In the case of Grateful Dead, the actual source music is distributed as flac files because this is "lossless" compression technology. Whereas so much other music distributed is mp3, mqa or some other compression/lossy schema.
In the case of DACs, it’s my understanding that delta-sigma is "lossy" (i.e. all of the source PCM music just isn’t converted and played back) whereas an RxR DAC is "lossless", so all of the music is played. And if that is true, it becomes hard to imagine how delta-sigma could possibly ever sound better, regardless of how it is used or any other factor for that matter.
So in my mind, if in fact delta-sigma is "lossy" and does not play *all* of the music, it is automatically disqualified as even being a contender for what this thread is mean to address, Absolute top tier DAC for standard res Redbook CD.
EDIT:
So as silly as it seems, the essence of what I'm stating is that even *if* delta-sigma does sounds better in any particular implementation, who cares, if all of the music isn't being played? |
@jwn, @audioengr, @charles1dad, @ctsooner, @shadorne As I stated, it was/is my understanding. By all means "school me" if you have something to teach. The following excerpt from the Schiit website (as well as certain other posts on this forum and in the past) is one reason how/where I have come to my understanding. I grant you, this could be erroneous and if so I’m open to learning. The followimg is from: http://www.schiit.com/products/yggdrasil-- Forget everything you know about DACs. Yggdrasil is the world’s only closed-form multibit DAC, delivering 21 bits of resolution with no guessing anywhere in the digital or analog path. We’ve thrown out delta-sigma D/As and traditional digital filters to preserve the original samples all the way through from input to output. When doctors are trying to diagnose whether you have gas or cancer from MRI results, or when the military is trying to ensure a missile hits an ammo dump and not a nunnery next door, they don’t use “24 bit” or “32 bit” delta-sigma D/A converters. Instead, they rely on precision, multibit ladder DACs, like the Analog Devices AD5791. This allows them the bit-perfect precision they need for critical applications, rather than the guesswork of a delta-sigma. ----- Note the operative word "guessing". Secondly and in addition, I already have two different delta-sigma dacs within two of my other components (Oppo UDP-205 and Emotive ERC-3). Note the chips are entirely different (one being Sabre ESS-9038 the other Analog Devices AD1955). I’ll grant you, those components do not rise to the level of the components you envision, and I understand there is much more to the overall sound than the dac chip itself. Nevertheless, I find it more than curious and coincidental that despite the fact those chips are different, those two components sound virtually the same (to my ears) and not as "authentic" to my ears and the ears of a close musician friend I had listen to my system, specifically for the purpose of DAC evaluation. Moreover, I have read numereous posts from folks on other forums (head-fi as one example) who are (purportedly) musicians and "swear" by R2R in comparison to delat-sigma. In fact, many of the "negative sound qualities" I’ve read about delta-sigma in general I can relate to with my own ears. |
Gdhal’s comments just didn’t correlate with my actual listening experiences of many DACs of both type of designs.
Hi @charles1dad I would like to point out that the comment to which you refer I made, prior to your post, is with respect to the technology of delta-sigma itself, which would have nothing to do with subjective listening preferences. Not true. I think you are maybe confusing this with lossy formats. Hi @audioengr I note the occurrence of the word (including stemming) "approximate" appears no less than five (5) times in the following article: https://en.wikipedia.org/wiki/Delta-sigma_modulation |
Not sure why some are so adamant to keep on trying to say R2R is always better etc... @ctsooner I suspect the answer to your question has something to do with feedback (no pun intended) from those - such as myself - who have heard both technologies and merely wish to convey their findings. |
In fact the oversample file contains only as much information and all the information as in the original file and no more as it adds nothing to the signal. It also contains the entire data in the Redbook file - nothing has been thrown out at all (as would happen in a facsimile). This is yet another statement (on their faqs page) from Schiit: ************************** "We can’t get over the fact that delta-sigma DACs throw away all the original samples". ************************* And in this context they are meaning all delta-sigma, irrespective of implementation. So am I to be led to believe by those on the forum that the statement is false? And, if there is even an iota of truth in that statement, wouldn't that mean that my understanding - that delta-sigma is "lossy" - is true? |
I think you don’t know what your talking about. Delta sigma beat discrete how could that be if information is missing? @jwm In answer to your question, please refer to the post I directed to audioengr 01-13-2018 7:34pm. Thanks. |
@almarg
Thank you so much Al for putting all of this into lets call it "much better perspective". Given what you are stating - that in a sense everything in the recording and reproduction chain is lossy - can/would you provide your opinion, even if mere speculation, as to what Schiit is meaning by their statement:
"we can’t get over the fact that delta-sigma DACs throw away all the original samples"
This appears to indicate samples - which I assume and perhaps is implied is equivalent to music - is in fact lost, at least to a greater degree than could or otherwise should be lost.
|
However that says nothing at all about the degree to which musical information is lost in the process. @almarg Hi Al. Please allow me to preface this post by stating that in no way am I looking or wanting to be argumentative, and you already know how much I value your input in particular. Does your statement (quoted above) in effect mean "yes, musical information *is* lost, however, exactly how much musical information is lost is not known (or undefined), and presumed to be irrelevant"? |
@shadorne
Thank you. I mean that sincerely because you do in fact present an abundance of technical data in support of delta-sigma. In my case, I suppose I'm fortunate to have components with both technologies, and have carefully listened to both. While I may not have a 30K divinci dac - which uses delta sigma - I'd have to imagine it sounds spectacular to say the least. All I can indicate with certainty, however, is that to my ears, I prefer the sound presentation of my multibit DAC to that of my delta-sigma. Not only because it sounds more analog, but because it sounds more "real" and authentic. And I get what Al is saying - that those qualities are a result of the implementation and not the technology. So be it. Anyway, who knows what tomorrow's technology will be. Enjoy the music while we can.
|
....but the issue with Delta-Sigma is the filtering. The issue with R2R is this non-linearity. @audioengr Perhaps I’m "fanning the flames" here just a bit, nevertheless, I’d like to comment on this in particular because I was/am well aware (so I think) of the "filtering" long before I posted in this thred. My Oppo UDP-205 (delta-sigma) has a user preference setting whereby the filter can be changed. Way back, I did some research as to what this entailed, and found it had all to do with "pre-ringing", "post-ringing" and basically everything at the 20 KHz point. I tried all (7) available filters and none made the sound "better" than the default (mini-phase fast). It only made it as good or worse, to my ears. So, what "filter" is the issue with delta-sigma? |
So, both techniques can have good S/N (Delta-Sigmas have improved a LOT since the 90’s), but the issue with Delta-Sigma is the filtering. The issue with R2R is this non-linearity. This is what I had said all along. Eliminate the bad digital filtering and you have something really fine with Delta-Sigma. This is what my ears tell me. @audioengr Thank you for your response regarding my question about the filter, and pointing out that you remove or replace it in your design, and how that differs from "apodizing". It would seem Benchmark also replaces the filter, given what shadorne states about how the ESS chip filter is overridden. Okay. Fair enough. Regarding the non-linearity, this is corroborated by Benchmark. According to Benchmark (this was forwarded to me by Rory Rall at the time I was in the market for a DAC and Benchmark was on the radar): "The distortion in a ladder DAC is caused by resistor mismatch. This resistor mismatch causes linearity errors. The step sizes between adjacent digital codes are inconsistent. Even if the resistors are precision trimmed, they drift with temperature. This drift is not just a function of the ambient temperature. The change in resistance can even be induced by the instantaneous heating caused by the audio signal." So it would seem (to me anyway) that the non-linearity issue you are referring to is a function of *resistor temperature*, which lends credence to why Schiit recommends leaving the unit on 24/7. That said, I can tell you that even from a cold (off 12hrs) start, and while playing and the unit is warming, my Yggy sounds better than both of my delta-sigma, which have "digititis" by comparison. This is what my ears tell me. |
gdhal - what source is driving your Yggy?
What are your Delta-Sigma DAC’s? Musical Fidelity M6si > Belden 5T00UP > Golden Ear Triton Reference (pair) Schiit Yggdrasil > Canare L-4E6S XLR > M6si Oppo UDP-205 > Belden 1694A (Canare RCAPs) > Yggdrasil Emotiva ERC-3 > Belden 1800F AES/EBU > Yggdrasil Samsung SMT-C5320 > Mitsubishi Eska POF > Yggdrasil EDIT: Steve, the Oppo and Emotiva (both delta sigma) which now act as transport were my primary source before I obtained the Yggy. I did perform extensive A/B testing, and that saga was documented here: https://forum.audiogon.com/discussions/new-yggdrasil-first-and-second-impressions |
I feel so confident in this dac that it would be interesting to see if MSB could actually beat it. I would hope so at 100,000.00.
@jwm I’m just curious, what music do you prefer to listen to? After looking at the MSB price sheet http://www.msbtech.com/image/Retail_Price_List_2016.pdf I’d also be curious to understand what you perceive is missing from lesser DACs, that these MSB DACs offer. Thanks. |
I remain to be convinced that there is still meaningful progress to be made in DACs for 16/44. However, some may find the following link on filters interesting: http://archimago.blogspot.nl/2018/01/audiophile-myth-260-detestable-digital.html Hi @willemj I looked over the article. I believe the author is referring to the same type of filters I mentioned a few posts back, which Steve states are merely "apodizing filters" and not the same type/kind of filtering required for good delta-sigma implementation. I could be mistaken though (and if so I have no doubt Steve or Shadorne will let me know :) ) |
@shadorne
Thank you so much for that filter information. May I ask your opinion then - and I realize this is somewhat off topic thread, apologies to the OP - which filter (all names are listed below) to use in the case of an Oppo UDP205, which you know uses the Sabre ESS9038 chip?
The manual provides no detail what-so-ever as to the nature or function of each filter. I'm told "brick wall" is meant for "diagnostics", so I've avoided that. Based on my own listening impressions, I find the default mini phase fast to be preferable. "Apodizing" is the term Steve mentioned, but that was applicable to all of these filters, true? (note there is only one specific filter named "apodizing")
In your previous post to me (01-15-2018 12:00pm), it would seem one of the "linear phase" filters would be your choice, true?
Brick Wall
Corrected Mini Phase Fast
Apodizing Fast
Mini Phase Slow
Mini Phase Fast (default)
Linear Phase Slow
Linear Phase Fast
|
If you are using Oppo as a transport with 1694A, you are getting a LOT of jitter.... Hi Steve. Thank you so much for your feedback. As to the jitter, this is somewhat of re-hash of recent communication between you and I in another thread. I’m not disagreeing in that regard, however, I’m essentially relying on the Yggy to "do its job" and cleans the incoming signal (at least as best that it can and purports to do). Of course it makes sense that to feed any DAC signal which is clean to begin with is preferable/better, and I imagine at some point I’ll investigate that. By this I mean both transport and cable. Also, at the moment I do not optimize the sound potential of the Yggy because I power it off/on each listening session, a cardinal sin by some folks standards. Keep in mind I’m happy and content with the current sound from my system. And while I haven’t personally heard hundreds of systems, as you probably have, I have heard more than a dozen rather high end systems, including one with a pair of 200K pair of Alta Statement tower speakers and comparable gear to go with it. I’m confident in the overall sound quality level of my current system. Can it be improved, sure. I suppose this is one reason why the forum is so valuable; to get ideas and implement where practical. I chimed in on this thread because it was (past tense) my understanding that delta sigma is lossy. To that end, you, Al and Shadorne have educated me to the contrary and I no longer believe that. Of course, I reserve the right to change that belief should some "higher authority" offer into evidence data to the contrary. Given the many posts in this particular thread, it would appear there is no absolute top tier dac for standard redbook, at least from the technology perspective. I get what you, Shadorne, Al and others are stating about the "implementation" of the technology. Nevertheless, in the very end I’d have to think that subjective listening impressions and preferences could prevail. And in my particular case I understand what you’re saying about a unfair comparison of multibit and delta sigma on the basis of my components (because of different overall quality and price points). But at the moment, short of spending more money, I'll have to live with what I have. I can only report my findings and preferences based on what I currently have, so in this context, I prefer multibit to delta sigma. Steve, I remain very appreciative of your kindness and information. |
Hi @shadorne Thank you so much for the recommendation(s). In the case of the filter option, I will change to linear phase slow as you suggest (even though I rarely listen to the Oppo directly at this point). As to Roon, I’m not sure why I would need that. One major source of my PCM based music is lossless legs. This is a bit torrent peer-to-peer trading community web site: https://www.shnflac.net/index.php?page=torrents&search=&category=0&active=0If you care to have a look at the source of the actual show/music, just click on a show and note the torrent details. While the precise lineage isn’t always available, the majority of the time it is. For example, gd1970-09-20.sbd.boswell.smith.miller.clugston.flac2496 details are: Source Info: Master Soundboard Reels (borrowed by Peter Kafer) > Will Boswell’s reels @ 7 1/2 ips Transfer Info: Reel (Technics RS1506) > Tascam DA-3000 (DSF 1bit/5.6 MHz) > dBpoweramp 24/96 > Adobe Audition 3.0 > TLH flac2496 I don’t think there is anything I need to do here in the way of upsampling or otherwise altering the source. Am I correct in my thinking, or is there something I could or should be doing to further optimize the quality? *********** By the way, @shadorne and at @audioengr in appreciation of your kindness, I’d like to offer the gift of music. Feel free to pick something (a show) you would like, and I can make it available to you as a direct download (in flac or shn format). http://halr.x10.mx/shows.html (grateful dead) http://halr.x10.mx/other.html (all artists other than grateful dead) The quality (sound and performance) of most shows is pristine :) |
@audioengr and @shadorne
You're most welcome. Give me a day or so and I'll PM you the download links. Alternatively, you can send me an email at the address on my web page (probably not correct etiquette for me to list that here) so I can reply to you directly. |
Tom Petty does a great live jam called Two Men Talkin’ that seems to be a tribute to Grateful Dead - do you know what is the original Grateful Dead jam this is based on?
https://youtu.be/HU9VBQBFQ_w Of course. While I never heard the Tom Petty song until I listened given the link you provided, I can instantly recognize the music. While completely different lyrically, musically this is the Grateful Dead’s "The Other One". You can hear a Grateful Dead - The Other One, from 9/28/72 (any date would do for sake of this comparison) here: https://www.youtube.com/watch?v=72iVSo6iXz8 |
They are probably the best band in the world to see or hear live! Have you seen or heard the Grateful Dead :) |
The Yggdrasil has a very good S/PDIF receiver, the AK4113. Same one I use. The results should be stellar with any good low-jitter source. Unfortunately, you are using very poor S/PDIF cables. See my cable jitter plots here:http://www.audiocircle.com/index.php?topic=154425.0 @audioengr Hi Steve. I did have a look at your web page with the measurements of different cables. Nice work on your part. Note that the Belden 1694A I use with RCA is because the Oppo does not have a BNC output option. It does have optical, but I’m not using it (with the Oppo, I am using optical with the cable box). Anyway, what is your opinion of using USB in the following configuration: Gateway NV79 > Belkin e129760-c usb > Yggdrasil Should USB produce better results than the Belden 1694A coax? Should there be improvement if I were to switch the optical from my cable box to the Oppo? (note that according to Schiit, AES/EBU is best to the Yggy, followed by coax) I only have AES/EBU out from my Emotiva. I also recall reading from Georgehifi that some type of glass optical is best, which I’m also not using. |
OP, getting right back to the heart of the topic thread now :)
Regarding the cabling and/or transport that could or do introduce jitter and other artifacts, isn't that a moot point where an "absolute top tier dac for redbook cd is concerned"? And why, because wouldn't the "absolute top tier dac" reconstruct (if needed) and clean the signal?
Consider the Yggy. Not only do they reclock and so on (as I;m sure other DACs do), but they (purportedly) go another extent. Their DACs have an indicator light to let the user know the signal source is "crap", and *even then* they claim to be able to make an improvement. |
@shadorne and @audioengr
Please check your private message box and/or private email for a link to download the music you requested of me.
Enjoy!! |
|
I thought the Emotiva was a DAC Yes and no. The ERC-3 is a CD player, and does have its own internal DAC (delta sigma). It also has digital output capability which bypasses the internal DAC. I am using the digital AES/EBU output and have it connected to the Yggdrasil AES/EBU input. (which by the way, Moffat claims AES/EBU to Yggy is best) |
post comments to ask questions and make suggestions! Nice room! The audio level on the YouTube video is a bit low. I find the audio level to be fine. The opposite of Steve, meaning I could lower the volume (on a laptop PC with internal speakers) mid way and it's still fine. |
For the money, the Schitt Yggdrasil. Everything you read about this thing is true and its convinced me that R2R design is the most musical and natural sound. I would be surprised if I ever purchase another delta-sigma design again. + 1 |
I’ve been moving around the whole thread. It’s always done meticulously and with purpose. And everything is compared A/B with all variables kept exactly the same. My side comparisons happen when DAC or server comparisons are complete. And every comparison is ALWAYS done with all other variables set in stone. Only one thing changed at a time when doing formal comparisons. @mattnshilp Given the number of posts in this thread, the nearly four calendar years since its inception and your meticulous standards, can you state at this time the answer to your post question? In other words, to date and in your opinion, what is "the absolute top tier DAC for standard res Redbook CD"? If you can elaborate as to why, that too would be appreciated. Thank you. |
|
I spent a lot of time on Sunday at the Schitt table. I know I'll get flamed for this, but I really felt let down. I heard their top offerings in the most recent form I'm assuming and wasn't impressed. You won't get flamed at all. The Schiit Yggdrasil isn't for everyone. It's akin to what Jerry has to say about the Grateful Dead. “We're like licorice. Not everybody likes licorice, but the people who like licorice really like licorice.” |
I wanted to convert these to the same bit depth and sample rate and in .wav format (storage is so cheap now). While I certainly do agree that "cheap" or "expensive" is subjective, I humbly submit you are probably making a mistake to store as wav (instead of a lossless compressed alternative). Assuming your collection grows, at some point you will be compelled to rethink storing files in wav format. In my particular case, I have 20+ terabytes of music stored as flac. If I stored in wav, I suppose I could say I have 40 terabytes. Besides the fact the cost of three additional western digital drives would be $600, I’d also need another cubic feet or more of physical space. Not clear to me where the benefit lies. |
How do you get time to listen to it all?
I listen at a rate of 12 hours of live concert recordings weekly, for many years now. Each day I listen to a completely different show. In my case, I primarily, but not exclusively, listen to Grateful Dead. To date, I’ve listened to 4772 *unique* show hours and 2026 *unique* dates . My 20 terabytes is not all unique shows. Much of the material is the same show, but a different source. For instance, a show can be a soundboard, audience or matrix. I have about 1 year of music in queue. This is music I’ve already obtained but haven’t listened to it yet. I also have videos of many shows. Videos are typically 12Gb whereas a show in 24bit is 3.5Gb. Details are on my website http://halr.x10.mx/shows.htmlEthernet renderers, as audioengr refers to, are the way to go. This isn’t practical in my case because I obtain and trade via client peer-to-peer bit torrent. I "collect" the music so as to retain in my personal possession. Grateful Dead can be streamed via ethernet rendering from archive.org, but my collection far surpasses what’s available on the site and is usually better quality sources. Note that at one time archive.org permitted downloading the files and not simply streaming as they do now. Moreover, IMO streaming is not as good as having the source files and playing locally. Streaming invariably involves occasional drop outs, especially if wireless is involved. My opinion on streaming music is that quality suffers in comparison to having the same music available locally. Streaming might be viable if I were just casually listening and not collecting on top of it. Additionally, not all types of music - especially live recordings - are available to stream. |
Why do you believe that Ethernet renderers = streaming? Hi Steve. If I'm mistaken here (and you're saying that I am) then I appreciate your pointing it out. My DAC (Yggdrasil) does not support Ethernet in, and I'm in no hurry to get another one. But would I be correct in saying that your wav files that are from CD rips
and downloads from HDTracks, that you are storing on Raid1 still has a cost of the storage device itself? And, is the cost of the storage more or less than $200 for 8 (eight) TB (terabytes), which is my cost for a western digital drive. And, aren't you using at least twice the storage space you would otherwise use if you stored as flac? Perhaps your configuration is capable of the best quality, and it that case I'll have to be content with less than best. That said, I'm curious if you are able to listen via "blind testing" and conclude the same results as your measurement data? |
I didn’t bother with that in this case. I am a trained listener and do this every single day, listening for changes, problems, nuances. Been doing that for 42 years. I don’t doubt my conclusions. I made business decisions based on them.
You are welcome to doubt them if you want. It’s still a free country, at least last time I checked CNN.
Hi Steve. I’m not doubting you at all. As I stated, I was just curious if you could blind test. I'll take your response as a "no". Given everything I’ve read from you over the years on the forum, invariably you have all the answers and the "best-of-the-best" gear and technology at play. Happy listening. |
Please lets be civil and well meaning in our conduct, or I’m out of here. +1 I absolutely agree with you amg56. But, being civil and well meaning in our conduct includes you too. You mayn’t go around questioning ones intelligence, simply on the basis of a forum post. |
...After researching the different types of digital topologies and learning about what the math and science actually means, I came to the conclusion that a 20-bit R-2R ladder DAC has the purest form of PCM decoding.... And so I understand from other sources as well. Thanks for your confirmation! |
What does that tell you? That my Amy > Bob > Amy > Bob test is best. |
charles1dad - Hello Steve N, George frequently to support his point says R2R is "bit perfect" whereas delta sigma is merely a "facsimile". As a listener I don’t hear R2R superiority. As a builder/designer is this claim true and if so is it an audible or meaningful distinction in your experience? @charles1dad With do respect, I would just like to point out that not only does George make the claim as stated in your quote, but so too do many DAC manufacturers. |
@charles1dad
Understood. Thanks. |
@sadono
You write with wisdom like a child, directly to the heart. I trust you will interpret my questions as genuine.
Simple yes/no questions if I may... no need to elaborate beyond yes/no but by all means do so if so inclined...
- does this 16 vs 24 bit "difference" effect your enjoyment of listening to music?
- are you able to discern a difference between 16/24 in a blind test (assuming you’ve tried)?
Thanks. |
|
shadorne - Here is a thesis paper explaining why R-2R DACs have lots of harmonic distortion. Looks like the distortion is consistent across all harmonics.....
Hi shadorne. I have no disagreement in what you are stating in your previous post. I would just like to submit as food-for-thought that Jimi Hendrix used fuzz distortion. Not quite the same as the distortion you are speaking of, nevertheless, worth mentioning, IMO. EDIT: ....that would be *intentionally* used.... |
audioengr - ....The question is: given both designs are excellent, which one sounds more live to you?.... Hi Steve. Not meaning or wanting to play on words here, but for clarification purposes, by "live" do you mean "authentic"? Thanks. |
audioengr - ....By live, I mean when I play a live track, say a jazz performance in a restaurant, can I close my eyes and believe that I’m actually sitting in the restaurant? Can I turn my back on the speakers and believe that there is a live performance behind me? Does a good piano track sound like my Yamaha Grand Piano in the other room? Hi Steve. I get what you're meaning now by live. To my understanding the audiophile term for what you've described is "suspended disbelief". True? I'd have to think the DAC design (DS or R2R) may not be the dominate factor in achieving what you describe. Big factor though, certainly. I would think the room, the speakers and the recording itself would be more prevalent. So when you state that the DS was better - at least with your designs - I would have to think your designs evolved over many years, and in that case the other variables I mentioned (room, speakers and so forth) may not have remained constant. In any case, I appreciate your clarification. Thanks. |
shadorne - Even harmonic distortion is pleasing. It is higher order odd harmonics above 7th which are not pleasing and harsh sounding.... Hi shadorne. Thank you. Not all that long ago, Ralph (atmasphere) provided a wealth of information regarding harmonic distortion in this thread https://forum.audiogon.com/discussions/full-detailed-sound-at-30-40-50-db (probably other threads to :) ) Your response is very much in line with what I’ve previously understood in this regard. In fact I understand there is "weighting" of distortion also at play. Noting that manufacturers typically provide "total harmonic distortion" and in some cases "intermodulation distortion", I had asked in that thread if manufacturer specifications could provide insight as to where the harmonic distortion is occurring. His response led me to believe "this can be tricky". Point is one may not readily be able to determine where the distortion is occurring, and may only be able to rely on subjective listening. Your explanation certainly does make sense. Would you say then that the *implementation* is the dominant factor of listener preference between DS or R2R, or is the *distortion* now the dominant factor (again, in your opinion)? Thanks. |
headphonedreams - I am also in the side that thinks that all common architectures (R2R, DS, FPGA) can work when implemented with care.... Hi everyone. I’m pretty good at using google. That said, may I ask the opinions of the folks on this thread for a very brief (sentence or two) explanation as to the advantages and/or disadvantages of "FPGA" in comparison to the other two architectures. Thank you. |
