Holo Audio Cyan 2

Actually when I spoke with Miska, he confirmed actually on the PCM side Cyan 2 has the same number 64 resistors on each side as the Spring 3, but probably to cut costs they use cheaper resistors on the Cyan 2, and commented that the Cyan 2 on the PCM side does not measure nearly as well as Spring and May.

On the DSD side, the Cyan 2 uses some thing different in a so called vector step resistor network, so even though they cut down in the number of resistors in half, but the measurements with DSD get closer to the Spring 3...

 

No, as @atomicbob indicated in his measurements the Cyan 2 PCM measures admirably well also in distortion and dynamic range.

With DSD, that means you would use a software such as Hqplayer to convert your PCM source on the fly to DSD for playback. (It means you don't need to do the actual conversion offline, when you click play the program converts your flac or mp3 to DSD only for playback)

 

Ah, I had assumed the ladder with less resistors is the PCM portion due to PCM performing a bit worse. Very interesting.

 

Interesting indeed and dac designs and implementation are way beyond my knowledge. But there are indeed pretty detailed measurements on the roon hqplayer forum comparing PCM input VS DSD input and DSD input at various rates to confirm that DSD256 measures best overall. DSD512 pushes ultrasonic noise further out compared to DSD256 but suffers a bit more noise in the audible band...

 

No, "this" (measurements - @dericchan1 would you be so kind and post a link) I am guessing (based on @atomicbob measurements posted so far) are a relative comparison of very good (PCM) vs. very very good (DSD), which means unless your a measurebator the decision to consider the Cyan should be based on other factors, such as the sound character of the DAC, how you plan on using it, how you plan on feeding it, etc.

Yes, your library (and any streaming you do) is most likely 99.99% PCM and compressed (i.e. mp2, AAC, etc.). The Cyan is a NOS (non oversampling) DAC, so you will have to understand what that means and if the consequences are for you. Since I don't like the sound of NOS 44.1/48k content (standard 1x, what is on your CD's) my recommendation is that you up sample your library to at least 88.2/96k. If you don't know what I am talking about you need to do some basic reading before making a decision to purchase any NOS dac such as the Cyan 2. Don't worry however, the learning curve is not that steep...

 

Sor

Sharing what I have read so far

https://community.roonlabs.com/t/best-native-dsd-dacs-for-use-with-hqplayer/132298/800
Starting from post 800 and on there were a lot of discussion on the Cyan 2, and measurements between PCM and mainly DSD

Cyan 2 jitter performance using USB input is usual Holo Audio quality, absolutely excellent!

 

Adding support for these comments: I own a dac2541, a Spring 2, and a May (different systems), For 44.1/48k PCM, Holo NOS sounds very lush, but ultimately somewhat smeared in the bass-to-midrange region. Before I installed HQPlayer, I would use Roon upsampling to 88.2/96k to tame the smearing. But HQPlayer conversion to DSD256 with the filters and modulators of your choice is where the Holo DACs really shine, and the only change that convinced me to sell my OG Yggdrasil. As for dac2541, it is a delightful PCM DAC that does not reach the clarity, tonality, and spaciousness of the most more complex and expensive HQPlayer>Holo setups, but is hard to beat if you want a simple, high-quality PCM solution.

 

Cyan 2 has excellent jitter measurements when using Dante Virtual Soundcard to spdif input at 88KHz, Bal output:


This is considerably lower than the DSD 256 reported above. Not that this is much more than academic.
PCM users need not suffer FOMO.

Additionally spdif in, Bal out, multitone performance:

Yes, there is a tiny increase in the noise floor above 10 KHz. This is at approximately -140 dBFS,
0dBFS =16.147dBu for Cyan 2. So the noise is at approximately -124 dBu or around 500nV.
That translates to approximately -23 dBSPL for most headphones and maybe 8.5 dBSPL for insanely sensitive
Westone UM2. Almost any amplifier is going to produce more noise and mask any residual noise from Cyan 2.

 

New YouTube video review just posted on the Cyan2 from New Record Day. Ron has listened to the Cyan 2 for months and compares it to his reference May KTE and concludes they are very close.

Ron is my favorite YouTube reviewer whom his comment I completely trust!!!

I have to believe Ron has good ears and know what to listen for. His room is well treated and the rest of his gear is up to the task. So as much as ppl looking at measurements and claim there is some gap in performance between the Cyan 2 and the Spring or the May, these differences just may or may not be totally audible or very subtle…

I think what makes Ron's review on the cyan 2 even more impressive is the fact that he didn't even test with dsd and hqplayer that everyone has measured would bring the cyan 2 much closer to the spring

 

I'd love to hear some impressions vs. BF2OG. Kthxbye!

 

Take his words with a pinch of salt. There’s so much he got wrong about the Yamaha AS-2100.

 

Well he got it spot on for the Freya +, the Benchmark stacks, Holo May KTE and now the Cyan 2

 

Thanks for the clarifications! I've ordered a Cyan 2 to replace my E70V, looking forward to hearing the NOS PCM. And I'll try to experiment with some DSD.

 

Added "2nd Update: HQPlayer (via Roon) PCM & DSD playback" to OG post

TLDR: getting a headless PC running Roon > HQPlayer for PCM > DSD upsampling duties has been worth it...I did not want it to be so

 

Welcome to the club PC specs?

 

Relatively light. I3 14100T (a 35 watt 4 core design) which runs @ 2.7 Ghz, 16 G of memory, a single SS drive, can't remember case manufacturer at moment. Even with HQP up sampling to DSD512 it's only pegging CPU to 60%. Jessi seems to have Linux better worked out, as the load is utilizing all the cores more or less evenly. All for under $1k shipped.

Ubuntu Linux keeps losing the USB lock on the Cyan however, so I need to figure that out or get a Red or something...

Edit: 60 percent @ dsd256 I think, will check when I get home. Does dsd512 however as well, will check CPU load....

Edit 2: I was mostly right the first time, DSD256puts about a 20% total CPU load, DSD512 puts about about a 60% total CPU load, though a couple of cores are near maxed. It takes DSD1024 to bring the computer to its knees (playback starts stuttering) Tested with 44.1 PCM > DSD256 via poly-sinc-gauss-long & ASDM7ECv3

 

U t

I think the iFi streamers and possibly the Rendu stuff will pass through the higher sample rates, too.

 

I have ultra Rendu running as NAA and it can do native DSD and PCM up to 768Khz as per spec but I never tried it as am limited to 384Khz due to DAC limit.

 

The

The Rendu stuff doesn't seem super popular on here, but I reckon they're pretty plentiful/cheap on the used market these days.

 

I should point out that with the new hqplayer dac output signal correction feature that provides improvements to the cyan 2, it does require more cpu processing.

it did take like a good 30 seconds for my i5 13600k to load the correction settings on my Cyan 2 but whatever it is Miska had done, it seemed to actually sounded cleaner at least on some of the recordings…

also the new switching pcm/dsd on the fly and changing sample rates on the fly also takes a bit more processing