I have stumbled on an interesting interview of Kusunoki San, who is accredited for bringing the NOS dac into the light. I find this a very interesting read. He is, in fact, an engine designer, which I think is fantastic as he is more able to draw upon design philosophies from different engineering fields and possibly think "outside the square". What makes his theory any less valid than any other?
I stumbled upon this as I have been on the crusade for answers upon purchasing the Muse Mini DAC TDA1543 X 4. After further listening, it starts to pull-away from my highly praised, hi-end Bryston DAC, further and further - with accordance to my preferences...which is to replicate my experience as a drummer. Just a drummer, mind you, as I have never considered myself a musician - nor have I ever considered myself an audiophile for that matter. This experience has let me enjoy music at a very intimate level and this has been the benchmark for my pursuit, saving money is of very little priority.
Here is also an interesting article by TNT audio, I find the information from section 2.4 onwards very interesting. Also another interesting article by Mother of Tone, also exploring the same issue.
What is of interest to me, as it coincides with my own conclusion, is that Kusunoki San himself, dislikes the "NOS" tag. He argues that oversampling is not the culprit that many blame on supposed limitations of modern DAC designs, rather that it was the digital filters that were the culprit. His evidence was presented in the form of paralleled TDA1543 - 4 in parallel was in principle 4 X OS, 8 chips = 8 X OS etc.
When I listen carefully to the Bryston, there are aspects to the audio signal which come off better than the Muse. I find that the treble is more refined, the bass is a little more accurate - this is most noticeable on less than perfect recordings, on excellent recordings, the gap is minimal. I can only attribute this to the superiority of taking hundreds of samples and averaging them out - bringing refinement to less than perfect digital signals. Whereas the Muse - simple relays whats on the recording with its 4 X averaging, being less efficient. Overall, the Bryston is perceived as a more refined DAC.
The advantages that the Bryston has, does not change my preference for the Muse Mini Dac. Moreover, the more tightly controlled bass on the Bryston might be a result of its superior power supply and regulation, with dual toroidal transformers, each separately regulated for the digital and analogue stages. Whereas the Muse Mini Dac is merely run from a 12v DC SMPS wallwart.
What the Muse does better - serves as a coupe de gras for the Bryston. The sound is "natural" with each instrument clearly delineated from the next. Each instrument occupies its own "space" - whereas on the Bryston, I find everything "congested", "enclosed" and instruments are less separated. Also, on the Muse, the instrument themselves have a convincing tone to them whereas on the Bryston, everything is somewhat homogenised in tonality and gelled together in an artificial enclosure. An enclosure that is less convincing with regards to a reality based reference point.
The Muse comes off with a bigger more spacious, three dimensional presentation than the Bryston. It seems to extend further up into the treble and further down the bass - bringing with it a certain realistic palpability in vocals and timbre that the Bryston cannot match. Upon hearing the Muse, it became apparent that the Bryston retains an "artificial sheen" that, in my experience, serves as a source of fatigue. This "artificial sheen" is worst on my inexpensive Dacmagic. I felt that the Bryston eradicated this artificiality when compared to the Dacmagic, but alas, upon hearing the Muse - remnants still remain.
This leads me to the issue that distresses me the most with regards to the Dacmagic and the Bryston, the perception of an "artificial sheen" that upon further investigation is a source of listening fatigue, for which the Muse either does not have or it is miniscule levels - for which I cannot, as yet, perceive. These two modern DACs share a common DA conversion technique - "Delta Sigma". The Mini Muse utilises an older - more expensive to implement "R2R" conversion technique.
"Delta Sigma" is the equivalent of a "SMPS" PSU, or a "Class T" digital amplifier. It relies on a single binary unit, on or off. Voltage present or no voltage present. On time (voltage present) and off time (voltage not present) - a varied voltage is obtained by changing the on time vs off time with a very high frame rate. This is the basis for modern vehicle CAN BUS communication - a process I am all too familiar with (for instance altering the pulse duration of an injector to accommodate for engine load or varying the speed of an electric water pump to suit engine cooling requirements). The fundamental principles are achieved via Pulse Width Modulation (PWM) or in other words "fast switching" or "switch mode". Also of note, I curse the introduction of digital sensors which were more prone to failure than traditional analogue, sine wave sensors.
This high frequency switching, I believe relates to the fatigue that many complain of in modern digital music reproduction, or "digititis" as is commonly labeled. I stumbled upon this phenomena on a grand scale when experimenting with car batteries to power up my audio systems. The use of inexpensive fast switching switch mode power inverters, when experimenting for a clean, viable power, led to what I perceived as a massive amount of high frequency hash or energy that was extraordinarily fatiguing. I would have abandoned this pursuit had I not stumbled upon a much more expensive industrial grade low frequency power inverter - using traditional heavyweight toroidal transformers. This solved the problem of listening fatigue.
What I hear from the Bryston is a similar effect - albeit in a drastically reduced level. I theorise that this is the side effects of Delta Sigma conversion techniques and most high frequency switching. However this technique can be used to increase the resolution of an analogue output by increasing the frame rate, or sampling frequency - as per SACD. However the benefits are questionable, if high frequency contamination is a side effect.
To conclude: of all the DA processors I have at my disposal (Yamaha Receiver, Pioneer SACD player, Ipod, Dacmagic, Bryston), the Muse Mini DAC TDA1543 X 4 has eclipsed all - in a most unique way, in reproducing an accurate, precise, neutral and revealing sound that is more reminiscent of my reality based reference.
What was initially meant as a shootout with the Dacmagic turned into a shootout with the 50 times more expensive Bryston DAC - emerging victorious. Yet the Bryston is universally hailed by professional reviewers as a giant killer - and to me sounds brilliant and streets ahead of the likes of my Dacmagic with very little "digititis". However, I feel it is a weakness that befalls upon all of its Delta Sigma relatives. A weakness the Mini Dac, for now, appears to be free from.