Columns Retired Columns & Blogs |
Time for Stereophile to review the Aurender W20 and/or the Aurender W20SE :-) .......
Being a Sharpener does affect how I've set up my reference system and how I approach components under review. What some may consider a needlessly complicated, overly expensive system, I've found essential for distinguishing subtle differences. To help ensure that my conclusions are fair, everything from cabling to equipment supports remains identical from one review to the next, except of course the component under review. Save for the times when I say "eff it all," turn off my critical mind, and sit back to enjoy, my nature as a Sharpener colors everything I contribute to this magazine. That's something readers have a right to know.
Why do I say this at the beginning of a review of the Wolf Audio Systems Alpha 3 SX audio server ($9295 base price with 2TB SSD, or $9895 as auditioned, with optional Flux Capacitor USB clock card)? Because as I went back and forth between the Alpha 3 SX and my reference Roon Nucleus +, I was aware of focusing on details that would be of minimal importance to many others, or even inaudible in some systems. As you read, please keep in mind that, once I had more or less figured out how to operate the Alpha 3 SXa complex component whose Linux-based operating system is quite different than this confirmed Apple user's referenceI really enjoyed what I was hearing.
What is it?
I've developed an aversion to using Swiss army knife analogies, and yet, here I go: The Wolf Alpha 3 SX is similarly versatile, albeit less potentially dangerous. It stores and plays back music, rips the contents of CDs to its internal drive, and includes a TEAC Blu-ray transport that can play DVDsboth video and audioas well as Blu-ray discs. The Alpha 3 SX handles many formats, including WAV, AIFF, FLAC, ALAC, AAC, M4A, MP3, OGG, DSF, DFF, DXD, SACD ISO, and MQA. It can decode PCM up to 32-bit/1536kHz, DSD up to DSD1024 (either Native or DoP), and it does multichannel DSD up to DSD256.
The Wolf Alpha 3 SX works with your choice of music-playback software, including (but not limited to) JRiver, Roon, Audirvana, or Amarra. Because those applications integrate various streaming servicessome or all of Tidal, Qobuz, Google Music HD, Spotify, Deezer, etc.it can stream music as well. The server's front panel is home to an on/off button, a Wolf logo, and a thin disc slot that, from a distance, looks like just a black horizontal line. (Apparently I'm not the only human who at first mistook the disc transport for a design element intended to make the server's plain front appear more interesting.)
On the rear panel sit, amongst other things, one network port (Ethernet), two HDMI ports, four USB-3 ports (two of which pass data through the optional clock card), and lots more. I can't detail them all because, during the review period, neither a manual nor online photos of the rear panel were available. (Once installed in my system, the back of the Wolf was not accessible.) In their stead, I received a Red Wolf 2 manual, which I promptly abandoned after discovering how different its rear panel looks. Hopefully, by publication time, a manual and photos will be available.
Lacking documentation, I was dependent upon the assistance of Wolf co-founder Joe Parvey, who often used TeamViewer to log into and monitor my server. We communicated a lot by phone, email, and text. I'm told that Wolf, which incorporated in 2015, has five people devoted to technical support, and that timely assistance such as I received should be available to all Wolf owners.
Wolf servers run WolfOS, a Linux-based operating system the company introduced at the 2019 Capital Audiofest. Before that, Wolf used a Windows-based version. I'm barely treading water when it comes to explaining such things, so it's best to quote Joe, who co-founded Wolf in Florida with fellow "technologist and audiophile" Fred Parvey, Joe's dad:
"We have designed a custom kernel, the back end or core of the OS, to offer the lowest latency possible, as well as keep the audio processes of the system carefully isolated and separate from the rest of the server. This allows a Wolf server to work as closely and seamlessly with DACs as possible and provides users with the ability to add on applications as they are needed or become available."In my opinion, the OS that runs all the apps is as important as the apps themselves, because how it is run and configured is absolutely critical to good sound. Our extremely low latency WolfOS is absolutely tuned for audio and has a big impact on sound quality."
The Wolf Alpha 3 SX is a one-piece unit, but Joe feels that "its full potential can be unlocked with a monitor and keyboard, which is especially useful for downloading video." Joe sent, with the server, a remote control receiver that can be used with an Apple Remote or a home automation remoteI didn't use itand a keyboard ($225), as well as a 22" Planar touchscreen monitor ($350 with 25' cables).
Another Alpha 3 SX option, which I also used, is the Flux Capacitor USB clock card ($600), which reclocks the audio signal prior to sending it to your DAC via USB. Yet another Wolf-provided upgrade is to one of three Audience power cords ($200$1310). "We were very lucky to meet some great people in the industry who helped answer my questions and explain things that I didn't quite understand," Joe said when he visited to set up the Alpha 3 SX. "Those people include John McDonald and Lenny Mayeux of Audience. When Wolf first began, we played with linear power, dismissed the commercially available linear power systems, and decided to make our own. Then, when we switched to using Audience copper internally, we had this huge bump in sound quality. It was a collision of circumstance and really good people that helped us start Wolf Audio Systems."
I did not explore the power cord option, which would have added another variable to the review. Instead, I stuck to my reference Nordost Odin 2 power cables. The Alpha 3 SX has Wolf's most powerful processor, largest amount of RAM, and best, fastest, and largest system drive. Case treatment, power supply, and quality of copper wiring also distinguish the 3 SX from the basic 3. Joe Parvey insists that the Alpha 3 SX is a "big upgrade over the standard Alpha 3," which uses an i5 processor, 8GB RAM, and a 250GB SSD for the operating system and a 2TB SSD for music storage. The SX ups that to an i7 processor, 32GB RAM, a 1TB system drive that runs at 3.5GB per second, and a choice of SSD music storage, currently topping out at 12TB. "Pairing the right RAM with the right motherboard and processor is very important," Joe said. "We spend a lot of time doing arduous listening and parts matching in every Red Wolf 2 and Alpha 3 SX server."
The "SX" in Alpha 3 SX stands for "Stillpoint/eXemplar Audio" and refers to grounding, vibration management, and RF/EMI rejection technology designed by those two companies. "The Alpha 3 SX contains roughly $2000 worth of Stillpoints, eXemplar, and Wolf technology that take the Alpha line to the nth degree," Joe said. Wolf intends to upgrade the company's flagship server, the two-chassis Red Wolf 2, to both Red Wolf 3 and Red Wolf 3 SX status by the end of 2020 because Joe considers the Stillpoints and eXemplar Audio additions major improvements.
"The 3 SX includes new, patent-pending grounding technology that was jointly developed by Wolf, Stillpoints, and eXemplar," he said. "It sits directly on the hot, neutral, and ground lines of the incoming power and pulls MHz- and GHz-spectrum noise off the power lines before it reaches critical components. It also contains the same Stillpoints standoffs inside every Stillpoint Ultra Mini filter. That technology, which consists of two pieces of metal separated by a pocket that holds ceramic bearings, is used to hold up the server's critical internal components. Imagine each internal component supported by four Stillpoints Ultra Mini feet."
By publication time, Wolf anticipates an additional optional upgrade that will replace the unit's stock feet with either three or four Stillpoints Ultra Mini or Ultra SS filters. Cost varies from three Ultra Minis ($375) to four Ultra SS filters ($1000). While screw-on Ultras weren't available during the review period, I tested their efficacy by using three and then four of the Stillpoints Ultra SS filters that I had on hand. The only difference between Ultra SS and the Ultra Mini is the size of the bearings.
Setup, software, setbacks
I placed the Wolf Alpha 3 SX on one of the two top shelves of my double Grand Prix Audio Monza rack; on the other sat the comparison server, Roon's Nucleus + ($2499), which was powered by an HDPlex 4-unit linear power supply ($485) I use in place of the Nucleus +'s switch-mode power supply (footnote 1). All three units were supported, initially, by Grand Prix Audio Apex feet and connected to Nordost Odin 2 power cables. A Nordost Valhalla 2 USB cable connected the servers to either the dCS Rossini DAC or EMM Labs DV2 DAC; I switched between them. A Wireworld Platinum Starlight Ethernet cable hooked the components up to the network.
Moving back and forth between servers involved a simple switch of three cables and devoting up to five minutes to rebooting servers and apps. When using Roon with the Alpha 3 SX, I sometimes needed to reboot again after I'd take a break for the night. After restarting, everything worked fine, most of the time.
During Joe's visit, we attached a 22" Planar touchscreen monitor to the Wolf via HDMI and positioned it next to me. The monitor kept going blank and then returning every time Joe tried to do something. Since he was loading my music onto the server and working with code, this proved frustrating. A replacement monitor worked far better, but its touchscreen needed far more pressure and more repeated poking than my Apple touchscreens require. Touchscreen-typing into "Search" proved frustrating, so I frequently resorted to the wireless keyboard Joe provided.
Time for Stereophile to review the Aurender W20 and/or the Aurender W20SE :-) .......
"focusing on details that would be of minimal importance to many others, or even inaudible in some systems"
The subtle changes can sometimes make the biggest change in enjoyment.
I often find that changes that I initially think sound either the same or better make me enjoy the music less over time. I don't consciously think "This doesn't sound as good as it used to" instead I just find myself listening less. Eventually I realize that I'm not turning the system on as much and then think back to why and realize it started at the component change.
I take it this can play a CD and send it direct to the D/A converter? like a CD transport without any storing and replay from the HD or memory?
If so it would have been good in the listening tests to A/B the CD direct to dac vs a copy of it from the units HD.
Cheers George
georgehifi wrote:
I take it this can play a CD and send it direct to the D/A converter? like a CD transport without any storing and replay from the HD or memory?
Well, you can play a CD from the internal drive but it is clumsy and inconvenient and depends on the software player you choose. AFAIK, for, example, JRMC can do it but Roon cannot. In addition, since it is playing through the same software player, it is not going from the drive "direct to the D/A converter.
I have a similar Baetis Prodigy-X and I reviewed a related Wolf Audio player and, until you asked your question, I had never even tried to play a disc in either. The drive is there mostly for ripping.
I wonder whether the drive can even spin at CD playback speeds like a conventional CD transport ...... It is possible that the drive in the server can only spin at 2X, 4X or more speeds, which is useful for CD ripping ....... It is also possible that the mechanism is designed to do both ie. playback and ripping :-) ........
"Well, you can play a CD from the internal drive but it is clumsy and inconvenient"
I don't care about that Kal, I care more about which way sounds better. To me that's the end goal for being an audiophile, not to sacrifice sound quality for a little convenience, especially for the princely sum of $10K.
And if it does "only spin at 2X, 4X or more speeds" as Bogolu thinks, then it's a non-event, as it's not reading direct from CD to dac.
Cheers George
Hi, I'm Wes from Wolf Audio Systems and wanted to respond to some questions that I've seen people asking.
What Kal said is correct, direct CD playback is not necessarily the point of any server. While the feature is there, it isn't the way that we prefer to go about things, which was why it was not specifically tested A/B in the review. We're of the belief that a better way to go is to rip your CDs to the internal storage, and not have to use the physical media any more than that. Personally, I find CD rips, when done properly, to be overall better sounding than direct playback, since once the files are there, there is no chance of the playback being affected by fingerprints, scratches, or mechanical noise from the spinning disc, and it's a bit-perfect reproduction every time. The drive however, does have the ability to spin at 1X speeds, in order to achieve accurate reading of the disc.
Hi, is there any relationship between the rate of disc spinning and digital data retrieval for storage in the server? ...... What is the maximum optimum spinning rate? ....... 2X, 4X, 8X or more for example? ........ What is your opinion? :-) ........
For $2000 it should play a CD or DVD smoothly. Build a pi music server connected to a big HD and control it remotely for $200 or less.
Kelaur wrote:
For $2000 it should play a CD or DVD smoothly.
It sure does play a CD or DVD smoothly but that's not what you are buying this for.
"Personally, I find CD rips, when done properly, to be overall better sounding than direct playback, since once the files are there, there is no chance of the playback being affected by fingerprints, scratches"
This is silly, as if there were errors playing the cd then the errors will still be present on the rips, plus whatever else is added by the ripping process!!
Cheers George
The error correction system commonly built-in the DACs will most likely take care of that ....... Of course, I'm not disputing about the errors in CDs being transferred to the storage in the servers :-) ........
The error correction system commonly built-in the DACs will most likely take care of that ......
DACs don't perform error correction. That occurs in the transport's decoder. If you rip a CD using Exact Audio Copy, you can set it to correct errors to the best of its ability, using multiple reads if necessary. and comparing the data against a database (a process called "AccurateRip"). If there are errors that can’t be corrected, EAC flags the time position when these occurred so that the resultant file could be fixed.
John Atkinson
Technical Editor, Stereophile
So, is it possible the CD rip stored in the server could sound better than the original CD with errors, like was mentioned by Wolf representative above? :-) .......
So, is it possible the CD rip stored in the server could sound better than the original CD with errors, like was mentioned by Wolf representative above? :-).
Yes, though with modern players, like the T+A 3100 reviewed by Jim Austin in the new (June) issue, they can have superb error corrections/concealment.
John Atkinson
Technical Editor, Stereophile
How does the T+A 3100 compares to the new $5,500, Mark Levinson 5101 CD/SACD player/DAC? ...... We can only find out, if JA1 reviews the ML 5101 :-) ........
Jim, the MP 3100 HV and PDT from T+A are amongst the finest players on the market today. We use both (among other things) when tuning our systems
-joe
Also, lot of CD, SACD and universal players are not connected to the internet ...... So, those players cannot verify with the database :-) .......
what we have is a device that sometimes takes a back seat sonically to a similar gadget that costs 1/3 the price. Admittedly, the high priced spread does give you an optical drive, but even that is not intended for real-time playback.
As an owner of both a nucleus and a Wolf (not this version), I feel the need to correct a couple of things. First, the Wolf does not take a "back seat" sonically. Jason's review mentioned a couple of passages (known to be "hot") and liked what Nucleus + power supply did. My older version Wolf certainly has not been bettered (it has been best sounding unit that I have used). Secondly, the Wolf does a LOT more than the nucleus. In fact, it is this flexibility (including video, which enhanced my movie watching experience and was great for the kids) that separates it from competitors, many of which use processing systems no more robust than last generation cell phones.
"I prefer my music neither warmed over nor darkened. If a recording is brash, let it be brash; I'll turn down the volume or hold my ears." I guess holding your ears is the price to pay for having really expensive equipment with no tone controls and no coloration, even euphonic coloration that helps old, dry recordings sound a LOT better. Some of the best music is found in some of the worst recordings, and if you want to hold your ears, so that you can maintain your sacred purity as a purist, well good for you. I personally prefer to enjoy my music when I listen to it rather than holding my ears. Isn't that what this hobby is all about? Why avoid pleasure so studiously? Jeez.
Of course !!! Finally someone nailed it!
It's about enjoying music through the hi fi system and not the other way around. It is and always will be that way, until "perfect music forever" become truly present, that day all hi fi brands will disappear. Meanwhile, like it or not all equipment MUST have tone controls to taylor the sound to the person is hearing it. Jeez...
latinaudio wrote:
Meanwhile, like it or not all equipment MUST have tone controls to taylor the sound to the person is hearing it. Jeez...
Sure although many other will disagree. I don't use tone controls, per se, but I do find it easier to tailor all sorts of EQ with file playback than with disc playback.
I recently played Miles Davis' 'Miles Ahead.' The sound on the CD is bright and a little sharp; as an experiment, I listened on a non-audiophile Onkyo receiver (albeit using the 21-bit resolution of an NAD M51 and very good cabling). Adding the bass that the original recording lacked gave me a very enjoyable experience -- it sounded WONderful. Speakers were Buchardt Audio.
I really wish more equipment gave us at least the OPTION of making adjustments like this. But purists insist the original recording -- no matter how bad -- is sacred, and they'd rather hold their ears than make it better or restore what should have been there but isn't. Seems kind of crazy to me.
By the way, the 50 watt Onkyo TX-8020 sounds amazingly good for the price ($200). Full-bodied, solid bass, rich midrange, surprising detail. Could be a little more controlled, but it's going to be one of my listening options from now on. Thank God for Japanese traditionalists.
I totally agree with the concept that music has the first fundamental goal of bringing us joy, and not a fight with our senses in order to preserve an alleged accuracy.
And I must reinforce the “alleged” part of the sentence, because even if we could be 100% sure that nothing was lost or changed between the master recording (and this is usually a big “if”), we would still hit the hard fact that it’s very unlikely that two studios have the exact same gear/monitors and acoustics.
And even if they did, not all audio engineers will have the same taste or be mixing for the same final consumer (most records are mixed and mastered with the best “compliance” possible between several lower to mid consumer grade products, not caring about audiophile standards).
And then we still have different headphones, loudspeakers and listening rooms introducing a huge “sonic signature”, which will have most likely the biggest impact on the so-called accuracy.
Unless the listener was in the recording venue, had the opportunity to compare the live sound with the one being recorded (making clear mental notes of the differences introduced right there), and then playing it back on a given sound system (which is a very hard exercise by itself given the range of human audible memory), absolute accuracy is a moving target.
I understand the concept of avoiding EQ (or simple tone controls) on some levels, but having to bear harsh sound in name of accuracy (without even knowing if the audio engineer who did the final mastering wasn’t over-compensating for a laid-back monitor system/acoustics) is the kind of self-imposed audio suffering that I honestly don’t understand.
On the subject of the review itself, if we’re comparing two very competent (and I write this as an understatement) digital audio players and one of them is harsh where the other one isn’t, I would be inclined to believe that one of them is anything but accurate on turning those 0 and 1’s back into an analogue signal.
We’re not talking about cheap components found inside budget smartphones that have to be able to be an jack of all trades.
I believe that having more examples and photos of the actual process of using this server would have added value to the review of this product.
But these are only my personal opinions, which can be as wrong as any other. :-)
"That's because there's nothing wrong with being a Sharpener."
Well, yes, there actually is- it reveals a lack of proportion and can make even the most informed reader believe that minor differences are in fact major differences. And once that's figured out by the reader, the reader may then tend to view the highlights of the review as hyperbole.
I enjoy JVS's reviews, and I bumped into once at a show and found him a very pleasant and gracious fellow. I've always thought of him as the audio reviewer equivalent of Linda Evangelista in that he won't get out of bed to review a components that cost less than $10,000. He certainly has a very fine reference system and is a meticulous tweaker yet we learn in this review that his room has a 'slap echo—on my end, an issue that will disappear once I find room treatment I can afford'.
I couple of things made me curious - why such a tweaker would not address room acoustics and why are solutions not affordable - compared to high end components acoustical treatments would seem relatively affordable? Over thinking this makes me wonder, then, is JVS's reference system loaned and not owned. And if your reference system is loaned do you then have to be a more positive reviewer to maintain good industry relationships?
My reading of this particular review is that JVS spent many words saying he found his nucleus+ with LPS preferable to the Wolf server. That doesn't mean you can't be positive about the Wolf but at more than twice the price I feel there should be a little reservation in the summary.
13DoW
1. The disc drive in the Wolf is intended for ripping discs which it does well.
2. The disc drive in the Wolf is capable of playing discs smoothly and clearly but that it not its purpose (see above). Note that I did not include the word "direct" in this statement because disc playback is routed through whichever music player software you are using.
3. In view of the above, the only significant comparisons should be made between file playback on the Wolf via a good DAC and disc on your choice of a disc player with or without an external DAC.
4. No one buys (or should buy) a music file streamer/server to play discs.
May I add ...... Don't buy a smartphone, if you just want to make and receive phone calls :-) .........
Poor analogy. A "smart" phone is a tool that is intended to do as many things as possible. Audio components have defined roles....... unless you include the "smart" phone as an audio component. I don't.
Aurender A30 is kinda 'smart' server/streamer/CD ripper/pre-amp/headphone amp/DAC with multiple digital reconstruction filters :-) .........
Then it should be judged against all the products with which it is designed to compete.
The disc drive in the Wolf, as well as the one in my Baetis, is a convenience, a tool. With both of them, I tested the ripping function for my reviews but, afterwards, I rip with other hardware and I definitely play discs with other hardware..................if I ever bother playing discs.
Aurender A30 was reviewed by Hi-Fi news with measurements :-) ......
BTW ....... LG smartphone reviewed by Stereophile is kinda audio component :-) ........
JA "DACs don't perform error correction. That occurs in the transport's decoder"
+1
Who ever heard of dacs fixing CD transport errors, 1's instead of 0's or 0's instead of 1's
So in the error department, it's impossible for this server to sound better than the CD disc it copied, it could only sound worse, with even greater error count than the CD it just copied.
And so we go back to my first question in this thread.
As I look at these servers as an inferior storage mechanism for my prized cd collection. As I've yet to hear one in my system sound better than my CD transport using the same dac, playing a copy of that same CD in an A/B.
So for $10k I would want better or at least equal to the sound of what it copied from.
Cheers George
Have you read JA1's comments about the CD errors could be corrected by the servers? ....... See, JA1's comments above :-) ........
It can't be as it doesn't know if the error needs a 1 or a 0 put in to fix the CD error read.
And you've got the same from the CD transport, so now you've got double the amount of errors, like I said at best it can equal the CD, highly unlikely odds of that happening, but the odds are it will be worse.
On top of all that now there will be error reads from the HD read as well.
Cheers George
See JA1's comments about server access to a database and 'AccurateRip' :-) ........
I just told you, it can't be the server doesn't know if the error needs a 1 or a 0 put into the already "fixed" (with a guess of 1 or 0) from the CD transports error correction. So sorry you are wrong and what you "think" JA said about it.
Cheers George
So, how come JA1 said the CD rip from the server could sound better than the original CD physical disc with errors? :-) ........
The database should be able to tell the server to put 1 or 0 in the appropriate place :-) .......
The database stores a small checksum per track, not full audio data. This is enough to determine if tracks match but not to correct errors.
See, JA1's comments above ...... What is 'Exact Audio Copy' (EAC) and 'AccurateRip'? :-) .......
CD error correction is made by CD ripper (transport), not the database :-) ........
It's a dead end argument, the server has no idea what the CD transport has corrected already, all it can do is error correct it's own HD read errors.
Now you have double the amount of errors corrected with a guess if they should have been a 0 or a 1.
The more reads and re-reads you have in the chain the worse thing get to being "bit perfect" That's why the stamped retail CD is the best.
Cheers George
The server has two parts ..... The transport and the storage ....... The transport rips, accesses the database, make any corrections, and then provides that data to the storage which stores that data :-) .......
That stored data is presented the DAC ...... I don't think any further corrections will be made by the server ...... JA1 could correct me, if I'm wrong :-) .......
The first part is an option that is not needed in a server. The files can come from many sources as have most of mine. Many are downloads. And all it does is rip wherever it is.
The server and its software do all the database operations as well as playback.
Yes, agreed ...... For example, some models of Aurender servers have built-in CD rippers, while some other models do not ...... However, an external CD ripper can be connected to their servers, if CD ripping is desired ...... Aurender recommends 'Acronova' CD rippers for use with their servers ....... Acronova says their CD rippers can rip CDs at 40X speeds! :-) .........
KR could review the Aurender A30 and the Acronova CD ripper all at one time :-) ........
"KR could review the Aurender A30 and the Acronova CD ripper all at one time :-) ........"
Not gonna happen. First, I have no personal interest in any streamer/server that does only stereo unless there is some other feature that I find attractive. Second, I do not think that a Stereophile review of an automated CD ripper fits the editorial profile. (Jim will advise if I am wrong.) So, here's my review: I bought an Acronova Nimbie USB NB21-BR with a group of friends to rip my CD collection. Running it under dbPowerAmp it did a great job.
That is great ....... That review of Acronova Nimbe is good enough ...... Somebody else at Stereophile could review the A30, if they choose ....... A30 was already reviewed by Hi-Fi News with measurements ....... If anybody is interested in A30, they could read the Hi-Fi news review ....... That review is available online :-) ......
As a side note ....... Stereophile has previously published reviews of accessories, for example ...... PS Audio Power Plants, AudioQuest Niagara power conditioners, Grand Prix Audio Monza audio racks etc. ...... May be you (KR) could write a short review/article about the Acronova Nimbe USB NB21-BR :-) .......
I do not believe that it is worthy of any page space. As it runs under dbPowerAmp, the results will be (are!) no different from any other CD drive that runs under dbPowerAmp. We all have those in our computers and/or can add one cheaply. The Nimbie's only significance is its ability to load up to 100 discs at once and that is well described on the website. What else do you want?
Ok, that is fine ...... I was just making a suggestion ....... You (KR) already mentioned in your previous comment that Acronova Nimbe works great ....... That is good enough :-) ......
Here are 3 CD's
https://ibb.co/kg5Gz4q
Stamped original retail
Burnt gold disc
Burnt standard disc
You can easily see which is going to give the most read errors
Cheers George
The examples you are showing appear to be CD-R/RW copies made from original stamped retail CD ....... We don't know what machine was used to make those CD-R/RW copies ....... We don't know whether that machine is connected to internet or not ....... We don't know whether that machine has access to the database or not, which JA1 mentioned above ........ The CD rippers usually don't work like CD copiers ....... CD rippers can spin at multiple speeds and can access the database like JA1 mentioned ...... CD rippers usually don't make CD-R/RW copies :-) ........
If it's possible to play CD from this Alpha 3 SX to a dac in real time, then there will be 2 different error corrections going on, if the same CD was written to the hd and then hd then replayed to the dac.
And therefore the CD direct will have had less errors ( 0's or 1's guessed) and corrected. And sound better for it, the blind can see this.
Cheers George
reading a CD via optical transport includes regular control of check sums but note, a CD has only approx. 10% of redundancy information [i.e. check-sums] stored compared to a proper PC file such as *.wav or *.flac on a HDD or SSD, so it is much more sparse in comparison.
In case check-sum errors occur, a real-time CD replaying CD transport may re-read the relating part: I would however not assume that all CD transports re-read the last turn ad infinitum or until all checksums are finally correct, as there is limited time for correction before the data buffer is empty (if so the music stops playing, or skips a tiny part, perhaps not noticed by the listener). Not sure about buffer size, but I'd assume rather a few kB or so, i.e. tiny compared to the track size.
Instead, when the server (PC) reads a disc it checks the CD check-sums (as any transport would do) but eventually also checks the check-sum of the complete scanned track (file) against values stored in data bases, and if it is not correct, it re-reads the track a number of times until correct, and it does so reading with 8-times or if need be even 0.1 times real time, dependent on how scratchy the CD. In the end, such a scanned CD is, unless deeply scratched, by all means 99.99% correct or better. You will most likely not get this data (stream) quality from a real-time played CD, perhaps also not with a 10k transport on good days.
Storing and re-reading the scanned music data in a file on HDD or SSD happens with 10x more checksums compared to data on a music CD, and as experience tells with normal files on a PC ssd or hdd, hardly ever a file has flipped bits or is corrupted, reading time from latest SSD is super fast, ca 1GB per second (~1.5times an entire CD, that allows for some comfortable re-reading if data is erroneous, which on the other hand would be much more obvious because of 10x more redundant "check data" stored with "music data"). So no issue here, neither with error correction nor timing.
Summary:
1) scanning a CD (to HDD or SSD): one more extra source to identify reading errors (data base), AND, should errors occur, there is plenty of time to re-read the CD until all check-sums are correct. Replay is not introducing errors (well.. no errors comparable to uncertainty of real-time CD replay).
2) real-time CD replay: very limited error correction (if at all), at some point re-reading must be skipped anyway to keep the music playing.
I, as a reformed IT guy who spent too many hours building decently high end computers and using them for hi-res recording long before it was a thing, can't get over that this is a 10k linux box. And that everything about it that is given as a reason for it's cost is A, nothing new or B, highly unlikely to improve it's sound. Fancy grounding schemes, while sometimes novel in audio, shouldn't be - but it's been the only way computers have been designed. Almost forever. Because noisy grounding affects speed and performance and may keep them from working, at all. So it's not something advertised, it's assumed.
I really hate to be a stick in the mud here, but this is a very expensive solution to problems that don't really exist in high quality computer audio. I'm sorry, it just is.
That said, different strokes for different folks! Enjoy it if you can, but you can buy a lot of music, good food, wine and whiskey for 10k bucks!
...in a custom case rather than one you can buy on Amazon?
https://www.amazon.com/Streacom-Alpha-Fanless-Chassis-optical/dp/B00P1TVK72/ref=sr_1_1?crid=3SI1ZA0RNHZ6K&dchild=1&keywords=streacom+fc10&qid=1588632100&sprefix=streacom+fc%2Caps%2C211&sr=8-1
Don't judge a server (book) by its cover :-) .......
...Bogolu. JVS and JA1 praised the sound and measurements. But you don´t want the BMW technology inside the Ford Pinto platform...
May be they could consider designing a box looking like a D'Agostino amp, with a large meter in the front :-) ........
JVS is waiting for that gee-whiz, state of the art, cutting edge dCS, 50 grand server ........ There won't be any 'vague testimonials' ........ JVS is gonna declare that dCS server as the world's best and the best server he has ever heard :-) .......
My first wife was a book editor and she had a sign over her desk which read, "Anyone who says that you cannot tell a book by its cover never tried to sell one."
May be the same thing goes for music album covers :-) .........
Take a look at the book cover 'Don't burn this book' by Dave Rubin ....... That book title is also interesting :-) ........
I always wondered why Stereophile print edition covers look so attractive :-) .......
... here:
https://www.quietpcusa.com/Streacom-FC10-ALPHA-HTPC-Chassis
Any mini-PC would do better and would have better form factor. It would also cost 400 USD.
I would be ashamed to be seen with such monstrosity in my home. It would show how ignorant I am.
Copying from it's CD transport to it's HD and then replaying from it's HD, has to have more read errors that were "corrected", than just replaying from the CD transport itself.
(when I say "corrected", that's 50% chance of getting the corrected errors right.)
If you can't see that you can't see the forest for the trees.
Cheers George
What makes you think it is only 50% chance of getting the corrected errors right? ....... It could be 60%, 85% or even 100% correcting the errors right :-) .......
You're saying the transport will interpret errors, there may be different errors in writing to HDD/SDD, and yet other different errors introduced in reading the HDD/SDD (worst case)? Instead of just the transport reading and interpreting a single set of errors, yes?
If I have that correct (or not), that thought intrigued me. So, I went to see my SMART error codes as an (admittedly anecdotal) test. I found no reports of read or write errors between the three. One with 20,000 hours on it unfortunately doesn't show how much has been written to or read from it, but it was my drive for music recording and mixing for years. A couple dozen or so projects had been worked on through it. I also know I've had it near full quite a few times before offloading the projects to other drives, so there will have been a good number of write/read/rewrite/read...etc. going on in it.
What does this mean for any of us? I don't know; nothing, likely. It's just interesting, especially thinking of how many errors could be introduced throughout the recording, mixing, and mastering process - if any.
George seems to be saying that the storage (HDD/SSD) will introduce additional errors, in addition to the transport errors ...... I don't think that happens :-) ......
There's no reason they couldn't, nor even something like a bit getting flipped in DRAM, I suppose could occur.
It got me to thinking 1) does this happen regularly, and 2) if so, is the occurrence enough that it pushes the errors above what a transport would impose? From what JA mentioned, though, the CD transport thing could be largely mitigated with Exact Audio Copy's AccurateRip (I can't believe I never heard of this before!). But, if EAC isn't used or doesn't have the data on hand to help the rip's accuracy, is it so much more that it would perceptibly affect audio quality? I don't know.
My guess is that the chances of DRAM and the storage, especially SSD storage, introducing errors is less than 1% ....... There is even less chance that being audible :-) ........
JA1 mentions about 'concealment' of errors in his comments :-) .......
I didn't mean that AccurateRip was a way to switc the bits or anything like that. Flagging suspect locations and (I'm assuming from reading EAC's website) they also use that information to direct the program where to attempt to read the area again at a slower speed. Really nice features for us who would like to know the rip was as good as it's going to get.
Like I said, I can't believe I never heard of it before. Or, I have and always just read about it with glazed-over eyes. Completely something I do.
since writing a file to a hdd or ssd to my knowledge ALWAYS includes test-re-read, looped until re-read results to original bit-sequence.
Plus the process of writing and reading files on a proper storage unit in a proper file format includes lots & lots of redundancy information allowing much more frequent checks than with music data on a music CD.
So the bottle neck is the CD & reading via CD transport, particularly without help from external check-sums (such as via AccurateRip).
Bogolu Haranath "What makes you think it is only 50% chance of getting the corrected errors right?"
Because it doesn't know when it's corrupt or missing, it has to guess 1 or a 0 which is a 50% bet in anyone eyes
And what it does, it duplicates what came before it 1 or 0 and sticks that in there.
Cheers George
Ok ...... So, how can that damaged CD with errors sound better in a regular CD player with regular CD transport? ...... Can it? :-) .......
May be the CD player does 'sins of omission', while the server does 'sins of commission' :-) .......
The only way to know is to play each and every CD on a regular CD player and do a listening test comparison with what is recorded in the server ........
The Acronova Nimbe CD ripper, which is mentioned above, discards unrecordable CDs to the server ........ I don't know whether the Wolf server has that particular feature or not ....... Probably the Wolf server has that particular feature also :-) ........
If you are using any other digital reconstruction filter other than the standard linear phase filter in the CD player's DAC, you have to use the same filter in the DAC which you are using with the server, for the listening tests ........ Bryston BCD-3 and Hegel Mohican CD players come with minimum phase filters, for example ....... Good luck :-) ........
"I have no idea what they mean"-Really JA?
What your measurements show is that the $10K server makes no difference once you have a competent DAC like the Mytek, and makes a small, almost for sure totally inaudible difference with a cheap, intro DAC like the DragonFly.
In terms of a PC, you can buy the same parts and software (except for a few audiophile add ons) for $3500. Pay someone to build it for you and spend $4000. You just saved yourself $6000. Is Wolf support worth that much?
You can also forego the $750 USB card, because it also probably makes no difference.
In the review of the Innuos Statement you had extremely similar results, and came close to admitting you really couldn't hear a difference between it and other PCs/servers.
Will Stereophile ever allow it to be said that these expensive audiophile devices contribute nothing to improved SQ?
If someone wants to pay thousands for the company support or the cosmetics - please do. But don't tell us that you are getting improved SQ.
How does it compare to Antipodes CX/EX? Also DX Gen 3?