Meridian Unplugged - Powered by MrTechGuy.co.uk Forums Discussion Meridian - Streaming MD600+MS600 review

How very odd. I find it hard to believe he is really hearing the differences he claims.

I am awaiting delivery of an MD600 and MS600 and will try out the two options he did. My plan is for the MD600 to be in the basement where it belongs, but I have plenty of space on the network switches to try it upstairs as well.

This review is not in context of an all Meridian DSP or analog system, which would have been interesting if it were. The Meridian philosophy as I understand it, is to avoid the sort of "Hotrodding" that is being practiced here.
With that in mind we will have to try tweaks for ourselves, but I am not expecting to gain much. I did try isolation feet under the MC200 and heard no difference.
Time will tell, Chris

I have always loved the Meridian approach that kept digital to the last possible moment in the chain. With Meridian I thought the days of different support systems and mains cables and interconnects was consigned to history (and the analogue boys!).

I was at a HiFi show recently and there was a room selling bamboo supports. For sure there was a difference between the glass, wooden and bamboo stands and I left the room smug in my belief that with a Meridian digital system, I did not have to worry about all that stuff.

Maybe not...but I am sceptical for sure...
/Marcus

For my money, putting the MD600 directly on top of the MS600 makes no difference. Placing isonodes underneath both in the same configuration certainly does - more musical sound. And the biggest surprise of all, using Russ Andrews Classic cables for both makes a big difference for the better. I also use isonodes under my 5200s - and yes, they make a difference too. thought I had left all of this messing about behind when I recently bought my system! It now sounds fantastic.

Hi Clem,
What exactly is the science behind this? Blind A-B testing under controlled conditions?
I am still skeptical about these claims for an all digital system and I would love to understand how the last meter of cable from the mains wall socket to your equipment can have such a marked effect when the analogue circuitry is actually in the speakers and not the "front end".
If your findings can be proved empirically then surely an even better sound would be had by tearing out all the copper in the mains rings in your house and replacing them with some sort of braided and twisted mains cables.
Has any HH here done this?
Best regards
/Marcus

Exactly my thoughts Marcus. Although I am not critical of people using such leads as the placebo effect (the only explanation I have so far) is very powerfully and routinely used in medicine.
I would love to have a proper provable scientific explanation, it would mean I could economically get even more from my systems.
Thoughts, Chris

AB testing-can of worms
Power cables-can of worms

If you start off from will a perfect supply (constant voltage, perfect sine wave) improve the sound (performance) of the electronics (I think yes) and work backwards, see where you get.

I can't help but think that spikes and variations in voltages are sub optimal. I think it was someone here who said- undo your mains plug sockets (powered off) and look at the black marks on the plug connectors, evidence of shorting. That can be good.

Ears always have it.

It is a fascinating topic when we wrestle with what our ears tell us vs what the theory/spec sheet tells us.

FWIW - I agree with the theory both here and in Ronnie's question to HQ via Rick which says it should make no difference (where a router or how many routers are involved).

However, my suspicion is that there IS something here that we (collectively) do not quite understand yet.

I will be sitting on the fence for a while I think until I get time to move a few units around at home.

I think the suspicion is that the network cabling could be producing differences. From memory Meridian found that network cabling made a difference (in the 808.3) and now the ID41 is shielded.

I think I can hear a difference between an MS600 hard wired to router (long cable) and having it connected via a bridge (Airport Express). If Ian tells me I can't, my ears will probably subsequently agree with him... There's too much psychology between expectations and what I "hear". eeek

Why might an MS600 sound better than an MC200??...................and most that have compared them think that the MS600 sounds better???

It essentially comes down to lower RFI / EMI levels within the MS600.......as it doesn't have a spinning drive or as many other RFI Generating components within it as the MC200

It remains likely that it is the RFI that is causing the Pollution of the Master Clock within the MC200 and therefore creating higher levels of Jitter on the MC200 Outputs


RFI is a well known side effect of moving Data around at Mega and Gigahertz levels.........and therefore RFI is a well known issue on Gigabit Networks

Is it not reasonable to assume that at least some of these high RFI Levels is finding its way into our Audio Endpoints from the Network Cabling...........and that the RFI could contribute to Jitter on the Digital Outs and internal transfers of PCM Data???

Thanks Ronnie, all clear.

Looking at my switch, the signal from the MC200 is Giga and the signal to the MS600 is Mega. Presumably that's deliberate re above? The MC200 core may be supporting a number of zones and therefore need Giga, the MS600 "quieter"?

So shielded cable to prevent antenna pick up and ferrites to reduce throughput?

At the application level, TCP network data, which is what it is (not audio) will (simplistically) behave as follows:

Sooloos sends audio data as PCM packets over TCP protocol. These packets will be received by the audio endpoint thus:

1. It will be received 100% correctly.
2. It will be received 100% correctly, but may be truncated. Subsequent reads will read the remainder of the data as per these rules. Truncation is entirely normal operation and may be caused by a number of things, such as exceeding MTU size in network equipment.
3. The data may not arrive.

It is impossible for an application to receive data that is 99% correct. Its either 100% correct, or nothing. Bits is bits at the application level of networking.

In #3 above, the effect is 'catastrophic' and will be clearly audible. I use the term 'catastrophic' to mean that it is an error that the application has to act upon to resolve rather than something like a thermo nuclear meltdown.
In #2 above, truncation is normal expected behaviour for long packets, which the PCM packets will be. It is possible that there is a bug which does not reassemble truncated packets correctly. I can see that this could have a subtle audible effect.
In #1, #2 above, the application will have received the data 100% correctly, but at lower layers (OSI 7 Layer Model), things may be quite different and things like error correction, retransmission, packet reordering may have come into play. But this is all transparent to the application and in the end, the application will receive 100% correct data, or nothing. There is no reason why a wired home single subnet network should not be 100% reliable 100% of the time, thus if error correction etc comes into play, then there is a network issue somewhere. Note that I also say wired home network. Wireless and powerline networks work quite differently and retransmission, error correction etc may occur on a frequent basis.

What we cannot see with Sooloos (sorry Meridian but Sooloos has kind of stuck) is the effects of non optimal network. Can we hear the effects?

When data is waiting to be received, the wait is often subject to a time out - I don't know whether Sooloos is or not. If the timeout occurs, then it is a catastrophic failure and the application has to resolve the problem. This is unlikely to be inaudible.

If data is received late, but there is no timeout, then receipt of the data may fall outside an acceptable timing window that the Sooloos audio endpoint can handle - possible with non wired networks. We also know that jitter affects audio. However, we are talking about a network packet here, typically holding 1500-4000 samples, not just a single sample. I believe that if the packet is received significantly late, then the result will be clearly audible as an audio glitch. For slightly late packets, bearing in mind that it is travelling over a general purpose network which will have other network traffic, I believe that at most, the samples at the packet start (and maybe end) may be subtly audible, but only for a few samples and entirely randomly. I don't buy this bass is better stuff, its going to be everything.

What if there is a bug in the software or hardware, or in network equipment? Well, it is conceivable that the application will receive network data that is not 100% correct but is told otherwise. I do not believe that this is an issue as it would appear that Sooloos uses much standard equipment and software (Busybox etc) and bugs such as this would have manifested itself in many other ways. It is possible that Meridian have written their own network stack, but I suspect it will have been based on a port from existing code. Finally, this is easy to test and I would like to think that network transmission has been tested independently of audio.

Can something else be at play? Well, I don't understand about emf, rfi, cables, ferrite cores, shielding and things but there was a pretty good thread that made me believe that this can all have a difference. But not at the MD600 end.

So, my 2p's worth.

I don't buy the moving NAS boxes, or putting them on feet stuff. They are totally decoupled from the audio endpoint (ensemble and MC200 excepted) and I just cannot see anything that can have an effect, short of introducing problems (bending a cable too much etc) into the 100% reliable 100% of the time network which may or may result in something randomly audible. Busted or go fix your network.

I don't buy into the one router/switch sounds better than another. If you stick with a switch, you get network speed access between ports - remember that it looks like audio endpoints are limited to 100mbs connections anyhow. If you go for a router, you should be using it as a switch if you connect it to the row of RJ45's on the rear, but if you aren't, say connecting to wan port, or you have an oddball switch that routes some of the ports, then you will get non network speed access as the packets are being routed at slower speed which may cause problems. The latter applies to some routers with wireless access points, where the wireless LAN may be routed. That said, Sooloos network requirements for audio are quite small. Busted on a wired network, plausible on a network with errors or on those that transmit audio over wifi or power line adaptors.

Can the design of the audio endpoint and cables feeding it have an effect. Plausible.

Can there be a software/hardware fault in Sooloos making it not behave optimally under non 100% conditions. Plausible but I doubt it as the issues would be far reaching and not just subtly audible.

Can there be a 'special sauce' in the core that makes one behave slightly differently to another. Meridian say this is busted.