1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)

Discussion in 'Dedicated Source Components' started by purrin, Dec 5, 2013.
400 401 402 403 404 405 406 407 408 409
411 412 413 414 415 416 417 418 419 420
  1. evillamer
    on the topic of mastering, check out this video:
    It seems that the mastering engineer tunes it by his trained ears and the inputs from the client.
  2. BassDigger
    On the contrary; I like your use of the word 'magic'. It implies the occurrence of something wonderful, that we don't understand.
    It seems that many of us agree that R-2R dacs do something better than D-S dacs, but we're not sure what or how.
    No, actually you're right; it's not magic; it's witchcraft! [​IMG] 
    (BTW: I appreciate the guys technical explanations, but not his analogies, and I do think he's doing it on purpose. I've had enough; I've other things to do.)
  3. evillamer
    I think this post by Jason Stoddard(of Schiit Audio) is the perfect counter arguement. However we must note that jason posted this story section before keith's post.
    I have truncated his post due to length, you can read his full post here: http://www.head-fi.org/t/701900/schiit-happened-the-story-of-the-worlds-most-improbable-start-up/6855#post_11730790
    Argo Duck likes this.
  4. jcx
    no one denies you can color or flavor your sound with electronics design and component choices - however it is hard to understand why this is the "superior" position to having accurate, transparent electronics which many of us with decades of EE experience and similar time hobby interest in audio think isn't that hard today
    and Jason is (ab)using "objectivists" in a classic rhetorical style - anonymous, context free "they said" attributions
    I am disappointed in fact given the early tone of "letting us in", "cutting the hype", his background as an engineer - but given his marketing and professional writing experience it is hard to believe he isn't being deliberate in his choice of rhetorical tools
    I don't know what others said but just recently somewhere on head-fi I pointed to reasons to believe leveraging ADI's institutional knowledge and $Billion fab line with the AD5791 was likely technically superior than what a MSB is doing for high multiples of Schiits pricing
  5. drez
    I can say quite easily what R2R foes well in audiophile terms. Timbre and tonal colours are typically denser and richer in texture and more natural. Treble is not as aggressive. I don't understand the concept of plankton or how this is better. I find in most cases there are tradeoffs to make these qualities. In my limited time with Yggy I was unable to determine any tradeoffs. Just a fuller acoustic, which I associated with improvement to my DS dac also. Perhaps the acoustic was a little less rry and reserved than I am used to. To be honest though the real question is why R2R sounds this way. Is this more accurate or less so. This really needs to be answered. Going from measurements would appear many R2R DAC's are less accurate. Apparently Yggy's chips are more accurate than DS, but I don't understand this. I was told DAC accuracy is different from effective resolution? ??
  6. AudioBear
    I'm big on neutral and transparent so I can use software equalization to produce the sound I want.  I expect the electronics to just plain get out of the way and disappear so I can shape the music the way I want it.  It's ok me with me if people want to be purists or subjectivists or mix components to achieve the same result because it's a hobby they enjoy.  I just hope that they realize that the "bit perfect" sound on the CD or download they are listening to has been acquired by an imperfect microphone, digitized by an imperfect AD and mixed, equalized and styled by software.  It's not real, it's shaped and there's nothing wrong with reshaping it any way you want (with the components you chose or the software you use)--that's your choice.
    I take Jason and Baldr to be trying simply to give us their best effort at translation to analog and amplification of the bits that are on the CD and then get out of the way.  That suits me fine.  Many companies take that approach, others have a house sound, and still others have many flavors available.  I prefer the vanilla so I can add any flavor I want.  I just don't understand why equalization held in such low regard.   I don't see it as any different than picking a bassy headphone.  But that's just me.
  7. evillamer
    Speaking of which it will roughly cost 40% more for audiophile grade R2R vs audiophile grade Sigma Delta if you based off Audio-gd dacs.
    TOTL Balanced R2R: USD$2180
    TOTL Balanced Sigma Delta: US$1380
    Single Ended R2R: $750
    Single End Sigma Delta: $499
    Is 40% more worth paying for? Maybe we have to ask @stillhart, @bimmer100 or someone who have owned both audio-gd sigma-delta and r2r dacs.
  8. BassDigger
    I get the point; comparing like for like; use components from the same manufacturer to keep everything as similar as possible.
    But, if it is concluded that the R-2R dacs sound an amazing 40% better, than their 40% cheaper D-S siblings, aren't people still going to say that a 40% more expensive D-S dac will be just as good?
    I think that maybe this kind of comparison should be more about defining what the differences actually are between the technologies, regardless of price. It's already a 'given' that R-2R is more expensive.
  9. wahsmoh
    I just got some powered monitors that I'm going to hook up this weekend after I build some speaker stands. Fostex PM841 are 3-way speakers tweet, 4'' midrange, and 8'' woofer. I figured I would go powered monitors since I do not have a large power amp for speakers. I am going to be using my Asgard 2 as a preamp and Theta DAC (R2R) being fed S/PDIF as my source. I haven't had the chance to test out the speakers  yet  but I am very excited to get them positioned and see how well they image.
    I do recall briefly testing out my M-audio BX5d2 (entry level cheapo monitors) and they sounded a lot better with the R2R DAC and more 3D in presentation but still noticeably weak in the midrange and tizzy on the highs. I have been working a really good job so my speaker/dac/amp budget is going to change here really soon. 
  10. KeithEmo
    I wasn't picking on you for using the word magic - and I'm sorry if I came across that way.
    Honestly, no, I don't just "think you're all crazy". However, I'm not nearly that certain that you didn't "just happen to hear R2R DACs that you thought sounded good and D-S DACs you didn't like". And, in my opinion, this possibility is made more likely by the current market. (If I'm a small company, selling an R2R DAC, I'm going to make sure that it sounds like what the majority of my potential customers - who've come to me because they want an R2R DAC - expect it to sound. So my DAC could sound good to you because the technology itself is superior, or simply because I've made sure to "voice" it to sound like what I know you want.)
    I don't believe that "all chips sound the same", but I do claim that the support circuitry makes a bigger difference than the chip. I also don't know for sure if there is some specific error that D-S DACs unavoidably inflict on the signal that R2R DACs do not... but I most certainly haven't seen any convincing proof that there is. To me, saying "they sound better" is a good first step, and the next step is to determine why, which then leads to "fixing the problem across the board". However, this discussion seems to have stalled at "they sound better". (And the few claims made against D-S DACs, like that they may modulate a noise floor that is already below the noise floor of most decent power amps, don't really seem to me to explain why they might sound audibly different.)
    (The errors introduced by oversampling, which we seem to agree aren't an audible problem, are pretty easy to explain, measure, and see on an oscilloscope trace. However, I haven't seen any plausible claims that the output signal of "a bunch of R2R DACs that sound good" shares some particular trait or lacks a certain error that is shared by "all D-S DACs that sound bad". And, yes, I would like to know, once and for all, if all the R2R DACs you especially like can reproduce a certain signal accurately that a D-S DAC cannot, of if they share some particular inaccuracy or flaw that you happen to enjoy, or, for that matter, if you're just imagining it, or maybe if I'm imagining that I don't hear it.)
    I also DO have a lot of faith in "the market". In other words, if we do eventually figure out what the actual "problem" is, and there are enough people who can hear it and agree that it's a problem, ad they're all willing to pay an extra $5 to avoid it, I figure that TI probably probably will "fix" it in next years $5 DAC chip.... which would benefit all of us.
    However, before they're going to listen, we need to tell them functionally what the problem is.... (as in "with this signal, the scope picture looks like this, when it should really look like that"). As a start, if the folks who seem to believe that "the big problem with D-S DACs is that they modulate the noise floor" are right, that should be easy enough to prove.... all they have to do is to create a test signal that contains the same modulated noise floor they insist sounds bad on D-S DACs and play it through an R2R DAC - if they're correct, then doing so should make that R2R DAC "sound exactly like a D-S DAC)... which would go a long way towards proving both that they do sound inherently different, and what the difference is... and show the guys who make the chips which problem they need to fix.

  11. jexby
    faith in "the market"? feh.
    the DAC chip makers market (to profitability) is not tied to audio fans searching for holy grail of quality,
    but to massive adoption in consumer grade equipment sales.
    now that the original author has left this thread, happy to do the same myself.
    bicker on.
    dan.gheorghe and BassDigger like this.
  12. KeithEmo
    If you were hoping that I was going to disagree with Jason, then I'm going to have to disappoint you... I don't disagree with anything he said. Delta-Sigma DACs were designed to deliver the best possible performance and price/performance ratio - BASED ON WHAT THE OVERALL DAC MARKET ASKED FOR. You can pretty well guess that this includes low noise, high precision (accuracy), low distortion, and other such things. Since TI, and their competitors, sell literally tens or hundreds of millions of DACs, and there are several alternative manufacturers offering competing products, you can indeed assume that they succeeded pretty well. However, you can also assume that the vast majority of their sales are to companies other than producers of small-market audiophile products, so it's quite possible that their priorities, and those of their customers, are different than yours or mine. Which makes it quite possible that a small company, with "audiophile priorities", may well be able to create a product that better meets "audiophile needs" by doing it a different way.
    (However, it isn't automatic, and so it's also quite possible that the massive development and research budgets available to big companies do actually often enable them to produce a chip for $2 that really is better all around.) However, because there is a lot of competition, and most of their customers almost certainly do have lists of what they do consider important, and would probably notice if the chips they were buying didn't work exactly as advertised, I think you can be pretty certain that they aren't "trying to sneak anything by anyone".
    The TI PCM1704 (one of the last remaining available 24 bit R2R DACs) was scheduled for end of life several years ago, but sufficient interest from audiophile manufacturers convinced TI to continue to produce them... at a price sufficiently inflated to compensate them for the aggravation of continuing to produce a product with high production costs and low sales volume. However, you will note that the part is now listed as "end of life" and the "recommended current alternative part" is in fact a Delta-Sigma type (it's actually a hybrid, where some bits are rendered "directly", but most are handled by Delta-Sigma circuitry). If you think about it for a minute, you will also realize that this status prevents most larger companies, and those with large design budgets, from using this chip; after all, who is going to spend a lot of money to design a product that uses a particular part, knowing that the part may soon become unavailable? Someone who builds products ten at a time in his garage, or even 100 at a time in a small facility, can afford to use a part until the bin is empty; but someone who plans spend $1 million designing a unit which he hopes to sell 10,000 units of next year, can't risk having to redesign it all over again when he hits the bottom of the bin.
    The problem that I find with many audiophiles is a tendency - which I share - to "enjoy the chase" and to "seek for treasure". Quite simply, we would rather believe that we've "found a treasure made by a little old guy living in the woods" than that the mass produced $2 part made by TI just may be superior. The subject then drifts into "the objectivists", who are usually characterized as "only paying attention to the numbers", and who like the $2 chip "because it measures well", and "the emotionalists", who honestly seem to just plain be unable to even consider the possibility that a cheap mass-market product might actually do what it's supposed to do pretty well, and often end up insisting "it just sounds better to me". (SOmetimes they are able to bakc up that claim in a real double-blind test that excludes their expectations from the equation; other times not.)
    (Another not-very-well-kept marketing secret is that "exclusivity and scarcity breed perception of increased value". In other words, tell someone that they're part of an exclusive group being offered a product, or that "they may not be available soon", or "we only have a few left", and most people will automatically "feel" that it is more valuable and want it a lot more. THis is very well known, and is the basis of every "quantities are limited" and "exclusive offer" ad you've ever seen on TV. Since R2R DAC chips, like the PCM1704, are legitimately "scarce and soon to become unavailable", you also have to assume that the fact that they are tends to bias people in favor of their being "rare and valuable" - which, in turn, leads people to expect "something special" from them.)
    There are more than a few small audio companies out there who produce badly designed audio products, based on faulty engineering, and which measure and sound as bad as that would lead you to believe, but who still have a "loyal following" of people who simply like the way they sound. (Please note that I absolutely do NOT include Schiit Audio in that category. Their products are well designed and, as far as I know, all perform as they were intended to.)
    As for what role expectations play in what we hear, and what degree marketing influences what companies sell, I'll just leave you with one comment about Schiit's products....

    If they didn't want the knowledge that it's an R2R DAC to influence your opinion of Yggdrasil, then they wouldn't tell you it was an R2R DAC.
    My point, though, is that the fact that it costs more to build an R2R DAC is simply a fact.
    It does suggest that the current market doesn't support the sale of quantities of R2R DACs.
    However, neither of those in any way suggests that R2R DACs are (or are not) inherently better.
    Therefore, as a bottom line, I'll give you the same advice that Jason usually does....
    If you've got $x to spend on a product, listen to all (or at least several) of the options in your price range, and buy the one that sounds best to you.
    gevorg and Articnoise like this.
  13. BassDigger
    Ok, I'm in a mood to humour you some more.
    Your desire for an explanation is perhaps expecting a bit too much; maybe a 'sound science' type discussion, with the mad scientists that lurk around there, might get a little nearer to the answer (or just completely deny that there's any possibility of a difference). 
    But, for what it's worth, and as I've stated before, my belief is that it's perhaps a timing issue; the fact (as I understand it) is that D-S does more processing and dividing of the signal, than R-2R, and then has to reconstitute it, at high speed, making it more likely that timing discrepancies (yes, such as jitter) will have some kind of an audible effect.
    We've been through this before; I said and you said, but the testimony of other experts, and in my experience the key thing that R-2R does better is musicality. Another adjective (that could mean anything to anyone), but I'm referring to the complexities of beats and rhythms that are much more realistic, 'funkier' with r2r.
    Now, where can we find a 'funkyness' meter? [​IMG] 
  14. evillamer
    Last I checked the pcm1704uk is still listed as NRND and not EOL/obsolete/end of production.

  15. preproman

    Great post.
400 401 402 403 404 405 406 407 408 409
411 412 413 414 415 416 417 418 419 420

Share This Page