HiFiMan Susvara
Oct 18, 2020 at 4:34 PM Post #4,892 of 18,858

Ciggavelli

Headphoneus Supremus
Joined
Mar 24, 2019
Posts
2,305
Likes
4,633
Location
Houston
How is Chord Dave plus Mscaler?
I love it. You of course need an external amp for the Susvara, but since the DAVE and M-scaler are so detailed, you can add an external amp without a noticeable degradation in resolution
 
Oct 18, 2020 at 8:12 PM Post #4,894 of 18,858

Shahrose

Headphoneus Supremus
Joined
Dec 14, 2006
Posts
3,979
Likes
405
Location
Toronto, Canada
I've only tried susvara with WyWires and AC Silver. I prefer the silver cable. the WyWires is no slouch, but I'm not a big fan of its build and ergonomics. That deal sounds OK but depends what u need the adaptors for. Otherwise, without extras, for the usual price, I'd rather get the silver cable.

I have X6 in gold+silver, and X8 in gold plated silver (don't know what sonic differences they would make considering both have gold and silver, but can't make 1:1 comparison as X6 is iem and X8 is headphone). Shared characteristics I would say are smooth highs and lots of bass (rounded and a bit boomy, rather than tight), crazy super lush and thick mids.

I may be just a silver kind of guy, but I find a well implemented silver cable plays well to the Susvara's strengths.

I'm with you. I almost always prefer high quality silver with headphones, but especially with the Susvara.

The realism silver imparts with its clarity, quick transients and extension on both ends is rarely matched by copper wires IMO. The design matters ofcourse, but the conductor usually makes the bulk of the difference IME.
 
Last edited:
Oct 19, 2020 at 1:15 AM Post #4,895 of 18,858

sahmen

Headphoneus Supremus
Joined
Apr 6, 2014
Posts
2,707
Likes
2,272
Welcome to the club. Your chain is very close to mine (Ambre > Morpheus > Class A amp) and I am sure that the sound is exceptional. But for others who may be putting together a similar chain from scratch, from a system budgeting perspective I am curious about the sonic benefit of the optical module and etherRegen and a couple of expensive LPS's. Everything up to the Metrum is asynchronous data in packet form and the Ambre comes with optical isolation at its ethernet input. The pi is further galvanically isolated from the signal processing and timing sections. The two clocks in the Ambre are very good. I2S is the optimum connection to the Morpheus. If you are streaming from the internet, the packets have already made their way through a high number of copper and optical switches. I have some difficulty understanding how the last one can become sonically important. I did a test at my home by replacing the wired ethernet with an optical link straight from my HP managed switch (which has optical ports) to a media converter at the Ambre. It is the same switch that my Synology NAS is connected to via a 4-ethernet trunk. So for FLAC files on the NAS this is about as direct and isolated as one could get. Maybe it is my tin ears, but I could hear no difference. We agree on the Metrum, Morpheus, and Susvara (which is the heart of the matter) but I would suggest that someone building this chain from scratch spend the remainder of the system budget on the highest quality and most system-synergistic class A power amplifier that they can get. I am just another anonymous opinion on the internet, but my ears can't hear ethernet and the stock Hifiman cables on the Susvara, though they have a cheap rubbery feel, are good bi-metal silver/copper cables and sound fine.

Thanks for your warm congratulatory message. Your questions about the use of my switches and power supplies are asked in good faith, so I welcome them. I am no engineer, but I know from some readings I have done that there are various types of electrical and digital noises (or to use the usual jargon, "common mode noise," "leakage current noise," "ground plane noise," "clock phase noise," "jitter," etc. etc.) that can plague an audio signal, and that not all these types of noise can be successfully addressed by one component alone.

You sound like someone with a mind for techie stuff (an electrical engineer maybe?) so I shall direct you to this paper by John Swenson of Uptone 'audio, who is the "guru" from whom I have garnered what little I know about all these types of "noises," and related "gremlins" and "ghouls" one might run across on the pilgrimage to audio nirvana :

>> Link John Swenson's white paper <<

Two more things:

1. I am supplying this info because you asked, and I am assuming you did so in good faith. I am not trying to convince you to buy anything or change your mind about the resolutions you've announced about not buying anything else.,

2. As for the Etherregen, it works, period, at least to these ears. I noticed its effects in minutes, if not seconds after placing it in my chain, which I ironically thought was already very resolving and practically noise-free. Well, it is not the first time I have encountered this irony of noticing the existence of certain types of noise only when they're no longer there, and my audio black background (which I mistook for the optimal in "black ') has become even blacker. (hence the joke about "gremlins" and "ghouls" :slight_smile: :slight_smile::slight_smile:)

There is a discussion of John Swenson's White paper on the Audiophile Style site, in case that interests you :

>> Link: Discussion of white paper <<
 
Last edited:
Oct 19, 2020 at 4:00 AM Post #4,897 of 18,858

MatW

Headphoneus Supremus
Joined
May 20, 2020
Posts
3,162
Likes
8,946
Location
EU
Atom to M scaler with USB or optical. I use USB. M scaler to Dave with double BNC cables. I use Wave stream. Dave to the amp with XLR or RCA, depending on whether it's balanced or not.
 
Oct 19, 2020 at 10:29 AM Post #4,899 of 18,858

beemerphile

100+ Head-Fier
Joined
Sep 10, 2015
Posts
166
Likes
212
Location
Danielsville, GA USA
There is a discussion of John Swenson's White paper on the Audiophile Style site, in case that interests you :

>> Link: Discussion of white paper <<
I am not an electrical engineer. My background is that I am a 40 year chemical engineer, but most of those have been in process control. Additionally I am a HAM radio operator and have learned a bit of electrical theory from that. There are many here whose electrical training and understanding exceeds mine and I don't present myself as an expert - a reasonably well-informed hobbyist, I guess. Regarding the "white paper", my personal tendency is to never buy anything that is sold on the basis of a "white paper" from the seller, and I am a guy who has even written a couple of white papers. Let's go through the noise types that you mentioned one at a time...

"common mode noise,"

Ethernet is, by design, inherently robust against common mode noise. Common mode currents flow with the same voltage and direction, and go to ground by the parasitic capacitances between the lines and ground. There is no common mode current flow through the load of a balanced differential receiver such as is contained in an ethernet transformer because there is no voltage difference on the load. The only way to induce noise into the unshielded twisted pairs of an ordinary ethernet cable is to unbalance the line. The magnetics of the ethernet connector contain chokes and shunt capacitors to control common mode noise caused by normal amounts of imbalance. Extreme imbalance could be caused by cable, connector or device problems - in other words, problems that don't generally exist, but should be found and fixed when they do. Almost any system built and tested to EIA/TIA standards is sufficiently resistant to normally experienced common mode noise. I've seen ethernet systems strapped to welding robots that functioned without error. No house has this level of EMI source in it. It is possible for a poorly designed switch to have unbalance caused by something like PC board layout, but bad design is fixed by design improvement, and not an add-on box. Such a product should fail in testing.

"leakage current noise"

There are two types of ethernet cables - shielded (STP) and unshielded (UTP). Hopefully you are using UTP so there is no opportunity for a leakage current to flow through the shield (also known as a ground loop). The only path to ground in an intact UTP circuit would be by parasitic capacitance of the transformer in the connector. I haven't measured and don't know the ground impedance of a typical connector, but I would expect it to be quite high - meaning it would take a very high frequency for it to conduct any troublesome quantity of leakage current to ground. Remember that equipment is tested for resistance to EMI and that any noise below the noise floor of the system is inaudible by definition.

"ground plane noise"

We are again talking about a noise source that is not universal, but one that only exists in significant levels in poorly designed equipment. Don't buy any of that.

"clock phase noise" and "jitter"

In general terms, jitter is a variation in timing of the signal edges from their ideal values. Phase noise is just a particular type of jitter where the noise is off-center of the carrier frequency, so I'll comment on these together. A clock with high jitter (of any type, including phase noise) can cause high bit error rates in an ethernet system. Bit errors in an ethernet protocol (whether UDP or TCP) are managed by a cyclic redundancy check (CRC) within the frame check sequence. If there is an error within a frame, it is re-sent. Errors are common within an ethernet system and so long as the errors do not exceed the maximum error rate that the process can handle (due to its speed, buffering, etc.) then there is no consequence. If the maximum error rate is exceeded, there will be a dropout. In audio systems, a blank or a click. It will not be an incremental change in sound quality that gets better and better as jitter is reduced. Sound.... No sound. A baseline level of jitter exists based on the accuracy level of the clock that is used. The solution if this is a problem is a better clock. That will not be a problem for any commercial ethernet equipment because it would not have passed its testing for throughput. As far as the other variable sources of clock jitter caused by such things as load, noise, heat, power supply fluctuations, etc., so long as the error rate that it produces is below the jitter tolerance of the system, it is of no consequence. Why? Because the timing is thrown away and re-created in the next step - where it's accuracy very much matters to the incremental quality of the sound. This is what conditions us to the mindset that less jitter is incrementally better, but that shouldn't be used to sell us an ethernet device. Ethernet packets don't even necessarily arrive at the receiver in the same order they were sent. Some were in error and had to be re-sent. Some arrived later by a different routing. Variations in network traffic and QOS rules within switches and routers can be causes of this. Having received this mixed up packet stream in its buffer memory, your Ambre goes to work unpacking the data and assembling it in correct order. It then paces this ordered data stream with a different and very important clock for presentation to the DAC. The Ambre takes clock accuracy very seriously. It has two separate high quality Tentlabs clocks for 44.1 and 48 kHz (and multiples thereof). Jitter in this clocked stream matters incrementally rather than by threshold.

So why did you and so many people hear an improvement that I do not believe can exist? I don't know, but I don't doubt you and the many others who believe this device to be useful. That is why you bought it, kept it, and recommend it to others. They say that if the Wright brothers had been to college they would have known that they couldn't fly. That means it is very possible that I am the one in error here and for that reason, I should not state my case too strongly. I hope that I have presented it clearly, rather than strongly. I am not telling anyone what to do or think. I could be wrong. Regarding the money back guarantee that this type of equipment is so often sold with, that does not persuade me to spend the effort to try it. If you offered me a pill with a money back guarantee that would allow me to lift the corner of my house I would not buy it, because I do not believe it to be possible for it to work. I will not spend the effort to try it. My doubt here is not to that level, but is sufficient to keep me from a "buy and try". I did my test with an uninterrupted optical link from the SFP port of an enterprise grade managed switch to a media converter on my Ambre and heard no improvement. Maybe it just matched my confirmation bias, but I buy and will continue to buy my network equipment from the best network specialists. I buy my audio equipment from the best audio specialists. I am a large fan of Nelson Pass and Cees Ruijtenberg, but I would not buy a network switch from either one of them any more than I would buy a Cisco power amplifier or DAC.

It is easier to show that a difference exists than it is to show whether the difference makes a difference. It is easy to show that oil filter A only traps 4 micron particles and that filter B is better because it traps 3 micron particles. But does the engine care? If the engine lasts 300,000 miles with the "bad" filter will it last any longer with the "better" filter? That is more difficult and maybe impossible to show. It may well not matter at all, making the "better" filter only better by measurement and not by function. I do not doubt that the differences exist in jitter, etc., but I have shown why I do not believe that the difference makes a difference. Showing two pictures of someone's eyeball and pointing out how under close inspection that this one is different and clearer than that one has been done and leaves little doubt that there is both a difference and an improvement. But at viewing distance with the frames passing by at 30 per second is there any outside chance that someone would see it and recognize it as better? I think not.
 
Last edited:
Oct 19, 2020 at 1:11 PM Post #4,901 of 18,858

MatW

Headphoneus Supremus
Joined
May 20, 2020
Posts
3,162
Likes
8,946
Location
EU
Do you use Uniti Atom?
No, I have an Auralic Aries G2. You called it a streamer so I assumed it was a similar device, but now that I looked it up I think it has a DAC built in. If it cannot be bypassed, and looking at the connections I suspect that is the case, you cannot combine it with the Dave. Maybe someone who knows the Atom can chime in.
 
Oct 19, 2020 at 3:55 PM Post #4,903 of 18,858
Joined
Mar 20, 2013
Posts
14,019
Likes
9,581
Location
Location: HQ in the UK, but spread far and wide.
The Susvara replaced an HekSE in the same chain, and it took me less than one minute of listening to notice the improvements in SQ performance, although I am still a very ardent admirer of the HekSE and wish I have not had to trade it up for the Susvara...

Yes, it is an instantly noticeable upgrade. Not "a bit here and a bit there" but immediate jump up, at least that's how I see Susvara vs HEKv2 :)
 
iFi audio Stay updated on iFi audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/iFiAudio/ https://twitter.com/ifiaudio https://www.instagram.com/ifiaudio/ https://ifi-audio.com/ info@ifi-audio.com
Oct 19, 2020 at 4:13 PM Post #4,905 of 18,858
Joined
Mar 20, 2013
Posts
14,019
Likes
9,581
Location
Location: HQ in the UK, but spread far and wide.
Any one heard of Nativ Vita Streamer?

I rember seeing it a long time ago as a "work in progress" development that looked really good and was up to date on functionalities. At this time it should be available I imagine, but I have no idea how it sounds.
 
iFi audio Stay updated on iFi audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/iFiAudio/ https://twitter.com/ifiaudio https://www.instagram.com/ifiaudio/ https://ifi-audio.com/ info@ifi-audio.com

Users who are viewing this thread

Top