• 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: 18 6.5%
  • 2. Not terrible (postman panther)

    Votes: 69 24.7%
  • 3. Fine (happy panther)

    Votes: 138 49.5%
  • 4. Great (golfing panther)

    Votes: 54 19.4%

  • Total voters
    279
Sigma-delta shenanigans,
Keith IMG_4647.png
 
ASR operates heavily on groupthink. Some things are considered anathema and R2R is one of them, even if you acknowledge the limitations as I have in this thread. I limited my inquiry strictly to 16/44 redbook CD, which good R2R implementations can very accurately decode, without any of the extra tomfoolery employed in DS implementations and look at the reaction it provoked.
Troll? Idiot? Either way, best not to engage with any further
 
without any of the extra tomfoolery employed in DS implementations
That "tomfoolery" allows DS dacs to achieve far better accuracy than R2R designs at lower cost (both/especially design cost and BOM cost)
 
ASR operates heavily on groupthink. Some things are considered anathema and R2R is one of them, even if you acknowledge the limitations as I have in this thread. I limited my inquiry strictly to 16/44 redbook CD, which good R2R implementations can very accurately decode, without any of the extra tomfoolery employed in DS implementations and look at the reaction it provoked.
Indeed.
Pushing a delta-sigma DAC to new levels of even more inaudible purity is a celebration of engineering and technical prowess.
Doing the same with R2R DACs is just audiophoolery.
Makes perfect sense to me ...
 
"Best not to engage with any further"
Neither, and this is the very definition of groupthink, right here. Look up the definition.

If a resistor network is capable of passively decoding a 16/44 signal without any oversampling or noise shaping, and the rest of the product is competently designed, as is the case with the Topping and Denafrips Ares II reviewed here, tell me why a DS implementation for 16/44 is superior.
 
Neither, and this is the very definition of groupthink, right here. Look up the definition.

If a resistor network is capable of passively decoding a 16/44 signal without any oversampling or noise shaping, and the rest of the product is competently designed, as is the case with the Topping and Denafrips Ares II reviewed here, tell me why a DS implementation for 16/44 is superior.
That's easy: it cannot. Turning off oversampling on these things at 44.1khz sounds close enough to the ear, but it is not really doing anything resembling accurate reconstruction of the sampled signal.
 
Neither, and this is the very definition of groupthink, right here. Look up the definition.

If a resistor network is capable of passively decoding a 16/44 signal without any oversampling or noise shaping, and the rest of the product is competently designed, as is the case with the Topping and Denafrips Ares II reviewed here, tell me why a DS implementation for 16/44 is superior.
Without oversampling of 44.1 kHz your back to the problem of very step filters very close to 20kHz probably a combination of digital and analog filters like in the late 80’s when CD arrived , this is not an improvement. Note correctly implemented reconstruction filters can not be omitted especially when i NOS mode .

Anyhow a bit nonsensical as DAC’s of somewhat reasonable design has been transparent to human hearing for decades.
 
ASR operates heavily on groupthink. Some things are considered anathema and R2R is one of them, even if you acknowledge the limitations as I have in this thread. I limited my inquiry strictly to 16/44 redbook CD, which good R2R implementations can very accurately decode, without any of the extra tomfoolery employed in DS implementations and look at the reaction it provoked.

I don't even own an R2R dac, I'm just trying to understand the potential strengths and getting this much pushback. Also, the ChatGPT hate is stunningly laughable. It easily passed the Harvard Medical boards, Harvard Law Bar exam and is capable of writing code surely better than 99% of ASR contributors, welcome to reality.

"R-2R DACs perform direct conversion from digital to analog by using a network of resistors to create discrete voltage steps corresponding to the digital input. This simplicity is a significant advantage in terms of signal purity, as there is no intermediate signal processing like in delta-sigma DACs.

  • R-2R DAC: The DAC directly outputs the analog signal without any digital modulation, oversampling, or noise-shaping.
  • Delta-Sigma DAC: Involves oversampling (usually 64x or higher), noise shaping, and the use of a modulator to convert the input signal to a high-frequency bitstream that is then filtered to produce the final analog signal.
References:

In the book "Data Conversion Handbook" by Walt Kester (Analog Devices), the author explains how oversampling in delta-sigma DACs leads to the reduction of quantization noise but can also introduce distortion and intermodulation products if not well-managed. In contrast, R-2R DACs avoid these issues by directly converting the binary input to an analog voltage.

The paper "Design and Optimization of DACs for High-Speed Data Conversion" published in the IEEE Transactions on Circuits and Systems discusses how the speed of a DAC architecture like R-2R is often superior to delta-sigma designs, particularly in systems where latency is a priority.

The paper "Performance Analysis of R-2R Ladder DACs" (IEEE Transactions on Circuits and Systems) highlights that R-2R DACs, with careful resistor matching, are able to provide high linearity and low distortion, making them ideal for systems where signal purity is prioritized.

The problem isn't the large language models like ChatGPT. The answers you got simply reveal the issues, or if you prefer the subsequent questions that need to be asked. For example:

- "Signal purity" has strong connotations that would seem to associate with maximum fidelity. But the fact that the design of an R2R circuit mimics the bit structure of PCM does not actually mean that the signal is any "purer" or "cleaner" than with Delta Sigma DACs. It's arguably conceptually and/or visually prettier, but that has nothing to do with what we hear. Don't forget that the entire procedure of digital sampling and reconstruction - the thing that this "purity" is supposed to preserve - is incredibly counter-intuitive and not at all "pure" seeming. Yet it works - and not just perceptually, but also provably and mathematically,

- "oversampling in delta-sigma DACs leads to the reduction of quantization noise but can also introduce distortion and intermodulation products if not well-managed." Is delta-sigma DAC oversampling well-managed? Is it difficult or risky or expensive to achieve that good management? Is this management issue a challenge or a solved problem at this time? Hint: it's the latter.

- "the speed of a DAC architecture like R-2R is often superior to delta-sigma designs, particularly in systems where latency is a priority." What is the latency of D-S DAC designs? Is it high enough to create issues for audio listening? (Hint: no.) What are the types of applications where "latency is a priority"?

- "R-2R DACs, with careful resistor matching, are able to provide high linearity and low distortion." No doubt. How does that linearity and distortion compare with that of inexpensive, commodity priced D-S DAC chips and fully assembled DACs? Does that linearity and distortion change over time as the resistors age?

If one asks these questions, one might find that the answers are ready at hand. No groupthink necessary.
 
Comprehensive rebuttal you've done here Oleg.
Of course you would think the spew of LLM nonsense (which scarcely makes it once sentence before a serious "citation needed" claim) you posted merits a comprehensive rebuttal, that's ChatGPT for you - people who don't know what they're talking about feeding garbage in, getting garbage out, and expecting other people to seriously engage with this ridiculous exercise.
 
Neither, and this is the very definition of groupthink, right here. Look up the definition.

If a resistor network is capable of passively decoding a 16/44 signal without any oversampling or noise shaping, and the rest of the product is competently designed, as is the case with the Topping and Denafrips Ares II reviewed here, tell me why a DS implementation for 16/44 is superior.
I own a few R2R DACs. They are expensive to make linear, and even more expensive to keep that way as the resistor ladders drift with time and temperature. Just because something is passive doesn't make it better, even if ChatGPT says so ;)

Here's ChatGPT answering a question about R2R DACs vs D-S DACs:

When to Use Each Architecture:
  1. Use an R-2R DAC When:
    • You need low latency and direct conversion (e.g., live audio applications, timing-critical systems).
    • The resolution requirements are relatively low to medium (e.g., 8-12 bits), and high fidelity is not absolutely critical at high bit-depths.
    • You are designing a simple, cost-effective DAC for lower-resolution audio (e.g., budget audio devices, microcontroller-based projects).
  2. Use a Delta-Sigma DAC When:
    • You need high-resolution audio (16-bit, 24-bit, 32-bit) with high dynamic range (e.g., audiophile systems, professional audio equipment).
    • You require excellent linearity and low quantization noise across a wide range of audio frequencies.
    • You need to achieve very high signal-to-noise ratio (SNR) and extremely low distortion.
    • The system can tolerate higher latency and possibly greater complexity in the design.

Conclusion:​

  • R-2R DACs are ideal for simpler applications with lower resolution, low-latency requirements, and cost considerations.
  • Delta-Sigma DACs are superior for high-performance, high-resolution audio systems where noise shaping, precision, and linearity are essential, and where oversampling and high-frequency filtering can be leveraged.
In the context of high-quality audio (especially audiophile applications), Delta-Sigma DACs are generally preferred due to their superior resolution, noise shaping, and overall performance. However, for simpler, low-to-medium resolution systems, the simplicity and directness of an R-2R DAC may be the better choice.
 
Thanks for the review!

@amirm, do you have any plans to review “Schiit Gungnir 2”? Since both of them were released recently, it would be great to see a comparison between two players.
 
Also important to mention that ChatGPT does a lot of "positive reinforcement". If the input is "tell me why x is better than y" then it will try to do so even if not really true.
This.

And if you then also ask why y is better than x, it will tell you that too. Guess what - both answers can't be correct.
 
Back
Top Bottom