The improvement is instantly obvious in particular I can hear better bass definition and clarity of vocals.
It is a vast improvement to my ears over the 1st generation.
My wife also instantly heard the improvement on first listen.
In Italy retailers no longer have Mu-so qb. I took one available and new at a good price … So the new qb2 comes out in September for the Berlin fair and it will cost over 1200 €. For a contraption so it is worth investing only up to a certain amount, as much use it in the bedroom and secondary environment …
And yet still no tone control.
1 Like
A what?
1 Like
It has 3 room correction settings and a loudness control.
I am considering buying the Mu-so 2 for my living room since available space is very limited and I would appreciate some feedback from owners or those in the know…
- Do you think that the sound will suffer considerably if the Mu-so was placed on a shelf?
- In the event that I can justify the cost of two of these, is there a way to set them up to be grouped together so that one unit is playing Left and the other Right? I assume I can do this with Roon via DSP settings and Grouping, but I have never tried. Is it possible to set this up without Roon? If so, I’d appreciate receiving any info.
- How would you rate the sound quality as compared to active bookshelf speakers in this approximate price range (of a single Mu-so) – up to the Elac Navis ARB-51?
[EDIT: I prefer neutral and natural sound and my musical taste is broad – but no metal, EDM, or most popular pop… :-)]
Thanks!
Just experimented in Roon with creating a grouped zone consisting of an iPhone and an iPad and applied a DSP (Procedural EQ --> Mix --> map L to both L+R and R to L+R for each device respectively)… despite how bad the iPhone and iPad speakers are for music listening (especially that they are not matched), the result definitely gave a much better stereo image. However, I believe there may be a timing issue between the two. It is difficult to tell because the two devices sound so different.
Anyone else want to try this out (if you have two Roon Ready / RAAT devices in close proximity) and give their impressions?
I seem to remember that there is a latency adjustment somewhere in Roon, so if they don’t play in sync, it may be possible to use this to get it right?
You are right – there is. But I wonder if this should happen at all as I believe RAAT is supposed to handle syncing. I also wonder if the offset might change from time to time thereby necessitating adjusting the latency between the speakers frequently.
I have also posted this int he Roon community, and I will post their response about the syncing here if it is meaningful.
2 Likes
It strikes me that using £2500 worth of Musos to create a stereo pair isn’t a good use of them.
They take up a lot of space and I suspect an Atom and suitably chosen speakers would sound better and take up less space.
I would be interested in if this would work with QBs as there are some pretty cheap ones knocking around at the moment. Already owning one and using Roon would make a pretty good 2nd system.
Thank you for your input. I understand that it is not ideal either in terms of value or acoustics (most likely), but the space available will not fit traditional bookshelf speakers.
At this point I am still researching what is technically possible.
Thanks again.
I guess it should be possible for Naim to add stereo pairing as an option. Apple do it with AirPlay, but as far as I’m aware, only for their own Home Pod speakers, and others such as Yamaha and Sonos do too. If you want Naim sound quality, though, I’d be inclined to go with the idea of an Atom. Or on a tighter budget, I reckon you could pick up a Unitiqute and a pair of used Neat Iotas for the price of two QBs.
1 Like
I was just reading the latest reviews on the Mu-So 2 and came across this quote:
The Naim’s native DAC resolution is 24bit/88.2kHz…
(https://www.stuff.tv/naim/mu-so-2/review)
The native DAC resolution is 24bit/88.2kHz, although the Mu-so 2 is capable of downscaling files of up to 24bit/384kHz via UPnP or wired/wireless network connection, and 24bit/96kHz via its digital optical input.
(https://www.trustedreviews.com/reviews/naim-mu-so-2)
That is a little disappointing since I would be using Roon as the primary source and have fair share of my library in hi-res. It is also a surprise as the Naim website mentions support up to 384kHz (a little misleading) and even:
From 32bit high-resolution streaming to multiroom music to HDMI TV sound; you’ve never heard it so good from an all-in-one wireless system…
While I understand that it might still sound “fine”, for me, and especially considering the cost, it is further cause for pause.
That wasn’t the meaning of the poster.
Do you this information verified by Naim?
I have no idea if it is verified by Naim, but as Naim’s site posted links to these reviews I assume that Naim would have corrected any misinformation.
I assume that Naim provided the reviewers with that info, but I would love to be corrected.
Hi @adeeb and forumers,
I would like to give some clarity regarding how Muso MK2 works, as a true end to end digital active system works differently to a full range single output analogue system.
The Muso system can be seen as 3 processing stages. Streamer engine, DSP that does active crossovers and optimises the signal for each drive units characteristic, then digital modulators that drives discrete output drivers, that finally drive the drive units.
Muso MK2 streaming core can natively accept files and streams upto 384kHz, with a depth of 32bit per sample. We use a dual core 1.2GHz iMX7 CPU for this task which is quite a serious processor. Data format wise it’s either 32bit IEEE float which represents the value of a fixed point 24bit sample value, or a 32bit fixed point value that is actually a 24bit value that is padded to 32bits. We accept all the permutations. From Roon we accept all data natively up to 384kHz and upto 32bit payload per sample which is leading edge for this type of product. The Roon audio chain view represents accurately what is going on and has been certified by Roon labs.
The next stage of Muso is that we have a dedicated multicore DSP that takes this signal and does an active crossover, drive unit management and getting the signal optimised for the class D amp modulators. The DSP does its internal processing at 88.2kHz, using 32bit maths which results in 24bit fixed point samples feeding to the amp modulators. At this stage we don’t process at ultra high sample rates as we know the physical frequency response of the various drive units and the art is to feed each drive unit as optimally as possible - not feed them signals out of their physical operating parameters, or have huge amounts of wasted digital processing in the system.
Finally the DSP output goes through a dedicated multi-channel modulator which drives a set of dedicated class D power drivers and output inductors. This is the stage when the bitstream become analogue and feeds each drive unit. This final stage is similar to how DSD works.
Naturally all the above is quite a science and an art to sound great, but when implemented properly can give excellent performance for the price, as all the components are optimised to work together.
Best regards
Steve Harris
Software Director
9 Likes
Thank you for the detailed explanation @Stevesky!
While a part of me still would have been happier if the internal DSP was performed at the full resolution, I trust that you had your reasons.
I wonder if in your testing — given the internal DSP rate — is there still an audible difference between feeding the Mu-so 2 music at 192kHz or even 384kHz vs 88.2kHz, assuming the same track and mastering? Also, would you recommend any up/down sampling in Roon for the optimal results?
Thanks again for taking the time!
1 Like
Hi @adeeb,
When taking a stream of sample rate x, its often quite undesirable to do all the active crossover processing at some uber high sampling rate. Everything has to work harder = more power consumed = electrical noise floor in the product goes up = typically negative influences in sound… then right at the end of the audio chain, it then has to be all filtered off anyway. In the world of cars, that’s the equiv. of a lot of tyre smoke and very little motion!
As for listening to different sample rates, typically if adding external sample rate conversion stages, the user ends up listening to the filters/distortions introduced in the algorithm, so it’s hard to appreciate what is being listened to and often comes down to personal preference on a given recording. It’s why great mastering engineers spend a lot of time getting downsampled material to 16/44.1kHz to still sound good.
On Muso we’ve spent a lot of time getting the audio processing chain optimal, so the end user doesn’t have to. As a rule of thumb, I would suggest sending the source material over natively to Muso, rather than upsampling it, converting to DSD or similar. Primarily it adds more filters into the audio chain on Roons side and gives the Muso more unnecessary work to do.
Best regards
Steve
7 Likes