Schiit Eitr Preview Thread

Yeah, that's what I was doing. Except I wasn't enjoying it as much as I have been for many months now, something was off. So I looked into it, it took minutes to find the issue, and was reproducible on later comparison. This "ain't my first rodeo".

On the other hand, I could just be full of shit. I never thought a shiny cable would do anything of use. But then if I'm right, it's more likely the lack of a poor cable that solved it.

 

How did you come to the conclusion that the difference you hear is Eitr? If you compare without and with the Eitr, is there less improvement than you first experienced?

 

Actually, the conclusion I came to was that it was an old thin random phono RCA cable that was the issue. Which is fundemnatally neither analogue or digital, though probably adheres to different manufacturing standards.

Before anyone posts again please go back and read the prior posts. The last two are replying to older ones in the chain, which are either less or no longer valid. Apologies to everyone, next time I'll take more time to problem solve before I post, and cause less clutter in the thread. I was caught out by "but it worked fine yesterday" thinking.

TL;DR Modi Multibit is better with Eitr, maybe don't use one channel of a rando phono RCA as a coax SPDIF.

 

I don't know what fundamentally is the difference but yes, a coax SPDIF cable has to have a certain rating/quality to do it's job properly. For example, the RCA cables that Schiit sells on their site weren't good enough. I went out and bought a Blue Jeans 6-inch cable (with their bendable wire) and it's a perfect fit for my Eitr to Modi Multibit.

 

Can you elaborate on this?

 

so I've read enough of this to get the sense that this is somewhat better than Auralic Aries LE to Yggdrasil via AES, but I like the convenience of the Aries. Would this plugged into the USB on the Aries be as foolish as I get the sense it would be? would there be a benefit?

Just wondering.

 

I found that the AES output from the full-fat (femto clocks, lower-noise internals, low-noise LPS, more RAM) Aries just edged out the Aries LE. I also found that the Eitr, regardless of source, sounds essentially indistinguishable compared to AES output from the full-fat Aries.

By that measure, which is triangulation not a specific test, I would expect Aries LE -> Eitr -> Yggdrasil to be better than Aries LE -> Yggdrasil.

The difference, if detectable at all, will be tiny ... audition-style listening required.

Depending on what functionality you're using on the Aries LE, you might be able to skin the same cat with a Sonicoriber SE + Eitr, or an RPi + Eitr, and save a little money.

 

eh.. I got the LE pretty cheap, and it works, and I whole heartedly believe you that the difference is tiny.
I think I'll stick with what I have.

Roon endpoint and Airplay.

 

Odd, been working fine for me, including at 24/192. Don't know. My PYST cables are several years old, so maybe there has been a production change, but the Eitr manual says something to the effect of sure you can use a 75 ohm impedance cable, but implies don't sweat it.

 

I emailed Schiit and asked and they said they aren't rated for SPDIF.

In general, RCA cables may work, but they aren't rated to work. As @BenjaminBore found out, they aren't perfect.

If you want to know all the reasons why it does or doesn't work, I'm not your guy. Mostly because I don't care and I wanted a cable that knowingly meets the spec.

 

I would use the Blue Jeans Cable SPDIF coax cable. They are the only cable that I've seen described as 75 ohm end to end (including the connectors). Even if I could not hear a difference, it would end any audio nervosa I may have about the SPDIF coax cable.

http://www.bluejeanscable.com/store/digital-audio/

 

So the thing to be aware of for digital transmission is that you need enough bandwidth and noise margin so the waveforms can unambiguously be read at one level or another for long enough (eg. the setup time for an input requires a voltage level to be held above some level long enough to be registered as a 1). This means that while a questionable cable can work in one setup, it may not work in another if it doesn't meet certain specs. Some environments might be friendlier and give you more leeway, while others aren't. A cable that meets its operating specs will work pretty much all the time anywhere. Given a friendly enough environment, even a coat hanger will work (as that famous experiment showed). The specs are there to ensure the cable works in the most adverse conditions.

The characteristic impedance (the 75 Ohms) of the cable isn't important at this cable length, because at 10 MHz, which is the ballpark of an S/PDIF transmission, the main harmonic of the squarewave is about 30 meters long, so a short (like 1 meter) cable looks like a point to the waveform and basically the voltage across the entire cable rises and falls at the same time along its length. Even the 5th harmonic of the square wave at 6 meters is going to be pretty long compared to the cable, and it's already pretty low in level.

What might affect transmission are the bandwidth of the cable: it could be rounding off edges of the square waves so the logic threshold is harder to read and unreliable. The cable's lumped parameters (capacitance, inductance, resistance) might be interacting with the receiver or sender's circuit, and acting as a lowpass filter. Lowpass filters also have phase shift, so you can get dispersion where the waveform's constituent harmonics spread out and separate: this can make it harder to reliably read the right level at the right time.

And the noise protection can affect things: how shielded is the cable, and how much noise it lets in. These things affect the operating margins of the receiver circuit: digital is inherently pretty robust, but you can degrade the signal enough or add enough noise that you might start to lose bits. Combine this with the stuff above, and it could add up to a situation that's not great.

But it shouldn't be hard to get reliable transmission. A composite or component video cable or other 75-Ohm rated coaxial cable should be perfectly reliable. I only say 75 Ohm not because that particular number is important for short runs, but because cables that have a rating like that tend to have specs and construction that will reliably support S/PDIF.

 

Cables rated for use with SPDIF are designed to meet bandwidth specs that analog cables aren't supposed to match.
That is why analog cables will sometimes work, and sometimes not.

SPDIF and AES are pushing squarewave signals with MHz frequency harmonics.
With analog cables, you really don't want to pass these frequencies along, let alone collect and then add them to the analog signal.

And those 50Ω or 75Ω spec'd ratings for those SPDIF cables is part of that expanded bandwidth.
Analog cables have an 'unspecified' impedance.

JJ

 

I like these cables and use them everywhere. The other important thing about their RCAP connectors is that the shield has a 360-degree, fully-circumferential termination to the connector, so there is no gap in the shield. With regular RCA plugs, you normally gather the shield into a pig tail, and then solder it to a lug pin inside the RCA body. That has two problems: the shield just went down into a pigtail, so there's a big gap in the shield between the RCA body and the rest of the cable (and the metal body of an RCA doesn't help either since it's not connected to the shield continuously), and the pigtail can be an antenna to introduce noise into the system.

It's also for this reason that I think AES/EBU XLRs are a dumb way to do digital because not only is it pretty much impossible to build an XLR jack with 110 Ohm impedance (this is determined by the physical geometry of the connector), but almost all XLR jacks require a pigtail termination of the shield. Neutrik makes one where you can crimp the shield around its full circumference, but almost no one uses that connector.

 

This is the bit from the Eitr manual that @Darren G was talking about:

"1 Coaxial SPDIF Digital Output. Connect this to your DAC with a high-quality cable. If you’d like, you can even use a fancy '75 ohm' version, though this doesn’t make a huge difference in performance, unless you’re using some unshielded, hand-braided thing that’s inappropriate to digital."

But yep, to get the very best out of your Eitr, it's probably sensible to pick up one of the Blue Jeans cables, or another similarly spec'd cable.

 

Hope @schiit starts selling PYST spdif cables.

 

Do I dare bring up the topic of SPDIF cable length? Some people think SPDIF cables should be at least 1.5m long to avoid reflections causing jitter, and it's been a topic of debate for years. Nervosa or legitimate concern?