dCS Ring DAC - A Technical Explanation
Apr 26, 2021 at 3:26 AM Post #31 of 187
Poor James having his thread hijacked so massively....

I am all for the scientists ensuring that their religion is artfully acknowledged (that plots represent information they are portraying etc), but I feel some leeway need be given here..
Why intelligent people need to ‘get in the way of’ simplified information for the masses (clearly the point for this thread) I am not sure- texts books would hint at ego needing acknowledgement.
”there there- we are all *special*“ (some clearly more than others)

When one is gifted with the beauty that is high intelligence, assuming society nurtures our wellbeing, and shelters us/meets our ‘hierarchy of needs’ (etc) - then we should give back in turn and not polute the good that others intend to do.

Now I am running with some assumptions here- first and foremost - that this thread is not simply an advert for some DCS kit. The thread seems to have started ‘the downward spiral’ with the mention of MQA (which I personally would rather be replaced with HDCD in a heartbeat - which is a format which gives to everyone/ not a case of ’haves/have nots’ and hits the magic number of 20 bit sound); I am not a fan of MQA, but I can see how it might reduce streaming bandwidth, which could prove a good thing. Of course I do not believe that was MQAs’ raison d‘etre (reason for being)..
Now as happy as I am to overhear a bunch of people talking about their experiences with high end audio kit, that wouldn’t be most peoples reason for coming to this thread.
Use your gifts people, and ‘if in doubt’... (say nothing) -try to let the person whose soap box we have come to criticise/hear (James) actually speak.

It kinda looks like it is a post that might grow with ‘one a week’.. the intention being to inform users why DACs are ‘not just the DAC chip’ but a ‘whole circuit’ and design matters.
Some education doesn’t hurt...
After reading head-fi for many many years, I finally had to start an account (in 2010) cause some muppet was running around, posting all over the place, some really negative reviews because of THIER BASIC MISUNDERSTANDINGS.

of course the internet gives everyone a voice..
Anyone who doesn’t want to hear what James has to say, give him the same courtesy as you would a person in public space- if he starts the conversation and is talking about a specific topic (DAC theory, which in this instance went ‘back to basics’ (as I listen to the album of the same name) and started with the initial encoding process), let the thread be about that topic, please.
This is really just *ahem* (common) courtesty.

If people want to talk about their experiences with DACs- go start your own thread. Clearly -hijacking- some elses’ nets you more attention. (If this sounds like an attack, well then, -=see it as that=-, as far as I can see, there are ‘drunk guests’ at this social get together who need to be kept in check, and whilst I don’t want to ‘mother hen’ someone elses conversation, I would like to represent the head-fi readers in the future who might come to read this thread years from now, and will be wondering WHY discussion of present market DACs are being spoken about in a thread giving digital theory.
And -YES- those smart enough to not need the education, or wish to correct another trying to get some general info out (burning the forrest to see the trees), maybe use the PM button, or do as some in the thread have done and give links that grow the discussion.

Lets assist our brethren to speak.. lets listen intently (requires mouths closed and EARs open), and lets acknowledge that the target audience for this thread might be a spectrum of headfiers, some new to the tech, some wanting to brush up their knowledge.
Now I know some ON THE SPECTRUM might wish to challenge ME over this; stating how important their opinion is on THIS TOPIC and WHY THEY NEEDED TO SPEAk.
(Go)Easy!
Un twist your panties.
All we need to do is let James talk about DAC theory.
End of (story)..

(I don’t need your ego battle- nor does this thread need anyones subjective experiences about their local hifi shops or why they went with a certain product- this isn’t the thread for that!)
I am not targetting anyone in particular- my sole purpose to write this is to return the ground work as laid out by James to be the topic, hopefully to be ‘resumed’.
If anyone actually knew me they would know I love to rant/rave as much as the next enthusiast and I AM a practioner of truth, and generally seek out accurate information. In any other thread I’d probably love to hear your respective stories, but ‘for those in the future scouring head-fi’ looking for nuggets of knowledge - lets let THEM have this one, shall we?
 
Apr 26, 2021 at 4:49 AM Post #32 of 187
The article is inaccurate and misleading from the start. The rectangles of sample showing a consistent value over the period of each sample is incorrect and misleading. Samples are values at instantaneous points in time, nothing more, a series of delta Diracs, infinitely narrow in time and with a given value. Nothing is defined about the value between each sample, so the sample-and-hold look of the rectangles should not be there.
This post is of course a simplified explanation in order to help those not as familiar with PCM get to grips with the topic, and to get some fundamental concepts across. This information doesn’t pertain to DACs directly here, but to PCM conceptually, as it could equally apply to ADCs.

You’re correct that digital audio samples are infinitely narrow Dirac pulses. The “staircase” plot here isn’t representing a DAC’s output, but the time interval between each Dirac pulse where an ADC would hold the value to allow for digitisation. Giving this notional impulse width (the stair on the graph) is necessary to move from the virtual world of PCM to the real world analogue.

The conversation around representations of samples as infinitely narrow would be better suited to a deeper conversation around digital filtering – which we do have on the cards for the future.
 
dCS Stay updated on dCS at their sponsor profile on Head-Fi.
 
https://www.facebook.com/dCSonlythemusic/ https://twitter.com/dcsonlythemusic/ https://www.instagram.com/dcsonlythemusic/ https://www.dcsaudio.com https://www.youtube.com/channel/UCQf2YCUG5UfXwZapkuTwxLw info@dcsaudio.com
Apr 26, 2021 at 8:39 AM Post #33 of 187
Theory is one thing, in practice is another.

Apart from having more bits to the depth for defining the signal better there's no point in recording stuff beyond the 44.1kHz sampling rate. Even this is true for recording and storing with all the limitations we have these days. It makes no sense when the music itself is produced using stuff which costs 1/10th price of what some of the OEMs are charrging to reproduce it, bear in mind the caveat that comes with mics having best case at 75dB tops.

Subjectively it makes no sense going beyond the 16/44.1kHz which Philips and Sony have nailed it. I will eat my hat if people can pick differences of a well recorded CD and it's HiRes counterpart when blind testing which is forbidden around these parts of the world anyway. HiRes only makes up for only 3% or 4% of the music anyway. Most of rock recordings are produced crap (Music is to die for though) anyway to start with.
 
Apr 26, 2021 at 10:03 AM Post #34 of 187
Theory is one thing, in practice is another.

I don't care what it is, but I prefer my music to sound pleasant. I always found the concept of "transparent" and "wire with gain" utterly boring to my ears. Wine is arguably not a faithful representation of grape juice yet I prefer the former over the latter. Likewise if some component adds coloration of its own that isn't present in the source but is to my liking, I'm all for it. for example my gsx mkii headphone amp seems to be praised as great and transparent but is impossibly boring to my taste.
 
Apr 26, 2021 at 12:27 PM Post #35 of 187
I am all for the scientists ensuring that their religion is artfully acknowledged (that plots represent information they are portraying etc), but I feel some leeway need be given here..
Why intelligent people need to ‘get in the way of’ simplified information for the masses (clearly the point for this thread) I am not sure- texts books would hint at ego needing acknowledgement.
”there there- we are all *special*“ (some clearly more than others)
Being scientist doesn't impose being intelligent. In fact scientists can be very dumb in the extreme case.
In a typical scenario some will use all their intelligence to make money - and then, if successful - they can afford to purchase a mysterious dCS Ring DAC. Those individuals don't need any technical education (even vastly oversimplified like it was made in the initial post). They need more a public appreciation for their money spent. It is why there is another type of trolling in this thread in addition to simply pointing of the evident errors.

Some more intelligent will go R2R route and get the same for a magnitude smaller amount of money. Those seems more educated, as they are resistant to the mass-marketting offering cheap-to-make Delta-sigma DACs with high markups due to the low BOM and those laugh on the efforts trying to explain advanced DSP methods, as they don't need any. They will stick to NOS, as it is a decoding method where the most listening pleasure comes from.

Whatever is behind of the idea, it is not profitable to make fundamental errors in the presentation or copy/pasting a material from the ADC section.
 
Last edited:
Apr 26, 2021 at 12:41 PM Post #37 of 187
Being scientist doesn't impose being intelligent. In fact scientists can be very dumb in the extreme case.
In a typical scenario some will use all their intelligence to make money - and then, if successful - they can afford to purchase a mysterious dCS Ring DAC. Those individuals don't need any technical education (even vastly oversimplified like it was made in the initial post). They need more a public appreciation for their money spent. It is why there is another type of trolling in this thread in addition to simply pointing of the evident errors.

Some more intelligent will go R2R route and get the same for a magnitude smaller amount of money. Those seems more educated, as they are resistant to the mass-marketting offering cheap-to-make Delta-sigma DACs with high markups due to the low BOM and those laugh on the efforts trying to explain advanced DSP methods, as they don't need any. They will stick to NOS, as it is a decoding method where the most listening pleasure comes from.

Whatever is behind of the idea, it is not profitable to make fundamental errors in the presentation or copy/pasting a material from the ADC section.
Wow you are a bit confused or just an angry person. I have owned nos dacs the metrum onyx which is a great dac in its own right. There are some r2r dacs that cost as much as any dCS have you ever seen an MSB? They are really pricey too. If you have not had a chance to compare them all in the same system your ability to judge them is skewed at the least.

I have had the metrum, chord TT2, and Dave both with the mscaler, Bartok, and MSB premire all in my system at the same time thanks to some friends. The dCS and MSB easily bested the others in terms of listening enjoyment.
 
Apr 26, 2021 at 12:52 PM Post #38 of 187
Wow you are a bit confused or just an angry person. I have owned nos dacs the metrum onyx which is a great dac in its own right. There are some r2r dacs that cost as much as any dCS have you ever seen an MSB? They are really pricey too. If you have not had a chance to compare them all in the same system your ability to judge them is skewed at the least.

I have had the metrum, chord TT2, and Dave both with the mscaler, Bartok, and MSB premire all in my system at the same time thanks to some friends. The dCS and MSB easily bested the others in terms of listening enjoyment.
All Chord DACs are Delta-Sigma, just in case you didn't know. And you make unfolded assumptions in your post. Are you angry? :)
 
Apr 26, 2021 at 3:30 PM Post #39 of 187
I don't care what it is, but I prefer my music to sound pleasant. I always found the concept of "transparent" and "wire with gain" utterly boring to my ears. Wine is arguably not a faithful representation of grape juice yet I prefer the former over the latter. Likewise if some component adds coloration of its own that isn't present in the source but is to my liking, I'm all for it. for example my gsx mkii headphone amp seems to be praised as great and transparent but is impossibly boring to my taste.
Good if you can enjoy but I am in pursuit of the truth and perfection!
 
Apr 26, 2021 at 6:15 PM Post #41 of 187
IIRC, dCS Ring DAC is like an evolution of DS and R2R, taking ideas from both but implemented in way that is unique (linearity error correction from temperature changes and other factors with DSP / FPGA) whereas something like Holo Audio R2R uses another R2R ladder for linearity correction, and Schiit Yggdrasil relies on "continuous run-in" to get the ideal performance / sound output. I'm new to the Ring DAC implementation so I'm hoping they cover that DAC section in detail
 
Apr 26, 2021 at 6:26 PM Post #42 of 187
This post is of course a simplified explanation in order to help those not as familiar with PCM get to grips with the topic, and to get some fundamental concepts across. This information doesn’t pertain to DACs directly here, but to PCM conceptually, as it could equally apply to ADCs.

You’re correct that digital audio samples are infinitely narrow Dirac pulses. The “staircase” plot here isn’t representing a DAC’s output, but the time interval between each Dirac pulse where an ADC would hold the value to allow for digitisation. Giving this notional impulse width (the stair on the graph) is necessary to move from the virtual world of PCM to the real world analogue.

The conversation around representations of samples as infinitely narrow would be better suited to a deeper conversation around digital filtering – which we do have on the cards for the future.
As this is a dCS thread and your products are not only highly praised but ultimately expensive, how about not pulling any punches with the dictation of the technical landscape. There are plenty of places to learn basic and fundamental architecture. This should be the place to explore and illuminate the nth degree.
 
Apr 26, 2021 at 7:07 PM Post #43 of 187
IIRC, dCS Ring DAC is like an evolution of DS and R2R, taking ideas from both but implemented in way that is unique (linearity error correction from temperature changes and other factors with DSP / FPGA) whereas something like Holo Audio R2R uses another R2R ladder for linearity correction, and Schiit Yggdrasil relies on "continuous run-in" to get the ideal performance / sound output. I'm new to the Ring DAC implementation so I'm hoping they cover that DAC section in detail
Not an evolution of R2R, no chance. It use multi-stream processing in a pure Delta-Sigma manner. Multistream allows to work with a lower clock frequency, also benefiting from randomisation of current sources, but it does not differ from a modern chip-DAC implementations. A time for a pure bitstream (one-bit) implementation was over already about 20 years ago and a balanced decoding is also a part of chip-DACs from a higher shelf.

To cut a crap short, Ring DAC is a discrete implementation of Delta-Sigma technology, the same as in Chord DACs. A discrete design allows to freely play with details of the implementation which is difficult or even imposible with chip-DACs, but it require more engineering expertise, so cost is increasing expotentially. A key point is to use FPGA for parallel processing, probably for both decoding and DSP, but it is a matter of convenience and cost reducing.

Before you lose a time reading this thread, follow this interview, you will see, nothing is special in Ring DAC. A shorth history of dCS is given there. There is also referal to MQA, no wonder, as both are trying to address some shortcomings in Delta Sigma decoding method
 
Last edited:
Apr 26, 2021 at 9:11 PM Post #44 of 187
All Chord DACs are Delta-Sigma, just in case you didn't know. And you make unfolded assumptions in your post. Are you angry? :)
I am well aware that all chord dacs are in essence delta sigma dacs. I was just stating that I found the dCS and MSB to sound better.

I find it funny how many folks will comment on dacs or other gear without a real world comparison. I have compared the above dacs in a headphone system with the Abyss TC, Susvara, Focal utopia, and the entire LCD4 lineup. They were also compared with the speakers that I had at the time a pair of Magico S1mkII. I would love an MSB Premire but I need to save up for that.
 
Apr 27, 2021 at 4:55 AM Post #45 of 187
I am well aware that all chord dacs are in essence delta sigma dacs. I was just stating that I found the dCS and MSB to sound better.
I am not trying to argue with, I don't know. However Ring DAC is not a new story. Many years ago when It was still innovative design I had an opportunity to audition, but was still thrown away and settled down for the inexpensive (chip-DAC) Burr-Brown PCM-63 implementation. It delivered more natural sound from a CD source and a cost? That was 10-times worth more comparing to other offering, but it is on a side note. I am trying to say, a DAC of your choice depends on a music of your preference. If your preference is electronic production type, you will be happy with dCS, Chord and more. If you attend a live performance of classical music, jazz, some pop, you would like to bring silmilar sound home.

The best Chord DAC is Mojo, everything above is vastly overhyped. Perfectly placed in the marketing target. There are better DACs in this price range, but at least it is a good match for your music taste. It is worth to note that while Chord is is offering such product, dCS do not.
 
Last edited:

Users who are viewing this thread

Back
Top