Grimm Audio MU1 and MU2 Music Players

The digital front end has made a lot of progress in the last years. Besides the DIY solutions, it’s great to see innovative turnkey products hit the market. Just to name a few I guess Antipodes is long known for its DX servers and now CX+EX combo, recently Silent Angel has introduced an affordable Rhein server (~ €1.500) and -on the opposite side of the spectrum- the Taiko Audio SGM Extreme server has received the highest appraisal from users and reviewers (at €24.000, see recent reviewed at 6moons). It kinda surprises me that the Grimm MU1 server went under the radar here at AS, so -with more spare time at hand these weeks- good to share and introduce this server here.

Coincidence or not, in about 1.5 hour drive from Taiko Audio another Dutch high end manufacture is located: Grimm Audio. Currently mostly known for its ultra-low jitter clocks (aimed at studio’s with Guido Tent) and LS1 speaker system line up, Grimm Audio has recently launched its MU1 server (€9.800). In short the MU1 is a Roon core server with Grimms oscillators, the inhouse developed power supply, endpoint with an inhouse designed FPGA board as the magic ingredient (which converts / up- or down samples to the desired format). Like other Grimm products, the MU1 is upgradable overtime (upgrade of the existing components and placement of a internal DAC).

The most interesting thing for me is that the FPGA is aimed to match the outgoing digital signal with the ‘default/native’ bit depth/sample rate of the connected DAC. For me this was a new approach, only later I discovered that Auralic has a similar approach with the Sirius G2 Upsampling Processor (€ 6.300). I’m not an expert by any means so I prefer to link to an interview where Eelco Grimm gives a proper explanation (time stamp, English subs). Have I been sleeping for years, or have more servers/streamers applied this DAC matching-approach?

Quick links for a fast introduction/understanding:

MU1 Interview Eelco Grimm (English subs)
The “Pure Nyquist” filters of the MU1

Company overview Hans Beekhuyzen
Productpage and reviews (6Moon has also a review of the MU1 lined up)

Besides Ian Shepperd (host) and Eelco Grimm, Bob Ludwig and Bob Katz are in the panel. Tool - Fear Inoculum has won this years award. Great discussion going on.

Hi Fred,

Do you have shares in Grimm? 🤪

I heard the MU1 or perhaps the prototype at XFI in NL 2 years ago, i don't remember which speakers but i would suspect the Grimms.

To be honest i wanted to like it, i know Guido and despite finding him somewhat unreliable in our past dealings, i like him, however i am afraid i was severely underwhelmed, although there was a lot of attention on it, lots of people in room and a presentation going on.

I can only think the circuit was tweaked since then as the sound was cold, thin and un invloving at that time, details sure but i do not want my wine at 60% proof, you know what i mean... Well, at least at the beginning of a night 😉

Glad to hear they are doing well and sounding good

Enjoy!

Haha, no shares or affiliation, just intrigued. I haven’t heard the version in the 2018 XFI event, I believe it was their first prototype in public!? Sorry to hear you where underwhelmed.

I first encountered the MU1 at the 2019 XFI event, in the Grimm Roon (together with the LS1be’s) and in a room using Kroma speakers, Zanden amps and Mola Mola DAC. I really liked both! Later I listened at a dealer (BB&G).

Recently a firmware-update has been released, which apparently brought some improvements (haven’t heard this). When you live close to the Netherlands, perhaps you have a chance to listen to the current/latest version?

I’m keeping my fingers crossed for XFI / Dutch Audio Event this year, hopefully corona is on its return by then..

Looking at the innards it seems to have a NUC board for the server/streamer side of things.

That means "low power" which can result in reduced dynamics.

Not sure if the special clock can recover this aspect.

3 hours ago, HeeBroG said:

Looking at the innards it seems to have a NUC board for the server/streamer side of things.

That means "low power" which can result in reduced dynamics.

Not sure if the special clock can recover this aspect.

That sounds like rather an over simplification/generalisation.  In my experience power supplies along the whole streaming chain have a major impact on dynamics. Especially important are the micro dynamics, as they affect note shape and development.....which is where a lot of the magic needed to convince us  lives.  SQ is a lot more than just low vs high power.....EMI and RFI, circuit topologies, oscillators and their power supplies,  screening, even things like DC cables, CPU selection etc all contribute or subtract to/from dynamic impact. 

On ‎4‎/‎9‎/‎2020 at 9:44 PM, FredM said:

 

As mentioned in the opening post, I was intrigued by the importance to match the source output format with the ‘native’ format of the DAC chip/module. An recent article from 6Moons goes in more detail, some snippets:

• “But when a DAC runs its own upsampler as the vast majority does but still allows you to defeat it, comparing it to the MU1's own filter can painfully reveal the resolution losses of generic on-chip upsamplers.”
• “ .. helps offload a DAC's heavy calculations when the latter converts music data from the digital into the analog domain.”
• “Throughout the listening period we switched between TotalDAC and T+A DAC 8 and each time found the TotalDAC in non-oversampling mode to be the preferred partner.”

 

Specifically regarding the Grimm MU1 server/streamer, for me it ticks all the boxes, and more. Perhaps in the Netherlands we’re spoiled, Grimm is well respected and the MU1 is reckoned as a state-of-the-art digital source (the Dutch site hifi.nl has even covered the MU1 in three in depth articles). Some other insights from the latest 6Moons review:

• "In Grimm's Eindhoven shop they spent a lot of time and effort perfecting a switching power supply which became a 2-storied affair. It is prepped for future expansions and was developed using jitter measurements on the audio clock.”
•  “The maximum sample rate for S/PDIF is 192kHz. USB supports 384kHz but eliminates a clock signal. S/PDIF's embedded clock signal meeting the MU1's superior clock is sonically far more relevant than sample rates above 192kHz!”
• “Compared to the Taiko SGM Extreme … the MU1 costs 2.5 times less, weighs 1/10th … and already stands out for its remarkably effortless listening quality.”

 

The Grimm MU1 in action, on top of a rather exclusive platform 😉

 

 

I’ve had the opportunity to listen at the MU1 on three occasions, leaving me (and others) nothing less than astonished with the impact of dynamics and accuracy and yet beautiful stress less sound (sets from €40k to well over €100k). Unfortunately it’s not my expertise to (even try) to describe a sound, I’ll just keep admiring those who can 😊.

 

For me the MU1 has all the advantages of an one box solution (synergy, in-house end to end design), without the disadvantage: upgrade flexibility. As I have no relevant electronic or software DIY skills myself, Grimm has a good track record supporting its customers with future upgrades for their products. This is very comforting in a digital area where new discoveries and improvements are a constant.

 

Oh dear.. I almost forgot, as @Nenon, @RickyV and @beautiful music have already requested, here's a link 😘

https://6moons.com/audioreview_articles/grimm/

 

Interesting that 6moons did not perform a listening comparison vs Taiko Extreme.......

Matt

6 hours ago, HeeBroG said:

Looking at the innards it seems to have a NUC board for the server/streamer side of things.

That means "low power" which can result in reduced dynamics.

Not sure if the special clock can recover this aspect.

Are you referring to upsampling or in general? 

1 hour ago, matthias said:

 

Interesting that 6moons did not perform a listening comparison vs Taiko Extreme.......

 

Matt

I agree if you read between the lines it appears that they didn’t include a direct comparison with the Extreme on purpose. Based on the price difference between the Grimm and the Extreme, a comparison of the MU1 with the Innuos Statement seems fairer.

42 minutes ago, McNulty said:

I agree if you read between the lines it appears that they didn’t include a direct comparison with the Extreme on purpose. Based on the price difference between the Grimm and the Extreme, a comparison of the MU1 with the Innuos Statement seems fairer.

.......or a comparison with the upcoming Taiko "Mini extreme" which costs about 50 to 60% of the Extreme.

Matt

It seems to me the MU1 is directed at audiophiles who have medium range/older non-USB DACs that might "strain" to upsample.

The price point of €9800 though would seem to to contradict this.

I think anyone who could afford that level of server would have the means to explore top-notch DACs

On 4/11/2020 at 10:55 PM, McNulty said:

I agree if you read between the lines it appears that they didn’t include a direct comparison with the Extreme on purpose. Based on the price difference between the Grimm and the Extreme, a comparison of the MU1 with the Innuos Statement seems fairer.

Reviewing is a commercial business, and so difficult to place direct comparison when one has revenue tied to it especially in these times. But I can say that for long listening sessions the Grimm will be more favourable.  

10 hours ago, HeeBroG said:

It seems to me the MU1 is directed at audiophiles who have medium range/older non-USB DACs that might "strain" to upsample.

The price point of €9800 though would seem to to contradict this.

I think anyone who could afford that level of server would have the means to explore top-notch DACs

 

The Grimm can output USB if so desired, and you can defeat the up or down sampling on it to provide the native signal.

I have a DAC with a USB input that upsamples everything to double DSD, but having trialled the Grimm 176.4 or 192 sampling via AES/EBU it will outshine the USB output due to it's FPGA card. 

Grimm were so entrenched in DSD earlier through their converters, that when they found out that 176.4 works better and required more stringent calculations to make it sound great, they pursued PCM.

33 minutes ago, nevillekapadia said:

 

The Grimm can output USB if so desired, and you can defeat the up or down sampling on it to provide the native signal.

Is this done by request to the factory, or can the user accomplish it at home?

15 minutes ago, ronfint said:

Is this done by request to the factory, or can the user accomplish it at home?

Accomplish at home. Just using one of the USB inputs, as it acts as an output too. Turn off the oversampling in the menu would let it deliver the song at its original recorded format.

3 hours ago, nevillekapadia said:

I can say that for long listening sessions the Grimm will be more favourable.

I am assuming you meant more favourable than the Taiko Extreme for long listening sessions? Can you please share your experience that led you to that conclusion? Have you had both in your system? Or maybe you've heard an A/B comparison somewhere? What DAC was used and what did the rest of the system look like? Thank you. 

7 hours ago, Nenon said:

 

I am assuming you meant more favourable than the Taiko Extreme for long listening sessions? Can you please share your experience that led you to that conclusion? Have you had both in your system? Or maybe you've heard an A/B comparison somewhere? What DAC was used and what did the rest of the system look like? Thank you. 

This was the feedback drawn by others when both units were in the same system, as I had to rely on reviews to make a choice which unit to pursue. 

Unfortunately I am left in a position not to be able to disclose this source, and it may seem vague or shrill to others in making such comments. But I leave that for users to audition and make up their minds.     

13 hours ago, nevillekapadia said:

 

The Grimm can output USB if so desired, and you can defeat the up or down sampling on it to provide the native signal.

I have a DAC with a USB input that upsamples everything to double DSD, but having trialled the Grimm 176.4 or 192 sampling via AES/EBU it will outshine the USB output due to it's FPGA card. 

 

Grimm were so entrenched in DSD earlier through their converters, that when they found out that 176.4 works better and required more stringent calculations to make it sound great, they pursued PCM.

Which DAC do you use?

In this use case, does your DAC take the AES/EBU input and upsample to DSD128?

G

On 4/10/2020 at 5:19 PM, Doh said:

.......however i am afraid i was severely underwhelmed, although there was a lot of attention on it, lots of people in room and a presentation going on.

 

I can only think the circuit was tweaked since then as the sound was cold, thin and uninvolving at that time, details sure but .......

Similar my impressions when visiting the Grimm room at HE Munich in the last years.

I never could endure listening to the LS1 speakers.

Matt

YMMV, personally the LS1 just stunned me, all good

On 4/12/2020 at 4:10 PM, nevillekapadia said:

I have a DAC with a USB input that upsamples everything to double DSD, but having trialled the Grimm 176.4 or 192 sampling via AES/EBU it will outshine the USB output due to it's FPGA card. 

Mixing and matching the server/streamer and DAC is growing on me. Good to see you echo the findings 6moons where they favoured the MU1 / TotalDAC D1 combination via AES/EBU without upsampling in the DAC:

“Throughout the listening period we switched between TotalDAC and T+A DAC 8 and each time found the TotalDAC in non-oversampling mode to be the preferred partner. “

With my limited technical knowledge on mixing and matching, as far as I now understand important point of attention are:

- connection type (USB, AES/EBU, I2S, ..)

- cable used for connection (connector, wire, shielding, ..)

- location and quality of the clock* and jitter level

- native ‘converter format’ of the DAC and output of the server/streamer

- upsampling and the quality of it

- ..

Very interesting to better understand these design subjects, but the more I read reviews or forums, the better I realise my technical insights are quite a bit under developed. I guess the list above is just a faction of the real world ‘matching criteria’. Bottom line that it’s all about synergy, and personal preferences of course.

* as an example, I even barely understand how this also relates to master & slave clocks and/or PLL & FLL. Way over my head I’m afraid 🙃 

@FredM or you could wait for the MU2.