ASR Review of Naim Uniti Atom

[windhoek]

Arguably, something that matters much more than how a given product measures and what reviewers have to say about it is how it sounds once you get it plumbed into your system in your listening room. It might measure really well (or really poor), and other people may say it sounds really good (or really bad) in their system in their room, but until you hear it as part of your system in your room all bets are off. Measurements and reviews may give you an indication as to how a product might contribute to (or detract from) the present sound of your system (as if anyone else could know that), but the influencing power of your actual system in your actual room trumps everything anyone has to say about this or that product and how it measures. That's not to say such things can't be a guide as to what you can expect, but really, what it delivers in the end wlll be determined by your system and room.

 

[Del monaco]

I agree. Same as live music. Even if the acoustic is just ok I can still enjoy the music just as much. Everyone enjoys a different kind of sound and good measurements don’t necessarily make a product I want to listen to. As long as they are within a reasonable tolerance, does it really matter?

 

[Purité Audio]

Anyone who thinks that site doesn't have an agenda should look at the recent hatchet-job on the Chord M Scaler. Amir neglected to test its best-performing 16x upsampling rate despite having the capability to do so (he was supplied a Chord Hugo 2 DAC to use with it but instead chose to conduct most of his measurements with it connected to a Topping DAC which could only accept 2x and 4x upsampling!). He also criticised its use of Gaussian dither over Triangular dither in 2x and 4x upsampling modes, pointing to a research paper that shows the former produces 1.25dB more noise (I can't recall the exact figure and I sure as hell ain't watching the 30+ minute video again but I'm pretty sure the figure he quoted was around 1.25dB, i.e. negligible!). He attributed this design decision to cost-cutting, when in reality that the designer chose Gaussian over Triangular because he preferred it in listening tests. If you subscribe to the ASR way of thinking then you are in effect stipulating that designers should not be allowed to use their ears in the tuning process.

Actually he did test it with the Hugo at 16x oversampling, it’s clearly there if you actually read the review.
Keith

 

[flatpopely]

Do ASR ever comment on how enjoyable a piece of equipment is to listen to?

 

[John Phillips]

Anyone who thinks that site doesn't have an agenda should look at the recent hatchet-job on the Chord M Scaler. ...

I have been reading that thread. It isn't pleasant. IMHO, hasty, not well considered interpretation of what was measured seems to be more of a problem than the tests themselves.

I have seen some voices offering considered views. However, many more seem to not have looked carefully at what can reasonably be taken from the tests and what cannot. There are reasons I personally would not buy the product but what has been discussed there has generated more irrelevant heat than relevant light. It's a pity. I like a resource like that to add to my understanding of a product.

 

[PerF]

Do ASR ever comment on how enjoyable a piece of equipment is to listen to?

What I thought about too.

Their criticism of Naim Atom is their own statement or opinion, I honestly don't understand me on the measurements they deliver, having no real interest in these.

I searched for their listening review but found none.

Guess that meant they never listened serious to Atom, which actually sound quite good IMO, price aside.
No matter how it measure.

 

[Tony L]

Guess that meant they never listened serious to Atom, which actually sound quite good IMO, price aside.

As I understand it ASR is a site where someone who bought some expensive test gear measures stuff readers send him and places it on a chart between what he considers objectively bad at one end, and objectively good at the other. Apparently he doesn’t even assess a pair of speakers, just one in isolation. This is just a game of Top Trumps to my mind. It has little to do with audio, science, or reviewing.

It is absolutely not a competitor to say Stereophile which balances high quality objective and subjective reviews to give a rounded impression of what they liked or didn’t like about a product, and indicate why that may be the case. Stereophile also give the manufacturer/designer some comeback and ability to challenge these findings, which is essential IMHO if you are attempting to state something as hard fact/as an absolute.

 

[jtgofish]

ASR stands for- Aspergers for Soulless Reproduction.

 

[Ryder35]

I find the whole premise of the site the exact oposite of why I buy Hi-Fi, to listen to music. Who cares how it measures!

 

[ToTo Man]

Actually he did test it with the Hugo at 16x oversampling, it’s clearly there if you actually read the review.
Keith

Where? All I see in his written 'review' is a 16x frequency response measurement, not the full suite of tests that were performed in bypass, 2x and 4x modes.

 

[sideshowbob]

Measurements matter, a lot, but I agree with Tony that measurement needs to happen in a context. There may be perfectly good reasons for buying product A, even if it measures worse than product B, particularly when talking about digital audio, a domain in which many of the measurements are of things average humans are vanishingly unlikely to be able to hear anyway. I sometimes use a little 2wpc valve amp I built from a kit. I know it measures not particularly well even within its limits, but it still sounds really nice, and I built it, so I like it anyway. I also use solid state power amps with much more power which measure much better, and I like those for different reasons. I like some NOS and vintage dacs, which don't measure well against modern competition, but still sound really good and are often much better built than what's currently available. I also have a small modern inexpensive dac (a Soncoz), and I like that very much too. There's no single answer.

Having said that, I'm pretty sure the review of the Chord M-Scaler is accurate. I can't see anything in the description of what it does that makes me believe the audible difference it can make will be anything other than tiny, probably tinier than moving my speakers a few centimetres. I am not capable of reliably hearing such incredibly fine upsampling distinctions, and even less capable of finding them musically important.

 

[Tony L]

I find measurements interesting as a tool to help understand what makes the really good kit stand the test of time and remain more desirable than whatever the current magazine or shop rave is. My concern is I just don’t think the standard measurement criteria captures what is important. As an example a really large high efficiency horn system like the Altec VOTT captures the lightness of touch and dynamic headroom of a great jazz drummer, the presence of Coltrane’s sax or whatever in the way no little box with a ruler flat response could ever dream of doing, and it does it even if you match the volume level to that which the smaller speaker can do without distortion. You can’t re-EQ a Kef LS50 to sound like a VOTT or KHorn!

There is IMHO way more to this than a ruler flat response, controlled dispersion and low distortion IMO. The things that make the truly great speakers truly great evade simplistic measurement. Same argument for ESLs, LS3/5As, SBLs, blue-face JBLs, Tannoys, Lowthers, Isobariks, Shahinians etc etc, in fact pretty much every classic audio product. What they do so, so right tends to evade current measurement thinking, what they do wrong doesn’t tend to matter to those who go to great lengths to seek them out.

 

[Tony L]

Having said that, I'm pretty sure the review of the Chord M-Scaler is accurate.

I had a very quick skim-read of it earlier and from what I could tell he didn’t even try it in the Chord system context it is designed to operate. No one considering one would be hanging it off the end of some random streamer or whatever he used. One doesn’t review a Koetsu in a Dual 504.

That said I’m not defending it, it seems very low perceived value and if I’d dropped the price of a Hugo TT, Dave or whatever it is intended to partner I’d be more than a bit pissed what I’d already spent so much on wasn’t state of the art as-is! I’d still listen to it with an open mind though.

 

[ToTo Man]

I had a very quick skim-read of it earlier and from what I could tell he didn’t even try it in the Chord system context it is designed to operate. No one considering one would be hanging it off the end of some random streamer or whatever he used. One doesn’t review a Koetsu in a Dual 504.

That said I’m not defending it, it seems very low perceived value and if I’d dropped the price of a Hugo TT, Dave or whatever it is intended to partner I’d be more than a bit pissed what I’d already spent so much on wasn’t state of the art as-is! I’d still listen to it with an open mind though.

Having had the opportunity of a month-long audition of M Scaler with Hugo TT 2 using several flagship headphones, I was definitely connected to the music more when listening through M Scaler's 16FS upsampling filter compared to through TT 2 on its own. The difference between M Scaler's 'pass-through' and lower upsampling rates was subtler, but like you I struggle to believe anyone spending £3.5k on a 16x upsampler intends to bottleneck it by running it at 2x or 4x into some random, budget DAC that it hasn't been optimised for!

 

[al2002]

Do ASR ever comment on how enjoyable a piece of equipment is to listen to?

Read the test report on the Neumann 420. There are others too: Genelec, Revel,KEF.

Edit: Of course, there is the question of whether it is the equipment or the music that is the source of one’s enjoyment. For me it is always the music. As an example, I’m listening to a low Fi 1938 record by Billie Holiday on my iPad’s tinny speakers and enjoying it immensely; far more than I would a synthetic state-of-the art recording of the latest pop megagroup played on my main system. Others might disagree and say both matter or that the equipment plays a significant role,which is fine with me. Perhaps a topic that deserves its own thread.

 

[camverton]

I had a very quick skim-read of it earlier and from what I could tell he didn’t even try it in the Chord system context it is designed to operate. No one considering one would be hanging it off the end of some random streamer or whatever he used. One doesn’t review a Koetsu in a Dual 504.

That said I’m not defending it, it seems very low perceived value and if I’d dropped the price of a Hugo TT, Dave or whatever it is intended to partner I’d be more than a bit pissed what I’d already spent so much on wasn’t state of the art as-is! I’d still listen to it with an open mind though.

As someone with an m-scaler and TT2 I certainly agree with perceived value for what is a subtle improvement in sound quality. However, that particular subtle improvement in sound quality has a very large effect to me personally on the enjoyment of my music. Others, sensitive to other aspects of sound quality might not even notice a difference, so value to them would be zero! Ultimately it is worth it to me but I still think it a high price to pay, although to be fair there has been years of research and one has to be careful of valuing everything by the cost of parts. On that basis a cup of tea in a cafe would only cost 2p!

The test at ASR does seem to have been undertaken with a predetermined view of the outcome in a sub optimal way. To hear an m-scaler in an optimal way you would need to use a Chord DAC with dual bnc connectors, and dare I say it a pair of Wave cables. They don’t and they make no proper effort to correlate measurements with listening. Still, crusaders will crusade and they don’t seem much bothered with the veracity of their crusade. One of the followers even called their leader a “veritable beacon” .

 

[Tony L]

As someone with an m-scaler and TT2 I certainly agree with perceived value for what is a subtle improvement in sound quality.

By ‘perceived value’ I was more commenting on the simplistic ‘what’s in the box’ attack on the ASR thread. It looks like a handful of typical surface mount components and standard connectors. To a layman not hugely different to say a Raspberry Pi or whatever. This observation obviously has no connection to Rob Watt’s work in designing the FPGA and the lifetime of digital expertise he brings to the project.

It isn’t an argument I’d use myself as my attitude to audio is just poles apart from this mindset and has far more to do with subjective performance, history, serviceability, sustainability, investment value, how it is constructed, who constructed it etc.

 

[Amber Audio]

Anyone who thinks that site doesn't have an agenda should look at the recent hatchet-job on the Chord M Scaler. Amir neglected to test its best-performing 16x upsampling rate despite having the capability to do so (he was supplied a Chord Hugo 2 DAC to use with it but instead chose to conduct most of his measurements with it connected to a Topping DAC which could only accept 2x and 4x upsampling!). He also criticised its use of Gaussian dither over Triangular dither in 2x and 4x upsampling modes, pointing to a research paper that shows the former produces 1.25dB more noise (I can't recall the exact figure and I sure as hell ain't watching the 30+ minute video again but I'm pretty sure the figure he quoted was around 1.25dB, i.e. negligible!). He attributed this design decision to cost-cutting, when in reality that the designer chose Gaussian over Triangular because he preferred it in listening tests. If you subscribe to the ASR way of thinking then you are in effect stipulating that designers should not be allowed to use their ears in the tuning process.

Do you mean he received a TT2 and Scaler or a Hugo 2 and Scaler, I assume they mean the TT2? To make the most of the Scaler you need a dual BNC DAC - TT2 or DAVE.

The Scaler has a defined role - it is aimed at TT2 and DAVE users, testing outwith that is fine but I am willing to bet that there are very few users of the Scaler with non Chord DACs, reckon a few will get the Scaler with a view to moving on from Hugo/Qutest to TT2, that’s the route I went. Testing it with a Topping seems strange given how much the Scaler costs. FWIW I owned the D90 at the same time as a Scaler/TT2 and a few other DACs like the Matrix Sabre MQA that ASR cheerlead on their SINAD scores, a year or so ago.

 

[Somafunk]

Apparently he doesn’t even assess a pair of speakers, just one in isolation.

When listening to a single speaker it is easier to check for correct tonality as a stereo image can obfuscate errors.

 

[Amber Audio]

When listening to a single speaker it is easier to check for correct tonality as a stereo image can obfuscate errors.

For Home Hi Fi don’t designers take into account the fact they will be sold/used as a pair in a typical living room and aim at that as a target environment. I can see the sense for a studio/musician application of gear sold in Pro Audio Shops in the test but not for gear typically sold in 7oaks/Richer etc.