dCS Network Bridge to Chord M-Scalar

Having bought myself a Tron 7 GT phono stage and being absolutely wowed by how marvelous my LP12 / Aro / Troika / 'Geddon sounds, even with ancient and far from silent vinyl, I have become disenchanted with my digital setup. But, was I getting the best from it?

dCS developed the dual AES output and wrote the standard, with which Chord have complied. Therefore, in theory, I should be able to connect the Network Bridge (NB) dual outputs to the Chord M-Scalar (MS) dual BNC inputs.

A quick trip onto eBay and I had bought:

  • Two Van Damme 1m AES digital cables (no expense spared);
  • Two neutrik 110 to 75 ohm impedance transformers (AES > BNC); and
  • Two male bnc to female RCA gender benders.

This enabled me to use my current selection of spdif cables.

Having set up the physical hardware I fired up dCS Mosaic. This is a bit more complicated than it sounds as I partition my physical LAN from my wireless. This meant that I had to find an Android emulator from my Win 10 laptop that allowed network bridging = Nox.

Via Mosaic you can set the ‘Dual AES Sample Rate’, I selected 352.8K. I had set up a playlist which included 10cc Rubber Bullets (RB), from 10cc Greatest Hits CD Rip.

I listened to the playlist via my usual set up using NB spdif out to MS BNC 1. I had started using Mosaic as I think it sounds a tad better than Linn Kazoo, but I can’t think of a logical reason why, and so I am still playing with this.

RB was a tad too hot on certain plosives, as usual. But this was still much better controlled via the AQ Diamond spdif than any other cable I had heard.

With the dual AES set up I repeated my listening. It worked! But wait, the MS was indicating that the input was BNC1, not dual BNC, and the input it was seeing was 44.1K.

Now, I had made an assumption. That by selecting 352.8K in Mosaic that the NB would upscale the input stream. IT DOESN’T. As dCS support kindly explained to me, the selection you make determines the threshold at which Dual AES kicks in, and the minimum is 88.2. So, counter intuitively to me, 88.2 is the selection to make. IF the input is less than this sampling frequency then the NB out puts the native stream via BNC 1.

Having made the change on my playlist I had included two tracks from LPs I had ripped at 2496, these correctly were played via Dual AES.

A bit later I played Bat Out of Hell (BOOH) 2496 from Qobuz. It sounded perfectly reasonable. I then played one of my LP rips also at 2496. This sounded bloody good.

As the day was drawing to a close I realised I had enjoyed my digital session more than any I could recall for some time. Still no conclusions.

Day 2

Having cogitated I decided to try upscaling my CD quality rips to trigger Dual AES …if possible. My experience with upscaling is not favourable. I have found it to ameliorate plosives but also blunt detail and dynamics. But, nothing ventured etc.

I was aware that the version of Minimserver I am running was behind the curve, but I am a great believer in ‘if it ain’t broke’. Now I bit the bullet, upgraded the server, reinstalled Minimstreamer and Minim Watch. This is an aspect of digital audio that I am getting increasingly impatient with, but I understand why it is needed and improvements can be delivered, so it proved here.

I have to pay for Minimstreamer periodically, but the level of support and development makes this worthwhile for a pretty minimal cost. To date I have simply used Minimstreamer to transcode my flac to wav. Now I added an entry to additionally upscale 44.1K to 88.2K. Now I get permanent 88.2K.

I tested the 88.2 via my nearfield system using Linn Kazoo. When I went back to my main system and played via Mosaic using the playlist I had set up the day before Mosaic reported 44.1!! I checked the M-Scaler (MS), IT was seeing 44.1?? I went direct to the album via Mosaic, 88.2 and Dual AES …weird.

I am also going to look at upscaling the 44.1 music from Qobuz via Bubbleupnp.

Time for some more listening. Ultimately will this prove just an interesting distraction, more difference than improvement? Good question.

M

Blimey. Liking your work @MrUnderhill :slightly_smiling_face:

Borne out of frustration :wink:

Maybe I misunderstood… why using dual bnc for 88.2. Why not go much higher bitrates? Especially when 382 is going over Single connection.

This really doesnt make sense if your using the same signal path from the UPnP server for both as they are control apps only and dont touch the audio. But you mention your using Bubbleupnp server? If so then Kazoo uses open home which may be conecting to an open home version of minimserver thats routing via Bubble if you set it up that way, where DCS Mosiac may be standard UPnP and likely going via Minimserver direct? Something to look at.

:laughing:

1 Like

Hi @CrystalGipsy ,

I COMPLETELY agree.

Yes, Kazoo is using Bubbleupnp Openhome, that may be having an influence as you say. Good observation.

M

1 Like

Interesting response from dCS:

Thanks for your email. What is happening here is that MinimServer uses a ‘virtual’ file system, so changes made to a library don’t mess up any UPnP renderers looking at the server. Some servers can become very confused when a change is made to the library, so for example adding new content to the UPnP server causes any playlists (such as those in Mosaic) to become scrambled with random content. The virtual file system avoids this happening.

So, when you set up MinimStreamer to transcode the files, a new virtual location was created. Browsing to the album finds the new transcoded location, but the playlist was created before the transcoding so till uses the old locations.

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.

Well, having listened over a protracted time I would say that:

  1. There is no difference between feeding the dCS NB with Mosaic or Kazoo;

  2. That I consistently prefer listening via the NB when the AES output is used, regardless of whether it is using the dual or single AES route.

No, this doesn’t get to the emotional heights of using the LP12/ARO/Troika/Tron GT, but it is a damned fine second best.

M

The network bridge is very good, shame that dCS no longer make it

1 Like

Chord Mscaler has no dual BNC inputs

It has dual BNC outputs only !

Recheck the manual.

BNC 1, BNC 2 and Dual Input.

In my case if the input stream has a frequency < 88.2 KHz - Input is BNC 1
If frequency >= 88.2 KHz - Input is Dual BNC.

M




2 Likes

I think you’re wrong here and @Analogmusic is correct. I have an mScaler and while it certainly has DUAL BNC output, it has TWO (not Dual) BNC inputs - i.e. it can be fed by two separate BNC sources (BNC 1/2) as well as two Optical sources (OPT1/2) and one USB source. You can cycle through the required sources using the Input button, but DUAL BNC output is only available for the source connected to BNC1.

2 Likes

Hi @Stilts,

If you go to the M-Scaler manual and go to page 19/33 you’ll see the input colours. This includes: BNC1; BNC2; and Dual BNC.

When I play a file with a freq <88.2 the signal passes down BNC1, and so the M-Scaler input is yellow. When >= 88.2 the M-Scalar input is turquoise. Hence the photos above.

M

No it isn’t… The INPUT light is second from the left, and Amber=BNC1, the OUTPUT (OP SR) light is third from left and looks white in your photo (it’s not colour balanced) which equals MAX upscaling. The input sample rate is shown fourth from left and, in your photo, RED=44.1 kHz.

I show my MScaler twice, showing two different incoming sample rates.

First photo shows turquiise dual bnc input. Incoming sample date yellow, 88.2.

Third photo shows yellow input, bnc 1. Incoming sample rate, 44.1.

As per the nanual.

Well it doesn’t mean that the Mscaler accepts dual BNC input

Do you have a source that output dual BNC ?

Have you tested it ?