay. This conversation brings us closer.

 

All this tis very interesting. Will the singular deliver the detail and clarity that I get from my HQPlayer>Holo chain - that type of 'clarity' that allow a person to hear the attack, inner note nuance, and decay in that beguiling "three dimensional" sort of way, whereas Yggdrasil's multibit (I have A2) has many other pro's but at the end of the day is "flat" when compared? How much is this advantage (for my preferences) related to 1 bit vs multi bit modulation?

Also, when is taco-fi happening again, would like to drive over if timing is right...

 

This is perhaps the largest blind spot there is. Analog circuitry has seen impressive developments in recent years, even op amps sound good nowadays.
Listening to MSB's totl dacs I believe they have cracked the code and come far already. They are naturally reserved when inquired about the tech (I have chatted with the guys on many occasions) - I suspect there is something in similar direction going on there. The sound is nothing like 'best' SD or 'best' R-2R. It has the best characteristics of both, none of the flaws and then some.

 

Fascinating! Thanks for sharing. Yeah, it's quite believable to me that some smart people have already solved this. I guess it will just take time for this tech to be implemented at scale.

Yeah. Overpromise, underdeliver. That's the Nvidia way.
I am super bummed that the AMD cards do not have UHBR 20 so I can pair with a TOTL high refresh monitor. Instead, this forces anyone targeting high-refresh but avoiding DSC to buy Nvidia. However, it seems there are a number of frequency upgrade-cycles all happening at once due to a number of simultaneous technological revolutions. Oled is enabling far higher refresh to be worthwhile thanks to faster pixel response times, mice are becoming far higher polling due to maturation of wireless tech for competitive mice, keyboards are now higher polling too due to hall-effect and "rapid-trigger" type tech. RAM and VRAM is also higher frequency thanks to PCIe 5.0. Therefore I think it could be worth waiting for one or two more years until all these things stabilise. phosphor will solve oled burn-in, GPU-to-display connectivity will become cheaper and more stable, mice and keyboard tech will necessitate the Mobo manufacturers figure out how to provide very stable fast usb connections, etc

 

I think we're all rooting for Schitt to overtake MSB in the DAC space. I'll never understand the logic of "we made an objectively amazing product, now let's price it out of the range of 99.5% of the people who'd want it". Sorry if that's stating the obvious.

 

I believe it is for marketability purposes, specially to appeal to asian markets. In fact, now MSB is not "fancy enough" because dealers such as Audioexotics don't carry it, they sell even more niche DACs instead. One example that first gained traction in Asia and more recently, in the US, is Wadax (which, IIRC, is even more expensive than MSB). Wadax even had an entry level DAC called Arcadia, at around 15-20k eur (yes, 20k eur was entry level) and had to discontinue it because they barely sold units compared to the higher priced models, despite using trickle down technology.

Speaking about Wadax, have you heard some of their DACs? There's people on WBF who switched from MSB Select 2 to the top Wadax, all claiming the latter is even more "natural" sounding. Kinda intriguing since the Wadax is not R2R but delta sigma AFAIK, using a custom ASIC chip. Maybe Schiit tried to achieve something similar with the Byygy? I didn't hear Wadax or MSB yet, but I find the sound descriptors from all the Wadax users quite interesting, one would think they are describing a well done R2R DAC and not a stereotypical delta sigma implementation.

 

Too stereotypical if true.

 

It indeed is comically stereotypical. Had to search a little bit from memory, but from just a quick search:
https://www.whatsbestforum.com/threads/wadax-dedicated-forum.34579/post-915767

Arcadia discontinued, coming from one of the first persons in the US who bought the top Wadax model (and used the MSB Select 2 before).

Audioexotics (asian dealer) Wadax threads, predating US and EU dealership and customer adoption by almost an entire decade, second thread even acknowledging it:
https://audioexotics.vanillacommunity.com/discussion/10649/wadax-aes-first-computer-audio-component
https://audioexotics.vanillacommuni...worldwide-wadax-atlantis-reference-dac-server

Wadax is a spanish company btw, it basically started selling to Asia. If you do a search on the Audioexotics forum (which is almost 2 decades old), you'll barely find mentions of "expensive but mainstream" brands other than for comparison to the more niche brands.

I can't find it now, but there was a thread on WBF where someone said that a dealer told him the Arcadia model was discontinued because the profit margins were "tight" and didn't sell much. One Arcadia was 6 months on Audiogon to be sold at a discounted price. Literally almost no one cared in the US:
https://www.whatsbestforum.com/threads/wadax-dedicated-forum.34579/post-916140

To not derail this thread more, I'm kinda curious how the Byggy could be similar to the Arcadia or the other Wadax DACs but on Schiit's market framework (focused on high value for the price) and mentality. It seems like the most similar match in the market at first glance (assuming that the Wadax sounds like people say consistently and that Byygy has a similar sound, since it is implied as very different from traditional delta sigma, like the Wadax).

 

Ehh, this is all relative. Depending on your income a Schiit Byggy is a Ferrari.

 

So if I want to keep the ability of my Yggdrasil (OG chassis, Unison upgraded) to swap back in my A2 board, I would just purchase Singular card for $1k correct? New ROM would (I assume) recognize/drive the older A2 card...

While I'm curious about Unison 384 (Unison 2.0 as near as I can tell) it would really have to be something to justify the jump to $3k full 'Byggy' and I don't see anything else (e.g. forkbeard, SPDIF upgrades, etc.) that interest me.

However, maybe the Singular card is gimped without the DSP/etc. of 'Byggy'?

 

Wtf?

 

Lavry and Bricasti have the gold casework for a reason.

 

Doesn't seem so. Over at HF it was stated that the high-capacity DSP card is there for "future functionality". Maybe Forkbeard-related DSP stuff.

 

Not sure why you disliked my post. Are you insulted? Confused? If this is not the time or place for a technical/explanatory post, just let me know.