SOtM Trifecta
Posted by: Mr Underhill on 10 October 2018
My front end for the past couple of years has been the Sonore ultraRendu > SingXer F1.
Last weekend a friend came over with his DCS Network Bridge. This had a tad more resolution and a tad more dynamics - as well as greatly simplifying the spaghetti of cables & LT3045s I have cleaning the DC. All for £3,250.
This got me thinking, the thoughts including:
- Is it really worth an additional £2k?
- What about the Auralic G1 or G2?
- Which effects the SQ more, the F1 or the uR? and
- How good is the SOtM Trifecta?
So, I have bought the SOtM Trifecta with a 30 day money back guarantee, this comprises:
- SOtM 200ultra 12V with clock input;
- SOtM sps500 PSU; and
- SOtM txUSB 12V with clock input; and
- DC Y cable - copper, they will send me the silver when it is in.
Will this simplify my system? NOT AT ALL. These three boxes will replace just my ultraRendu!
Reports to follow.
M
Brilliant posted:Are using the Celeron NUC? I have thought about trying one.
Yes, for the princely sum of £109.
I agree, the pace of digital audio change is breathtaking!
Thx,
M
Mr Underhill posted:nbpf posted:Mr Underhill posted:Hi NBPF,
The RPi3 that Metatron compared to the DCS NB that he owns is one with a Allo spdif board, comments above. This seems consistent with what Hans Beekhuysen is reporting. However, if you REALLY want to put the cat amongst the pigeons then look at using a cheap NUC with AudioLinux.
One of the things this session confirmed for me was the importance of the server feeding the endpoint. At one point I switched to using my NAS to strem the files, the immediate drop in SQ was plainly obvious.
M
Has Metatron compared the Allo DigiOne or the DigiOne Signature to the dCS NB?
The DigiOne Signature is a major improvement over the DigiOne, especially when its clean side is powered via batteries or through a LPS-1.2. It brings the same system simplification as the dCS. Thus, at less than 300$ it might be worth a try.
I have been following the latest development of the original "A novel way to massively improve ..." thread on CA. I have ran some experiments with real time kernels in the past and I am not very surprised that low-power, low-latency sources can dramatically improve the sound quality. I have seen there is also an AudioLinux version for RPis, I'll check if it comes with the DigiOne drivers.
It will be the DigiOne. Hans B. was impressed by the Signature, not up to the SOtM ultra in his opinion - from memory.
If you can get AudioLinux for RPi I would recommend it - esp. if it will work in RAM Mode - all the OS in memory, makes a big difference.
M
You are right, although the comparison between SOtM ultra and DigiOne Signature is problematic and Hans B. was not powering the DigiOne Signature with a LPS-1.2.
I am currently running latency tests on my minimal Raspbian installation. I am then going to get the AudioLinux for RPi and compare it with my installation. Do you know whether the latency plots on the AudioLinux web page have been computed for an idling system or for a system during replay?
No, but Piero is good at replying.
M
Mr Underhill posted:No, but Piero is good at replying.
M
Thanks, I have been in touch with Piero. As it turns out, the AudioLinux version for RPi does not yet have a RT kernel. I am currently testing RealtimePi, it's a project on GitHub. Anyway, my custom Raspbian installation seems to have very decent latency figures and the DigiOne Signature sounds gorgeous in my system. It also brings a great simplification!
I am very keen on having a listen at the new Bartok streamer. According to dCS streaming board on par with the standalone NW bridge and a Rossini level DAC inside - albeit with one power supply only (and we know in this forum that this will have a detrimental impact... but still... sounds like an interesting alternative)
DrPo posted:I am very keen on having a listen at the new Bartok streamer. According to dCS streaming board on par with the standalone NW bridge and a Rossini level DAC inside - albeit with one power supply only (and we know in this forum that this will have a detrimental impact... but still... sounds like an interesting alternative)
I am not sure that I would like to buy an integrated streaming board + DAC if the streaming board was not replaceable. Perhaps it is in the new Bartok. The dCS NB seems to me a very appealing proposition. The only reservations that I have are about the software. I do not mind if a device comes with a proprietary OS as far as this does not violate open source licences and the device is certified for an open source OS. But the experience shows that many commercial products come with outdated, proprietary software and very little support and documentation. Does the dCS NB run a real-time kernel? Will one be able to upgrade as this becomes available? These questions are of course not dCS specific, they also hold for the Naim streamers or for the Sonore network transports. It is conceivable that the next most significant improvements in sound quality are going to be achieved through software rather than hardware improvements. Or through a combination of both. Which companies are in the best position to readily integrate new softwares in their products?
DrPo posted:I am very keen on having a listen at the new Bartok streamer. According to dCS streaming board on par with the standalone NW bridge and a Rossini level DAC inside - albeit with one power supply only (and we know in this forum that this will have a detrimental impact... but still... sounds like an interesting alternative)
I'm sure it's fantastic, but at its price point (around £12,000) it is very expensive and if you are willing to spend just a little bit more then it will be competing against the likes of the ND555 and Linn Klimax DSM Katylist. Will it be better than those?
As a matter of interest, Mytek is apparently about to release the Mytek Brooklyn Bridge - a Mytek Brooklyn+ DAC with an incorporated streaming board. At around £3000, this will no doubt be nowhere near as good as the Bartok but it will be pretty 'good' if my experience with the Mytek DAC is anything to go by, and it's certainly a lot cheaper. Of course you will need a good external power supply as well. My experience with the Mytek Brooklyn + DAC is that a good external linear power supply is a mandatory purchase from a sound quality perspective. With a good linear power supply, the Brooklyn + is definitely on a par with, and possibly slightly better than the Chord Hugo and similarly priced products
I wonder if Hans B will review the Brooklyn Bridge anytime soon, and compare it with his sotm/Brooklyn+ combo?
I see it differently: the Bartok (without the headphone amp) is not that much different in price from an NDX2+XPS2 and it is against them that should be compared. I am personally disappointed that the new NAIM streamer range has not worked on the DAC section per se and expect dCS to have the edge here, soon will find out; it is the Rossini that should compete with ND555+PS555 and my budget cannot stretch to either of them
Good luck if you end up going down that route.
If I was looking for a solution for my main system for that sort of money (I use a Klimax DS/1 in that and am very happy with it) and I didn't have any curiosity about MQA, then the DCS Bartok would probably be pretty much on the top of my list of options as well along with perhaps some of the new Auralic products.
Although I haven't heard any of them, the dCS products certainly appear to be more or less unanimously praised.
Mr Underhill posted:Docv posted:Thanks for this update....the diagram was complicated for me but very precise.
It seems that the DCS NB is compelling and reasonable value looking at what it can replace: would I be correct in assuming that it cannot connect to a Hugo? Is SPDIF better than USB?
I have read up on the Audio Note 2.1, it does seem to be a wonderful DAC; do you prefer it to other DACs you have may have heard such as the Border Patrol, Mytek Brooklyn etc.
Are you now going to return the Trifecta and buy the DCS!
Cheers
Stewart
Hi Stewart,
I know, digital audio is a pig - everything makes a difference but the difference isn't always better!
I think the DCS NB is an excellent piece of kit, and I prefer it to the ultraRendu and SOtM stack.
My friend who owns the DCS NB uses it with the Hugo via its spdif.
From what I have read the spdif works better with the Hugo, but all these things need testing as it is your ears and your system. N.B. DCS promised that they would turn the NB USB input into an output as well, they have backed away from this. The hardware they use for the NB was unable to do this in the way DCS have managed this before.
I have returned the SOtM - but not without some thought. I am grateful to Elite Audio for their generous support.
I am currently trying AudioLinux, headless and in RAM mode - so the device boots from a USB stick and runs without a HDD/SSD. I have loaded this onto two devices, one acting as the stream aggregator ( a la uR) and one as my Logitech Media Server.
If I was to buy Sonore, SOtM or DCS the DCS would be the one ....and still may be.
Cheers,
M
Thanks M
Please forgive my ignorance but does the spdif connect to the coax or the optical on the Hugo?
Stewart
As you can tell I haven’t a clue about electronics. I bought a Nova recently and love it...as do the family. It’s in the kitchen where we spend a lot of time. It’s (deceptive) simplicity and operation via Roon is a joy.
I am allowed to disappear to my own snug occasionally where my developing DAC based system is slowly evolving. First stop is the dedicated spur being installed tomorrow. I like my olive 52/250...had them for years...the sheer effortlessness of streaming and prospect of huge leaps in SQ is driving my interest and ambitions. I was very much drawn to the sotm kit by Hans B...but the neatness of the DCS solution together with pretty much universal approval, and now your own observations are building up a persuasive case for me.
One thing at a time though!
Off topic, but as a matter of interest I wrote in an earlier post:
"With a good linear power supply, the Brooklyn + is definitely on a par with, and possibly slightly better than the Chord Hugo and similarly priced products"
I'm normally sceptical about the 'burn in' effect that many people seem to report in respect of audio equipment and cables. I have always believed that the effect is at least partly down to psychology or to a gradual process of 'the ear' getting used to a change of sound.
I now think I might be wrong, because I am pretty sure that I am experiencing the effects of 'Burn in' changes in a significant way for possibly the first time with my Mytek Brooklyn +. On the basis that there was no harm in accommodating the possibility of burn in, I have been running the DAC more or less constantly now for 3 or so days. A few hours after I posted the above comment about the Brooklyn+ sounding slightly 'better' than the Hugo, I almost posted again to say that I was completely wrong and the DAC now sounds more than a little thin and edgy compared to the Hugo. However, today it once again sounds very good indeed.
Not something I have really experienced before, although I have read elsewhere that both the Brooklyn and the linear power supply I use with it really do require a lengthy 'burn' in period in order to perform at their best. Of course it could still be a psychological effect, but I'll now be a little less sceptical when others post about 'burn in'.
I have definitely experienced a fairly extended burn in with the Nova...it was worth the wait
Re connecting a DCS to a Hugo...I can see that the DCS has 4 SPDIF outputs but how does one use them to connect to the Hugo?
Sorry if this is obvious.
Docv posted:Re connecting a DCS to a Hugo...I can see that the DCS has 4 SPDIF outputs but how does one use them to connect to the Hugo?
Sorry if this is obvious.
Assuming you mean electrical SPDIF rather than optical, Hugo has an RCA input. I don't know what the DCS has, but just get a cable that has the correct connections at each end. Hugo 2 has 3.5mm jack input instead of RCA.
Thanks Chris
Is that ×2SPDIF into the single RCA (on my Hugo 1) or 1:1...how is the stereo signal maintained (I know it must be, and with top quality on the strength of Mr Underhill' s posts).
Cheers
Stewart
Docv posted:Thanks Chris
Is that ×2SPDIF into the single RCA (on my Hugo 1) or 1:1...how is the stereo signal maintained (I know it must be, and with top quality on the strength of Mr Underhill' s posts).
Cheers
Stewart
Just a single cable is all you need. I use a Hugo from NDX to Hugo, and a regular Naim DC1 cable with BNC connector at the NDX end, and phono at the Hugo end, works fine - although there are cheaper alternatives!
Could Mr U please let me know what cable his friends use for the Hugo-DCS connection. My dedicated spurs are in and I’m dreaming of a musical Christmas.
Kind regards
Stewart
Hi Docv,
Sorry been wrestling with AudioLinux for the past few days!
I have asked my friend to either post here or drop me some notes for me to repost on his behalf.
M
Mr Underhill posted:Hi Docv,
Sorry been wrestling with AudioLinux for the past few days!
I have asked my friend to either post here or drop me some notes for me to repost on his behalf.
M
Great thanks.
Good luck with AudioLinux (I don’t know what it is but I will do some research!).
stewart
Mr Underhill posted:Hi Docv,
Sorry been wrestling with AudioLinux for the past few days!
...
Would be interested in your findings about AudioLinux. Have you been running the OS from memory?
Last week I compiled a real-time kernel for the Pi 3B+ and compared it to the standard kernel. I cannot hear any differences in sound quality but on the Pi neither the standard nor the rt kernel can run in memory.
Also, I have noticed that, on the Pi, the rt kernel effectively bounds the max. latency to values well below 100ms. But this comes at the expense of min. and average latency. The latter goes up from about 6ms with the standard kernel to about 9ms with the rt kernel. I have spoken to Piero and he assured that on more powerful architectures rt kernels also decrese the average latency. With the Pi, this is not the case.
Hi DocV,
My friend is using: '...... I'm just using a StudioSpares RCA digital IC ..... dCS SPDIF output to Hugo coax input.....'
Context: A home built massively over specified PC for hosting Roon Core >> DCS NB >> Hugo >> Modwright LS36.5 >> Pass Labs X250 >> WB ACTs + Torus.
M
nbpf posted:Would be interested in your findings about AudioLinux. Have you been running the OS from memory?
....
Hi nbpf,
Yes, I saw your posts on 'that' thread.
I have a lot of respect for the Pi. Moving my audio files from my Synology NAS onto StarTech hosted USB sticks into a LPSU powered Rpi 3 made a nice difference, further accentuated by the Audiosence Prestige 2, and further improved by replacing Vortexbox with AudioLinux.
I am running two boxes with Headless AudioLinux Extreme, one with LMS and one with Squeezelite. Bothg are running in memory, and that makes a lot of difference.
When I initially set up the Squeezelite NUC I had it sat next to me while I sorted out the BIOS and initial config, then I moved it into position and had a listen. It was not bad, but was overly sharp. I then SSH's onto the box and moved it into RAM, there was an immediate improvement.
Over the next few days the box settled down and on day 2 the bass arrived.
Two days ago a replacement R-Core 19V 5A PSU arrived and had the expected effect of removing a very small level of high frequency tizzzz that I was getting with the SMPS. I am currently just running it in before reporting back.
I have a level of complexity with my system as my Audio Note DAC uses valves, so I have to turn it off between sessions. This can cause the chain to glitch. Sometimes this can be in the software, sometimes in the firmware in the USB chain. This has caused me to do a level of digging to isolate where and what the recovery procedure should be. This was easier with the SOtM & Sonore boxes and there a few applets you could simply restart.
In terms of quality I would rate AL > SOtM & Sonore. WRT DCS, not sure - I suspect it is better, but I would like to listen for a longer period to the DCS. Either way there is a slight bar in that you have to be comfortable with the Command Line.
Next steps: Consider a new case to get rid the fan, although the box isn't any more than slightly warm; and, play with PoE - although I am enjoying what I am hearing now.
M
Mr Underhill posted:nbpf posted:Would be interested in your findings about AudioLinux. Have you been running the OS from memory?
....Hi nbpf,
Yes, I saw your posts on 'that' thread.
...
Thanks for the report!
I am currently waiting for a JS-2 to be delivered next week. That is going to complete the DigiOne Signature source and power both the dirty side of the DS and the LPS-1.2 that powers its clean side.
Once that system is settled, I'll try AL on my old fitPC3 + Eitr system and run it in parallel with the DigiOne Signature system. That should make a comparison relatively straightforward.
That should be interesting. I saw that John Darko was VERY impressed with the DigiOne Signature. Look forward to your thoughts.