Schiit Eitr Preview Thread

Don't you mean:

5 > 3 >> 2

Except that with Eitr, the USB on the receiver is not used so the USB version on the receiver is moot. So, as Torq points out, the only difference is the DAC when Eitr is used.

 

I had Jude down as more of a Judas Iscariot character, collecting 30 pieces of silver from the HF site sponsors.

Ordered my Eitr yesterday. Already shipped and on the way to my part of the planet. Looking forward to comparing it to my Singxer SU-1. If it tops that device (as per the review comments on page 2 of this thread) then it will be another amazing value win from Schiit, right up there with the Yggdrasil.

 

Eitr ordered! cannot wait to finally hear what my Yggdrasil can really do.
thanks schiit for the insane value of the eitr and thanks to all other and Torq for your many insight into how the Eitr compares to the Aries or other streaming options.


Edit: is there a way to convert the coaxial output to a toslink output? id like to also buy a second eitr for my ec design mosaic t that only accept spdif Toslink input... Please help?!

 

The Eitr handily bests the Singxer units. In my case I didn't even really have to A/B them.

The Singxers are merely the latest iteration of Xmos and they still have that sort of whooshy vapid USB harshness. They are better and smoother sounding than earlier Xmos solutions though.

The Eitr has a sense of "rightness" that I have not heard via a USB connection before and it's a very different sort of quality than is provided by the Singxer.

 

Not surprised, I just checked the layout of that... a big honkin' routing FPGA in the quiet section, with it's output pins going straight to the outputs, no flip-flop reclocks on the i2s or spdif post FPGA that I can see!

 

Brilliant - thanks for the feedback. I am "moist" with anticipation from the Eitr.

 

Thanks @Wfojas for a fun morning of listening at his house. We got to compare Berkeley Audio USB, Auralic Aries, and Schiit Eitr on his speaker system.

The one thing that seems definitive is that the Berkeley sounds the most natural of the three. The sound was fuller, voices more chesty (LOL) or corporeal/physical, and -- perhaps as a result of this -- the sound felt more dimensional. The BADA was somehow smoother and more detailed than the rest. I really liked the BADA a lot, sounded the most "analog". Worth noting that we plugged his microrendu into both the BADA and the Eitr and streamed roon, and that worked really well.

In terms of Eitr vs. Aries, we both liked the Eitr better. They both sounded flatter compared to the BADA but the Eitr was less flat and had a bit better depth than the Aries. Eitr also seemed to resolve detail better than the Aries. But the functionality of the Aries kills, so there's that, depending on your needs. To quote Willy, the heart wants what the heart wants. From memory, the Eitr definitely bests the Singxer stuff, and the Mutecs as well. Wish we had a Lynx on hand but i'm an impetuous asshole so I didn't have mine anymore. Hopefully someone will do this comparison or else ship us one to play with.

So here's a brief summary:
Detail: Bada > Eitr > Aries
Body: Bada > Eitr ~ Aries
Dynamic loudness: Bada > Eitr > Aries
Soundstage: Bada > Eitr > Aries
Price: BADA > Aries >>>> Eitr

If anyone wants to take issue with any of this, I'll paraphrase The Simpsons and say that this was all at this time of year, in this part of the country, localized entirely to Willy's living room. Btw, Willy feel free to add more detail, nuance, wisdom and maturity to this post.

 

I'm wondering if this is due to the filter on the Aries. The 5.1 beta firmware lets you have access to them.

 

Interesting. I had a chance to listen to BADA back to back with a Singxer F-1. I found that I preferred the Singxer F-1. This was on a headphone setup using unmodified HD 650 headphones. I felt like the background was cleaner and there was less of a veil over the music. I just ordered a Eitr to compare to my Singxer F-1 on my 2 channel setup.

 

Why eitr, when you can go G5 & skip the spdif shenanigans!

@cskippy the bada's got those crystek clocks I see, hmmm, simple arch + better clocks?

The concept and layout is very similar to the eitr, the eitr's decoupling itself from the CM6631A's performance... and the DigiOne's decoupling itself from the WM8804 performance (along with *overkill* decoupling from the pi). It's really interesting.

 

In my case? Flexibility! I have found it nice to have a USB->S/PDIF DDC around. I do like that the Singxer F-1 does not use a wall wart. Plus, my desktop system now uses a Modi Multibit which certainly benefits from a USB->S/PDIF DDC. At some point I suspect I will self update my Yggdrasil to Gen 5 USB. But, I have no plans to use it with USB. The streamer I will be getting has AES/EBU out (and no USB at al)l so I will use that.

 

Cool sounds perfect, grab a G5 though when you can!

 

My measurement queue overflows but I thought I would toss this in for a quick peek.
Setup:
Dell E7470 laptop with USB3 only
dScope soundcard output to Schiit Eitr via WDM driver
Vaunix Lab Brick USB Hub
Audioquest Forest USB cable
Schiit Eitr input to dScope via spdif
Tecnec 75 ohm BNC cable and Tecnec BNC to RCA adapter
44.1 KHz, 24 bit
Noise floor with -40 dBFS 1 KHz sine is below -175 dBFS !!!!:


dScope setup (partial, there are many more control panels not in view at the moment)


Now go and buy it.

 

Wow! Just effing wow!

 

I'm not Torq, but it makes more sense to upgrade Gungnir Multibit to Gen5 USB rather than put Eitr in front of it.

 

Torq responded to the post completely. Your post adds nothing......

 

While -175dBFS sounds amazing, this is just digital data, so one would expect it to be perfect. As a control, what does the dScope with a loopback and same test signal look like (ie. its digital output straight into its digital input)?

BTW, 16k FFT points = 2^14 buckets. So that's 14*3 = 42dB more noise if you add up the entire measurement bandwidth. -175dB+42dB = -133dBFS, or around 22 bits of resolution. 24-bit data should be around -142.72 dB with TPDF dither, but there's probably some eyeballing of the noise floor in the measurement, or something else is going on.

 

USB quality seems asymptotic. It gets closer to AES, but may never get there. But who's to say. Try them both if possible, I guess?