Trying the Metrum will be next step. I’m currently selling my dac/preamp, as soon as I have sold that I will might look into the Metrum.
I was just wondering if the was any improvement when switching to roon on the Nad M50. But you never used the NAD with Roon?
Sure, remarkable difference in sound … no, not in my opinion, unless you start working with parametric EQ. The roon app is addictive in the positive sense, get used to it but once you realize it is a pleasure to go through your library + with Tidal it becomes really fun
In my case the step from M50 to Metrum Ambre is a logical step, he fully responds to my sound experience + the CD and rip function of the M50 I never used or maybe a few times when he was new.
My one on one comparuson of the M50 running with Bluesound and Roon learnt me that the sound quality with Bluesound was slightly better. Perhaps difference was due to adding a laptop with the roon server which was not using my optimised switch and cables.
I liked the parametric eq and the idea of doing room correction with the convolution filter. For now the M50 Bluesound combination won. But i will try to evaluate the ambre and roon combination if that outclasses the M50.
Just received one for test in front of a dCS network Bridge and an auralic Aries G2. I need more time, especially because the Ambre was new and need some time to settle especially for the clocks but first feeling is promising.
The Ambre should be better, but give it some time, ask a week or three to sound at its best.
Really looking forward to know how it compares against those giants!
Please report back soon
I have been using my Ambre for about 2 weeks now, connected to a Metrum DAC (the Onyx) via the I2S, and it has been rocking on several levels, exceeding my expectations in many ways, even though up until today, I have been listening to Ambre/Roon music only through my Hifiman He-1000 headphones, which is hooked up to my Cavalli Liquid Carbon v1.
However, as I mentioned in August, I have been always curious about how the Ambre would compare to the Ultrarendu as a Roon endpoint. Now, my Ultrarendu, which is powered by the Uptone Audio LPS 1.2, is itself hooked up to my Schiit Yggdrasil (Analog 2) via the Stello U3 USB ==> AES/EBU converter (I think this hookup sounds better than the Yggy’s USB-5 one, even if the difference is slight), and they serve my main 2-channel speaker system, which I typically enjoy quite a lot.
Today, however, I replaced the Ultrarendu with the Ambre in the Yggy A2 based chain which serves my main 2 channel speaker rig, using the Ambre’s AES connection, and the results are what I can only describe, for now, as “spectacularly ambiguous,” but in a good, “win win,” kind of way.
Not everything is as yet conclusive, but these are what my impressions are so far :
I suspect the Ambre’s I2s output, which serves my Onyx ==>>LCv1==>> He-K headphones rig, sounds a wee bit more spacious, airier, more resolving, and overall, noticeably better, even if it is by a hair, than the AES connection that is going to the Yggy, but I cannot completely verify this until I eventually hook the Onyx to speakers, which I am still yet to do. Note that I am not claiming that the Ambre’s AES output, now hooked up to the Yggy A2, is a slouch. Quite the contrary.
As compared to the Ultrarendu/I2S ==>>Stello U3==>> Yggy A2, the Ambre==>Yggy (via AES) is slamming and rocking as hard on all fronts, if not a bit harder in some key areas. This is a well-matched, pound for pound combat, which will take a lengthier, more careful, and extended comparative listening session to definitively call. So far it is more of the Ambre calling attention not only to its own sonic virtues, but also to those of the rendu, in the process if you know what I mean. It seems to me sometimes that the Ambre sounds a wee bit more detailed, slightly more resolving, and blacker in the background than the rendu, which is, in its own right, no slouch at all either. Conversely, it seems sometimes that the Ultrarendu might have a tad more bass-heft and extension than the Ambre, which seems more gifted in texturizing the bass response, but in either case, the results are too close to call, and I think more extensive comparative listening is called for before the differences can be properly and more satisfyingly sorted out or nailed down,
What is strikingly interesting about this experimenting, so far, is that I am not coming out feeling any type of disappointment with either the ultrarendu or ambre, sonically speaking, although overall, I have to admit that I am more impressed with Ambre, given its higher relative versatility, as compared to the Ultrarendu.
Still, the one unambiguous area in which the Ambre completely leaves the Ultrarendu/LPS-1.2 combo in the dust, lies in its relative versatility, mostly, in the superior number of connectivity options it offers. Whereas the Ultrarendu offers only one USB output, the Ambre offers a total of 4 that are all simultaneously active. In that respect, the Ambre is currently serving my main 2-channel rig via the AES connection, but I can also listen to it with my headphones on the Onyx/Cavalli rig via the I2S connection, so you could say, it is serving 2 rigs simultaneously at this very moment. It still has 2 ports left (an S/PDIF coaxial output, and a Toslink Optical output) which can be hooked to two other extra rigs if necessary…
From this latter viewpoint alone, I feel very confident in declaring the Ambre the winner over the Ultrarendu in two important areas, namely, in its versatility of connectivity options, and also as a value for money proposition, and those are nothing at all to sneeze at, in my view.
I still love my Ultrarendu as much as I did on the first day I got it, and since I need more than one reference calibre rig, I feel certain that it is going to stay… However, the Ambre must be given its due : as a roon endpoint, it is currently the undisputed, best-performing, king of the “castle,” that I also call my humble abode 
+1. I’ve been using an Ambre with a Holo Spring (via AES, since I’m still waiting for the correct RJ-45>HDMI I2S wiring from Metrum), and with an Onyx that Metrum lent me for evaluation. I agree with you that the Ambre is an excellent streamer. Even using AES into the Spring rather than I2S, it’s a win over my earlier USBridge>SU-1 I2S source. I have downstairs an Yggy A2 fed by an Auralic Aries Femto. Now you are tempting me to go try the Ambre there…
I would say try it, if doing so does not seem too complicated, or involve to much effort. This looks like one of those “nothing to lose” situations. even if not much should come out of it. On the other hand, you might well discover a pleasant surprise, and that is always a nice thing to have
A pleasant surprise for Metrum, if I end buying a 2nd Ambre for downstairs 
Hello, I tested and finally bought a ambre few weeks ago. I tested it on my main system in front of a dCS Network bridge. The dCS is definitely doing better in my system but gap in sound quality does not worth the price in money. Tests have been done in AES. dCS is going more far in details and somehow on stages but it is not a world.
Finally I kept the ambre and bought an Metrum amethyst modified with a i2s input. I use now the amethyst and ambre with my Sennheiser hd800s headphone and I can testify that the i2s is really making the difference in terms of sound stage, resolution and fluidity (silence and sound stage differentiation) with a spdif connection (few cables tested). The SPdIF is the only other connection available between the ambre and the amethyst.
I linked both devices with a sotm cat6 hg Ethernet cable (20cm) or a patchanko one (30 cm). No big differences between cables.
The drawback is that it could have been a good idea to test the ambre in i2s with my major system but I don’t have i2s DAC and even if metrum is providing the pin compatibility you need a bit of diy to link ethernet to hdmi for example or another ethernet pins.
A friend of mine also bought it and use it with squeezelite (moode 4.3 or pcp). Except the blue led blinking everything is nice and he can enjoy qobuz streaming through LMS or Roon in squeezebox mode.
Sorry for my question. Not sure what the ambre does and how would you fit it with the amethyst device between a windows computer and an AVR ? Can someone provide some details about how the connections between the 4 devices will run windows-ambre-amethist-avr ? also would the sound become much improved? I now stream from computer through HDMI to a Maranz AVR, but I am not convinced I like the sound, bass sound very muffled and I think I might gain some sound quality by using an external DAC; I looked into the Aurelic Vega but then I stumbled upon the Metrum products. Aurelic vega had rave reviews but just released a Vega 1 and 2 but those are beyond of what I could afford and willing to pay. Can someone recommend one solution over the other ( vega vs Metrum ambre/amethyst) ?
Ambre is a networked Room endpoint. It receives the audio stream from the Room core and outputs it to a DAC. Room endpoints are useful if you have multiple audio systems with a single core (like I do) or to avoid connecting your DAC directly to the Core server, which may be inconvenient or suboptimal from a sound quality point of view (servers can be electrically noisy, and USB audio is not the best connection method for some DACs).
I use the Ambre with Metrum Onyx(DAC). I initially had Bluesound node 2i going to Rotel 1572 and while it did sound good, the Metrum sound (IMO) is in a different league all together.
I’ve been told many times that the Roon endpoint you use will be very important to the overall sound of your system. Obviously the DAC plays a role as well, but I think many often overlook the endpoint.
just responding to your October impressions of the Ambre. How do you feel about it 2 months later? I have the same setup.
Thanks Fernando
Thanks Jimmy, I think I will get my feet wet and purchase the Ambre /Amethyst combo at 2100 $ ( can’t afford more then 2.1 k now) instead of the Auralic Vega at 1900 $. I saw a Srajan Ebaen review on 6mooons
where he compares the Amethyst to the Vega and he states the following : ‘’ But reviewers dare not leave it at that. They’re expected to parse and analyze. Enter the Vega, here tapped RCA not XLR to simplify A/Bs. Colour saturation went up, rhythmic perspicacity down. In a sense, it mirrored going from DSD to PCM. To my ears, the former is softer, fuzzier, less informative in the treble but often spatially more texturized. Amethyst took up the counter position of PCM. This meant more articulated transients, more focus, better top-end visibility and depth specificity but all of it at a lower colour intensity and with leaner textures. Like DSD vs PCM, both were equally valid. And like DSD vs PCM, my personal preference, in this system, was with the latter. ‘’ This gives me confidence in purchasing the Metrum ambre /amethyst bundle.
What bullshit. Some people have great imaginations.
And all, probably, from an SBC with the same ARM processor and architecture as an RPi. Wonder what interconnects they used.
Hi Slim, so what are you saying is that one should not bother thinking about purchasing those Dac’s. I am new to this and therefore rely on reviews and such. Thanks
Amethyst and ambre « review » : https://metrumacoustics.com/blog/2018/10/28/amethyst-ambre-a-wonderful-set/
Bullshit or not bullshit … that is the question