"What a refreshingly honest blog about listening to music through hi-fi. So happy to see views based upon the enjoyment of music rather than so-called sound 'quality'." - Peter Comeau, Director of Acoustic Design at Mission / Wharfedale

Friday 17 June 2011

Streamers and DACs from Linn, Cyrus, Logitech, Rega and Naim

Spent an interesting couple of hours in the hifi shop this afternoon with Naim enthusiast Richard trying some streaming / DAC combinations.  The excuse was created from having to pick up my newly serviced Cyrus DAC-X.

Objective - is it the streaming or the DAC that makes the difference?

Kit tested:

Linn Majik DS (as streamer and with internal DAC)
Logitech Squeezebox Touch (as streamer and with internal DAC)
Naim NDX (only as a combined unit)
Cyrus DAC-X (no PSX-R)
Cyrus Stream XP (as streamer and with internal DAC)
Rega DAC

System consisted of Cyrus 8 XPdQX integrated amp and a pair of Cabasse Artis Baltic Evolutions with their subwoofer.  Not sure what the speaker cables were, but the interconnects were from Atlas.  For the comparisons we used 16 bit CD rips in FLAC from Florence & The Machine and Elbow.

Lets get the speakers out of the way first.  I would describe them as an acquired taste, but its not a taste I acquired in the 2 hours in the shop.  They're too high, leading to a rather distant and detached soundstage and the subwoofer isn't well integrated with a somewhat underwhelming bass that lacks any real weight or slam - that could be down to how well they were set up of course.  They do have an excellent mid-range though, producing some really stunning female vocals.  The top end is ultra detailed, but can lead to some brittleness, particularly with some of the combinations we heard this afternoon.  They were, however, quite capable of showing the differences between the source components - the notes below refer to subjective differences and personal opinion - its not about absolutes.

So we kicked off with the Touch / DAC-X which I normally listen to at home, albeit it normally has a PSX-R playing along.  The sound was good, within what I think are fairly compromising limits of the speakers with some confusion on complex passages.

Next up is the Linn as streamer and DAC and here we get a decent step up in performance - more detail, much more space around the instruments and better playing of the tune.  The complex passages are better resolved too.

So which is making the biggest difference - the streaming bit or the DAC?  We hook up the Linn as a streamer into the DAC-X and find that this makes the music somewhere between the Touch / DAC-X and the complete Linn solution.  So its true, the streaming 0s and 1s bit really does make a difference and the Linn as a streamer is a worthwhile upgrade over the Touch as a streamer.

What next? We went back to the Touch and used its internal DAC.  Its very polite, a little muddled and dynamically a little flat.  Let's try that Stream XP with its internal DAC and see how Cyrus's latest performs.  Well, its better than the Touch, and gives a good account of itsself, but its not as big an upgrade as you would expect for £1400 extra investment.  So we go back to Touch / DAC-X for a while and then to the Stream XP as streamer only feeding the DAC-X and we find that the Stream XP is a better streamer than the Touch and the DAC-X is significantly better than the DAC inside the Stream XP.  This is a good sounding combination, but at £2600 (I'm allowing for the DAC-X's list price when it was new, rather than what I paid for it) it's not competing with the £1880 Linn.

The Stream XP was a bit of a disappointment as a complete streamer / DAC, but performed well as a streamer into the DAC-X and equalled the Linn /  DAC-X but not the Linn as a complete unit.

So in comes the NDX.  And here we have the typical Naim sound - very detailed (probably a very small amount of detail more than the Linn) and very pacey.  But we agreed it was all a bit too much and rather aggressive - going back to the Linn was a joy - much more musical and the music has time to flow much more than it does on the Naim.  Even Naim fan Richard agreed that the Linn was producing the best sounds we'd heard.

So that was the main session that we'd planned to go through.  Richard's interested in a DAC for his second system that currently uses a Touch only so we gave the Touch / Rega combination a try.  This is a good DAC - very refined and smooth but I was hoping for more, having seen the reviews and forum recommendations.  Keeping the Touch as a streamer we compared the Rega and DAC-X DACs, finding that the DAC-X is a better sounding piece of kit in this particular system - being more detailed and more dynamic without being overpowering.

We finished with some more tracks on the Linn including some 24 bit stuff.  Interestingly the difference between a 16 bit CD rip and a 24 bit version of the same track brings a much bigger difference on the Linn than you can differentiate on the Touch.  Some great music playing here, within what I think are some pretty severe limitations in the speakers.  Still need to hear it in my own system of course.

So here's our conclusion for you to pick apart!  Best to least best with prices just to add a bit more interest:

Linn Majik DS streamer and DAC    £1880
Linn Majik DS streamer and Cyrus DAC-X     £2880
Cyrus Stream XP as streamer and Cyrus DAC-X     £2600
Naim NDX streamer and DAC     £3000
Touch streamer and Cyrus DAC-X     £1200
Cyrus Stream XP as streamer and DAC £1400
Touch streamer and Rega DAC     £700
Touch streamer and DAC     £200

We didn't try the Cyrus Stream XP / Rega DAC combination.


  1. Great read, thank you!

  2. Subjective comparisons are flawed and meaningless. If you wanted to really know if there were differences, the test HAS to be level-matched and blind. The brain too often hears what it wants to.

  3. Agreed, to a limited extent. I think for small nuances you are correct, for major changes I think subjective A to B is enough. For what its worth, in this test, I wanted the Logitech / DAC-X to sound best as it was the combination I already owned and therefore would have both been giant-killing and cash saving. So you might think that the brain hears what it wants to, but in this session it conclusively didn't hear what I was very conciously hoping for. How does that fit with your point?
    Besides, we made a lot of comparisons here in 2 hours. To do this fully objectively, with measured level matching and blind we'd need a couple of days. Unfortunately listening to music is a leisure activity for me - I don't have the luxury that might come from being a professional reviewer.
    Also, for what its worth (which might not be very much), I can quickly tell what I like the sound of and what I don't and believe that this is much more valuable than reading tests in magazines, which is all the majority seem to do.
    I hope you enjoy reading the blog, despite the subjectivity, and I'm pleased you didn't fall into the trap of objectivity meaning only measurements count for anything - as you know, we don't know everything we need to measure yet, otherwise everything that measured the same would sound the same.
    Can you please post some links to reviews that you consider are adequately objective, I'd be interested to see what there is to learn from them.