Announcement

Collapse
No announcement yet.

Have we forgot about the Server

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #31
    Originally posted by Kingrex View Post
    I think our idea of a server is a little different. It sounds like you're really using the device in a more robust Pro Audio type environment. My idea of a server is more simple. It is more for playback only. It is most always Network compatible to utilize a NAS or the internet for streaming services such as Qobuz or Tidal. These devices generally run $5,000 to $20,000,
    The functions are the same, only the names are changed. NAS and all streaming services and devices fall into the server/client category, and as such, all use TCP/IP for transport, which is asynchronous, packetized, error corrected (including request to resend bad packets), and assumes an "unstable" network.
    Originally posted by Kingrex View Post
    In my scenario a non audio computer is a device with a large switch mode power supply and multiple programs running such as QuickBooks possibly CAD, Excel spreadsheets and worksheets, photo editing software, email and virus protection Etc.
    Ok, but just saying "switch mode" isn't always bad. Just like there are good, bad, and ugly in everything, there are good, clean, low noise switch-mode power supplies. Apple products, for example, generally have had outstanding and clean switch-mode power supplies.

    While I appreciate the concern for the processor load of other apps, it's not how many, or even how much RAM they occupy, its their usage of CPU resources that matters. For example, QuickBooks hogs almost zero processor resources unless it's actually being used. Most Microsoft Office apps also can sit resident and mostly asleep indefinitely without impact to the system. You can see this clearly by looking at CPU resources in Task Manager.
    Originally posted by Kingrex View Post
    The purpose-built server I am alluding to would have either a custom Linux software extremely strip down and incapable of most all general purpose computer tasks. It may also have a Windows environment operating system but will most likely be running software such as audiophile Optimizer to turn off as much processes as possible.
    Sounds like a nightmare. Pick one OS, stick with it. We have servers running Windows and Linux. I personally find the Linux servers to be more difficult to manage. I've also had a few macOS servers. They all do the job just fine, but managing multiple OS's is a major PITA requiring more mental gymnastics than I like to perform.

    Serving audio files is a very easy task. A garden-variety NAS should be able to sever several concurrent clients with hi-res files without a problem. 24/192 isn't even 10mbps, and we're dealing with gigabit networks routinely (that's 100X the rate of a 24/192 bit stream). It's just not an issue. In fact, we can now have several virtualized machines on a single server, each doing audio tasks, with no issues. Now, serving a few clients with 4K uncompressed video? That's another problem.

    Yes, ideally you'd have your server running only server software, nothing else. But that's a reliability issue, not one of sonic signature. And paying attention to CPU load management solves the problem anyway.
    Originally posted by Kingrex View Post
    The environmental condition where a server I am envisioning is utilized is as follows. It would be on a well damped audio rack. It would be fed with the same isolated circuits used by the playback preamps and amplifiers as well as DAC and other audio devices.
    While I don't think it's a problem at all, if one wanted the best, most stable power fed to that server PSU, the last circuit you'd want it on would be the one shared by power amps. Depending on the amps PSU design, they can present a significantly variable load to the line. But like I said, it's a non-issue. Every PSU of any type is highly regulated, so what goes on on the power line has been taken care of.
    Originally posted by Kingrex View Post
    In some cases it is direct connected to the DAC with a USB cable.
    Direct to DAC via USB is a very different animal from network transport.
    Originally posted by Kingrex View Post
    At other times it is either direct connected or through a network connection by ethernet. Even when the server is direct connected to the DAC through USB, there is an ethernet connection to the server reaching out to the internet and or a NAS.
    I wouldn't term a network/ethernet connection to a DAC a "direct" connection, if that term is used for a USB-connected DAC. Not much is "direct" in network communications.
    Last edited by Tech7738; 09-04-2019, 08:53 AM.

    Comment


    • #32
      Tech, I get how life work can burn out life passion if the overlap. kind of like a commercial fisherman doesn't usually want to sit on a boat and sport fish. Have you experienced that in your personal life as it relates to audio? I ask as it sounds like you have no desire to test the envelope with your home audio. Provided you even have a home audio system. Do you even have a home system? If you do, I believe you would be pleasantly surprised by your digital audio performance if you applied a few of the modification members here advocate, as well as what Ultrafast advocates on Audioshark. If you put my Mojo Audio Deja Vu server, or an Aurender or Innuos server in, I'm confident you would immediately hear the change. Call it what you want, you should hear it. Over a week or 2 you would also find the lack of distortion allowing you to play comfortably at volume levels 6 to 8 db higher than you previously use to. Reapply your original computer and a week later you will be backing down the volume. I don't profess to know why, even though I just use the word distortion, but it happens. I hope you have a home audio system and give it a try someday.
      PAP Trio 10/Voxativ & PAP Trio 15 Horn speakers, Ampsandsound Casablanca monoblocks, First Sound Audio Mark 3SI preamp,
      Mojo Audio Deja Vu server, Mojo Audio Mystique V3 DAC, The Linear Solution Ethernet Switch, Blue Jeans Ethernet cablling,
      Akiko Corelli, Custom power strip direct wired to panel with OFC copper wire. Inakustik Ref Air 2404 Speaker cable. Genesis and Inakustik NF2404 Air Interconnects.

      Comment


      • #33
        Tech, I was not trying to be insulting in any way. I did not think it was an out-of-the-ordinary question to ask if you engage in home audio playback. If I may, it appears to me you might have a little difficulty in separating your professional from personal usage of audio. You seem to always be focused on understanding.

        But I need to know why. Always. An unknown cause doesn't further the art, it only confuses it. If there was something that improved any audible aspect significantly, and the cause was known, it could be applied broadly with unfathomable benefits to millions.


        I do 100% agree bias will always influence your subjective opinion. That was partly my reasoning for stating that weeks later you find the way you interface with your equipment has changed. I believe bias his less influence with that much time. I believe something has changed which through repeated use and time shows Itself. It has altered the habits you engage in. The habit being a certain volume level you normally play your music at. When that shifts over a week or two span something more is happening than personal bias, in my opinion.

        I get your adherence to scientific methods. Yet I still believe you could perform a very valid "personal" test at your home by inserting a modestly inexpensive "Gigafoil" optical isolator right before your DAC. Assuming your dac is ethernet connected. I'm sure your microphones will pick up a change. Whether you can, IYO decide whether you like the change? I just think it would be interesting because I believed you were saying that a device such as a optical isolator should in no way altar the sound. It's still only packets of data. Yet I firmly believe your microphones will pick up a change. If indeed your microphones do pick up a change, at that time you can decide whether you want to apply a more true scientific methods to determine why that occurred.


        Tech, I don't push the idea to be confrontational in anyway. You're obviously far more skilled and knowledgeable in the field of digital audio recording and playback than most all of us on this form. I'm sure I speak for many in saying we as a group would be extremely interested in hearing the results you might quantify if you were to replace a home computer with a high-end audio grade music server or simply inserting an optical isolator. In short, it would give us actual technical data to validate our own beliefs. It would be wonderful to hear a person with your skill utilizing microphones coming back with information confirming noise levels, distortion, bass, high frequency Etc were altered. I would even say I'm interested enough that I could probably get you a demo unit to use in your trials. Your writing is of a high enough caliber your results could easily be picked up by one of the many audio magazines and published.
        PAP Trio 10/Voxativ & PAP Trio 15 Horn speakers, Ampsandsound Casablanca monoblocks, First Sound Audio Mark 3SI preamp,
        Mojo Audio Deja Vu server, Mojo Audio Mystique V3 DAC, The Linear Solution Ethernet Switch, Blue Jeans Ethernet cablling,
        Akiko Corelli, Custom power strip direct wired to panel with OFC copper wire. Inakustik Ref Air 2404 Speaker cable. Genesis and Inakustik NF2404 Air Interconnects.

        Comment


        • #34
          Originally posted by Tech7738
          Not in the least!

          I would appreciate if you could please try to control presumptions. Just because I don't share an opinion doesn't mean I'm not immersed in home audio, or that I no longer care. What it does mean is I have my opinions based on my listening, knowlege of audio, and personal and professional research. Unlike a commercial fisherman, I love audio. I'm privileged to work within a field that I also love recreationally as well. Most of the audio professionals I know are the same way, the passion transitions work and private life.
          Seriously? Of course I do. Invested thousands...more like 10s of thousands.
          If I were to do any of that, and it's not out of the realm of possibility, I would pretty much have to ignore everything you just said because it scales expectations. I need my observations to be un-scaled, and as rooted in reality as possible. So, to do any of the above, would take the construction of the two identical systems I've outlined in a previous post, then application of my ABX comparator. And lots of listening. So, don't expect results quickly here. Listening and compiling subjective observations would take a very long time.

          But I need to know why. Always. An unknown cause doesn't further the art, it only confuses it. If there was something that improved any audible aspect significantly, and the cause was known, it could be applied broadly with unfathomable benefits to millions. Ok, a bit grandiose, but please understand my point here. I'm not trying to debunk anything, I'm trying to first, determine if there is any real change caused by <fill in the component change>, and if there is, what the cause is, so that it may further understanding and the improvement can be universally applied as much as possible.

          I audition audio every day.

          If you guys would consider this:

          1. Just because someone doesn't share an opinion, they may not be wrong.
          2. Just because someone else hasn't experienced the same audible change doesn't mean they or their system is defective, or burned out, or they don't listen
          3. There are those that endeavor to further the art purposefully, but with the understanding of what is being done on a scientific level
          4. Audible changes affect the wave arriving at the ear (or they wouldn't be audible changes)
          5. Changes in the wave arriving at the ear mean the signal arriving at the transducer (speaker, headphones), must also have changed
          6. Changes can be detected and measured, perhaps not in traditional ways, but they can be detected far below audibility (see "Spectral Contamination" for one example)
          7. If we can explain has changed and why an improvement is heard, our ability to benefit the art is amplified and...
          8. The improvements can benefit others!

          I would also like to open this possibility: If a change is even slightly anticipated, and uncontrolled subjective observation is the only metric, the resulting opinion can be skewed.

          Years ago when I first got the ABX system I "calibrated" the test setup by using a null...both A and B were identical, so there could be no difference in X. Calibration produced a null, 50% result. However, if I injected a bias, a suggestion that A and B were absolutely NOT identical, and that there was some characteristic of A that made it superior to B (pick your favorite here, color, cost, quality of build, pseudo-scientific mojo), I could produce a reliably positive opinion that A was superior to B. But when X was compared, I still got a null, 50% result. "A is made with the highest grade components, including multifilar transformer windings, ultra-low ESR caps, all discrete components, and 24K gold contacts. We feel the sound is more open, airy, and has palpable dimension including height". That would always produce a preference for A. I could increase the degree of preference by adding, "90% of our testers have clearly heard the improvement". Artificial peer-pressure. And that bumped the results to nearly 100%. But in reality, A and B were identical. The results were skewed by the injection of a bias.

          When you watch a magic act what your senses tell you is in conflict with reality. If you had never seen a magic act, and didn't know it was an illusion, you would put complete trust in what your senses were telling you because they had proven untrustworthy in the past. So your expectation would be that your senses are giving you reliable information. Now, magic and illusion is a deliberate attempt to place conflict between senses and what we know is reality. Most audio tweaks are not deliberate attempts to mislead, but expectation is powerful, and there is usually a lack of knowlege...the "why" is missing, even attributed to the mystical, unknown, or an inadequacy of science.

          My interest here is to explain the "why". But to do that, I must first confirm there actually is a change apart from expectation. And to do that, I must listen apart from bias. That's not easy to do, and I wouldn't expect the typical audiophile to attempt it. But I would also ask that readers consider the possibility that if you do not understand what's going on, or perhaps are committed to a particular belief, or that because of what you've heard or read, that your expectations could at least play a small role in the opinions formed by your subjective listening and observation, particularly when the difference is small.
          ABX is a joke because it tells us what we already know. Nothing like rediscovering the wheel but any good book on motor learning and testing will tell you that short term memory is notoriously unreliable-Like police getting ten different descriptions of the robbery suspect from ten different people. The reason is simple. Short term memory has a limited amount of storage not unlike a1986 IBM computer. Look up books in motor learning by Shea or Schmidt-Wrisberg or the work of the great Nikolai Bernstein. Learning is learning to the brain whether it’s math, language or motor skills.

          Throw into the mix complexity and perception goes down the drain. That’s why concert pianists want to be as relaxed as possible when playing so they can enhance their perceptual skills; that’s opposed say to an Olympic Lifter who has one task to perform - lift as much weight as possible - and wants maximal CNS stimulation possible.

          Hence the difference between double blind clinical trials where there is simple criteria by doctor to be evaluated and audio where there are so many things going on within the context of the test, perception is lowered and the person being tested is giving feedback. And add to the limited number of people in the trials where you have to do voodoo stat to analyze. If you want to see that even 30% are positive responders in a trial, you need a whole lotta more people. I should know. I was involved in clinical cancer trials.

          So this comes from the Admin. No more ABX discussion here. Any further discussion will be deleted.

          Now let’s return to Rex’s original topic.
          Myles B. Astor, PhD, Administrator
          Senior Editor, Positive-Feedback.com
          ________________________________________

          -Magico S5 Mk.2 speakers with SPod feet
          -cj 40th Anniversary ART300 monoblock amplifiers
          -Merrill Audio Elemente 116 monoblock amplifiers
          -cj GAT preamplifier Series 2 preamplifier
          -Doshi V3.0 phonostage
          -VPI Vanquish direct-drive turntable
          -VPI 12-inch 3D Fat Boy dual pivot tonearm, VPI 12-inch 3D Fat Boy gimballed and SAT LM-12 arm
          -Lyra Atlas SL, Fuuga, vdh Colibri Master Signature, MOFI Master Tracker, Sumiko Songbird cartridges
          -Technics RS1506 with Flux Magnetic heads, Doshi V3.0 tape stage (balanced)
          -Assorted cables including Transparent XL Gen. 5, Allnic cables, Skogrand, Viero, Kubala-Sosna, MG Audio, Audience Au24SX, Genesis Advanced Technologies Power Cords
          -Accessories including Stillpoint Aperture panels, Cathedral Sound panels, Furutech NCF Nano AC receptacles; Silver Circle Tchaik 6 PLC, Symposium ISIS and SRA Craz racks, Audiodharma Cable Cooker, Symposium Ultra and assorted SRA platforms.

          Comment


          • #35
            But long-term memory may be no more reliable than short-term memory, and is in some ways worse because it can be manipulated so easily. Each time a memory is "recalled" from "storage", it is re-remembered and then re-stored as that "new" memory. And not to get too political, but it is well known (at least in some scientific circles) that repeating a false statement enough times will cause it to be remembered as true. In the same way, convincing oneself (over and over) that one hears a difference while listening (over and over) will result in believing that there was a difference, even if there wasn't (and isn't).

            Similarly, saying that "everything matters" is misleading, because how much does everything (or anything) matter? Not to pick on anyone in particular, but Kingrex's (and others') continual fine-tuning, if it is as effective as claimed, should result in a huge audible difference comparing the final, "finished" version with the starting point (presumably something like a high-performance but not "optimized" Windows 10 laptop). I've never heard that an unbiased listener has found that to be true.
            Modwright Oppo 205 full tube mod
            Pass Labs INT60
            Daedalus Audio Ulysses v.2
            REL S3
            Daedalus/Wywires and Acoustic Zen cables
            Torus IS5

            Comment


            • #36
              Rbbert, are you saying you know someone who sat in front of a $5k to $12k dac and the owner played a laptop through it, then played an Innous Statement through it and they heard no difference,

              And I don't think I have said any partucular step in itself is owe my god, has to be done. I believe my many post on many threads are that many incremental steps are what is needed to get your digital to a place where it rivals your analog. I just dont see a laptop sitting on top a speaker achieving the same satisfaction with the end result as the analog guy who sets his table on a great stand, uses excellent cables and fine tone arm wire, uses a top line cartridge and spends weeks getting it dialed in with tools, then by ear. Its not going to happen. IMO the vinyl will slaughter it.
              PAP Trio 10/Voxativ & PAP Trio 15 Horn speakers, Ampsandsound Casablanca monoblocks, First Sound Audio Mark 3SI preamp,
              Mojo Audio Deja Vu server, Mojo Audio Mystique V3 DAC, The Linear Solution Ethernet Switch, Blue Jeans Ethernet cablling,
              Akiko Corelli, Custom power strip direct wired to panel with OFC copper wire. Inakustik Ref Air 2404 Speaker cable. Genesis and Inakustik NF2404 Air Interconnects.

              Comment

              Working...
              X