• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Topping Centaurus R2R DAC Review

Rate this R2R DAC:

  • 1. Poor (headless panther)

    Votes: 23 7.7%
  • 2. Not terrible (postman panther)

    Votes: 70 23.5%
  • 3. Fine (happy panther)

    Votes: 147 49.3%
  • 4. Great (golfing panther)

    Votes: 58 19.5%

  • Total voters
    298
He's going to be reviewing the Centaurus - gives an initial impression on the Centaurus a few videos before this one.
It's not about the technical performance. Audiophilia is the practice of attributing certain characteristics or properties of audio products (real or imaginary) to specific (often esoteric) types of equipment, components, implementations or materials not based on or contrary to any objective evidence.

The Centaurus performs not bad objectively. The problem is that its R2R topology is perceived and marketed as the main merit of this product and an advantage in its own right.
 
the practice of attributing certain characteristics or properties of audio products (real or imaginary) to specific (often esoteric) types of equipment, components, implementations or materials not based on or contrary to any objective evidence.
We are really fortunate to be able to buy SOTA gear for so little nowadays, and indeed really show up 'esoteric' equipment ... especially amps and dacs.
 
Mnyb said:
This is one of the best examples of R2R currently. They should just remove the the NOS 2 filter setting altogether for public safety, just like mandatory seatbelts and other regulations :) pure dunning Kruger territory to be fa filter tinkerer imho .
listeners enjoy it , why not?
I suspect that the NOS 2 filter setting is one of the key elements to what people want from NOS DACs.

If we accept that we can measure anything that we hear, there seem to be two aspects of NOS DACs that stand out - a spray of distortion harmonics, and an absence of digital and analogue filtering (which are often muddled up). The latter is usually switchable, though some NOS DACs (like the Holo Cyan) don't even give you the choice. The lack of filtering trades time domain performance for frequency domain performance. Impulses don't have (pre- and) post ringing, but both ultrasonic noise and distortion and images of wanted musical signals are allowed through. I suspect this is what NOS DACs supporters are actually after.

If the NOS filter option was removed I think it might defeat the point of having a NOS DAC at all.
 
I suspect that the NOS 2 filter setting is one of the key elements to what people want from NAS DACs.

If we accept that we can measure anything that we hear, there seem to be two aspects of NAS DACs that stand out - a spray of distortion harmonics, and an absence of digital and analogue filtering (which are often muddled up). The latter is usually switchable, though some NOS DACs (like the Holo Cyan) don't even give you the choice. The lack of filtering trades time domain performance for frequency domain performance. Impulses don't have (pre- and) post ringing, but both ultrasonic noise and distortion and images of wanted musical signals are allowed through. I suspect this is what NAS DACs supporters are actually after.

If the NOS filter option was removed I think it might defeat the point of having a NOS DAC at all.
There is a lot in this.

From what I can gather many want to send 'pre-upsampled' data to the NOS set input on the DAC.
I don't understand a manufacturer not providing an analogue filter option above 20khz. I suppose there might be subharmonics people might hear ... but that is a big 'suppose' on my part.

The big 'if' in your statement I think is in the measurement of timing, or that is what I am seeing / reading in my general foray into this topic. Not sure why oversampling would cause 'pre/post ringing' and upsampling would not!

Post edit: those 'impulses' you refer to are what are being proposed as audible, and they seem to be measurable, but not a part of standard measurements used (there is a question in there I think)! ?
 
If the NOS filter option was removed I think it might defeat the point of having a NOS DAC at all
Removing reconstruction filtering removes the point of having any digital system at all.
 
The big 'if' in your statement I think is in the measurement of timing, or that is what I am seeing / reading in my general foray into this topic.
Not sure why oversampling would cause 'pre/post ringing' and upsampling would not!
I never thought of that, but I see your point. Oversamping is a time domain function, and upsampling is amplitude domain. However (thinking out loud) if you're doing both to 16 bit data, I guess the interpolated samples will want to sit on 24 bit data points.
Post edit: those 'impulses' you refer to are what are being proposed as audible, and they seem to be measurable, but not a part of standard measurements used (there is a question in there I think)! ?
I think the audibility of pre and post ringing is a key factor. Meridian published the result some research at AES convention 137 that showed that different filtering schemes were in fact audible; I think they were trying to prove the benefits of apodising filters. Goldensound also performed a successful ABX DBT on different DAC filters and got a similar result. The ringing is generally around 22kHz, which should be inaudible, but I think these results show otherwise.

Therefore I think there's some mileage in what NAS fans want, though it's got to be understood that it's trading one undesirable artefact for another undesirable artefact, and you have to take the lesser of the two evils. I don't know what the answer to that it, but it may be different depending on whether the audio is CD or HR format, that doesn't get discussed much.
 
I never thought of that, but I see your point. Oversamping is a time domain function, and upsampling is amplitude domain. However (thinking out loud) if you're doing both to 16 bit data, I guess the interpolated samples will want to sit on 24 bit data points.

I think the audibility of pre and post ringing is a key factor. Meridian published the result some research at AES convention 137 that showed that different filtering schemes were in fact audible; I think they were trying to prove the benefits of apodising filters. Goldensound also performed a successful ABX DBT on different DAC filters and got a similar result. The ringing is generally around 22kHz, which should be inaudible, but I think these results show otherwise.

Therefore I think there's some mileage in what NAS fans want, though it's got to be understood that it's trading one undesirable artefact for another undesirable artefact, and you have to take the lesser of the two evils. I don't know what the answer to that it, but it may be different depending on whether the audio is CD or HR format, that doesn't get discussed much.
Yes 22kHz certainly well above my thresholds ... maybe my 7 year old ... or if there is a knock on effect at lower frequencies.
Interesting point of debate the timing domain, and amplitude domain, indeed.
 
I never thought of that, but I see your point. Oversamping is a time domain function, and upsampling is amplitude domain. However (thinking out loud) if you're doing both to 16 bit data, I guess the interpolated samples will want to sit on 24 bit data points.

I think the audibility of pre and post ringing is a key factor. Meridian published the result some research at AES convention 137 that showed that different filtering schemes were in fact audible; I think they were trying to prove the benefits of apodising filters. Goldensound also performed a successful ABX DBT on different DAC filters and got a similar result. The ringing is generally around 22kHz, which should be inaudible, but I think these results show otherwise.

Therefore I think there's some mileage in what NAS fans want, though it's got to be understood that it's trading one undesirable artefact for another undesirable artefact, and you have to take the lesser of the two evils. I don't know what the answer to that it, but it may be different depending on whether the audio is CD or HR format, that doesn't get discussed much.
ABX testing of different, non-ideal filters has been known about for some time. In the earliest days, before mathematical, IIR, FIR filters, the choices between ringing and insufficient attenuation were potentially audible. Properly implemented modern filters have ringing artefacts outside of human hearing with still excellent attenuation.

Having no filter is broken.
 
Agreed. But what if no filter was preferred in a controlled test?
What if there is a teapot orbiting the Sun? Even if it was, it does not mean this is the correct way.
 
so what, people like what they like . And sometimes it's koolaid and that's fine too.

Absolutely. As long as we don't have to listen to any claims about NOS filters giving higher objective performance.

People can play their music through a fuzz pedal for all I care, but please call a spade a spade, and an effects box an effects box.
 
Agreed. But what if no filter was preferred in a controlled test?
That's a preference for corrupted sound. Consumers should be made aware.
 
Effects box?? I like that term!

I like both my Soekris R2R effects box, and my superb SINAD Topping DM7.
 
Do they sound different?

< Teflon Suit On >

I'll walk right into the trap question... YES ... they do sound DIFFERENT ... just so very SUBTLY different.

I'll use "audiophile" terms to describe the SUBTLE differences:
  1. Soekris dac1541 R2R (4 strings, discrete class-A buffer)
  2. Topping DM7

Guess what?? I enjoy both DACs !!!

Let the bashing begin !!!
 
Last edited:
< Teflon Suit On >

I'll walk right into the trap question... YES ... they do sound DIFFERENT ... just so very SUBTLY different.

I'll use "audiophile" terms to describe the SUBTLE differences:
  1. Soekris dac1541 R2R (4 strings, discrete class-A buffer)
  2. Topping DM7

Guess what?? I enjoy both DACs !!!

Let the bashing begin !!!
You're not going to get bashed. You likely are going to get asked whether you've done a proper (ie level matched etc) double-blind test. Because if not, it's possible you say one DAC is "warmer" (whatever that means) simply because you know it's "that dac", but that the differences are not detectable.
 
< Teflon Suit On >

I'll walk right into the trap question... YES ... they do sound DIFFERENT ... just so very SUBTLY different.

I'll use "audiophile" terms to describe the SUBTLE differences:
  1. Soekris dac1541 R2R (4 strings, discrete class-A buffer)
  2. Topping DM7

Guess what?? I enjoy both DACs !!!

Let the bashing begin !!!
:D :D It's all bias ... it's just because you saw those charts that you think the Soekris sounds different.
I actually would expect you to hear a difference between those, so I'll get my teflon suit on too.
 
< Teflon Suit On >

I'll walk right into the trap question... YES ... they do sound DIFFERENT ... just so very SUBTLY different.

I'll use "audiophile" terms to describe the SUBTLE differences:
  1. Soekris dac1541 R2R (4 strings, discrete class-A buffer)
  2. Topping DM7

Guess what?? I enjoy both DACs !!!

Let the bashing begin !!!
Next question is how does the R2R one measure ;)
 
Back
Top Bottom