Jump to content

jabbr

  • Content Count

    7852
  • Joined

  • Last visited

  • Country

    United States

Reputation Activity

  1. Thanks
    jabbr got a reaction from R1200CL in Diretta audio protocol   
    Err ... sorry this isn't how modern NICs work. My Mellanox ConnectX-4 and 5 NICs ***most certainly*** don't generate a CPU interrupt for each packet.
    In ancient days, perhaps, but modern networks just don't do this and the topic is far too complex for a robust discussion here ... see RDMA, ROCE
    Also look at how an FPGA processes Ethernet.
     
    A lot of these things is frankly why I'm hesitant to look at Diretta without access to a highly technical whitepaper. It's just handwaving to me otherwise.
  2. Upvote
    jabbr reacted to Miska in Low cost high quality excellent performing LPSU for audio.   
    It is not so clear.
     
    Medical grade PSUs, like the one I referred to, have been specifically designed to have very low leakage currents, in µA range.
     
    Even with standard SMPS wall-warts, it is entirely possible to design very low noise analog devices. I have measured enough many DACs running from SMPS that I know it is more about the design than the type of PSU. Overall, 60/200 kHz is much easier to filter out than 50/60/100/120 Hz. Especially since the fundamental and it's harmonics are outside of audio band.
     
  3. Upvote
    jabbr got a reaction from Duke40 in GROUNDING   
    Points all well taken, but furthermore consider the other benefits of a good isolating transformer ... and I use the beautiful Equi-Tech Q balanced isolating transformer... the readers are left to consider what the ground impedance to the main power box is, in fact. Furthermore given your balanced audio circuit, what do we expect the ground current to be exactly? ZERO!!!
  4. Upvote
    jabbr got a reaction from Solstice380 in AES and Its Papers   
    This is a well known "trick" when giving a lecture at a conference, always write up the talk as a paper because when it is published, it will get accepted without much if any real peer review. Don't believe every scientific paper as the gospel truth. If 10% of the papers published in Science and Nature every week were breakthroughs then humanity would be a lot collectively smarter than we are today.
     
    The bottom line is that you need to know the biases of the author when reading anything including a published scientific paper. We need independent journalists to keep those biases in check by exposing them.
  5. Upvote
    jabbr got a reaction from sandyk in AES and Its Papers   
    This is a well known "trick" when giving a lecture at a conference, always write up the talk as a paper because when it is published, it will get accepted without much if any real peer review. Don't believe every scientific paper as the gospel truth. If 10% of the papers published in Science and Nature every week were breakthroughs then humanity would be a lot collectively smarter than we are today.
     
    The bottom line is that you need to know the biases of the author when reading anything including a published scientific paper. We need independent journalists to keep those biases in check by exposing them.
  6. Upvote
    jabbr got a reaction from botrytis in AES and Its Papers   
    This is a well known "trick" when giving a lecture at a conference, always write up the talk as a paper because when it is published, it will get accepted without much if any real peer review. Don't believe every scientific paper as the gospel truth. If 10% of the papers published in Science and Nature every week were breakthroughs then humanity would be a lot collectively smarter than we are today.
     
    The bottom line is that you need to know the biases of the author when reading anything including a published scientific paper. We need independent journalists to keep those biases in check by exposing them.
  7. Funny
    jabbr got a reaction from AudioDoctor in GROUNDING   
    apropos leakage currents ... I mean leaking discussions, the comments make their way into my feed despite blocking ... the AS blocking needs to be higher impedance!
  8. Funny
    jabbr got a reaction from barrows in GROUNDING   
    apropos leakage currents ... I mean leaking discussions, the comments make their way into my feed despite blocking ... the AS blocking needs to be higher impedance!
  9. Upvote
    jabbr got a reaction from barrows in GROUNDING   
    Points all well taken, but furthermore consider the other benefits of a good isolating transformer ... and I use the beautiful Equi-Tech Q balanced isolating transformer... the readers are left to consider what the ground impedance to the main power box is, in fact. Furthermore given your balanced audio circuit, what do we expect the ground current to be exactly? ZERO!!!
  10. Funny
    jabbr reacted to AudioDoctor in GROUNDING   
    @barrows There is a reason I blocked Frank...
  11. Upvote
    jabbr reacted to barrows in GROUNDING   
    OMG, I just came across another review of a very expensive "Audiophile Grounding Box", and there was so much mis-information and outright lies (in an interview with the designer!) that it just about made my blood boil.
    A quick few words on "grounding":
     
    1.  If any audiophile is concerned about grounding, the very best thing one can do is use true balanced/differential components.  There is no reason to use single ended (RCA) components in a high end audio system anymore, as there are numerous choices one can purchase to avoid single ended circuitry.  And please, do not give me that old trope about how most components are not truly balanced internally even though they have XLR connectors, this is just not true in so many cases as to be silly.  Some components which offer true balanced circuitry:  Hypex, Purifi, Benchmark, PS Audio, Bricasti Design, Ayre, Aesthetix, Chord, Constellation, etc, etc, etc.  The list is long!  Additionally, contemporary DAC chips and phono cartridges both have balanced outputs standard.  With balanced signaling there is no audio circuit reference to 0 volts (or ground if you prefer)-in other words, the signals do not care what is on ground at all.  Balanced signals only reference the voltage difference between the two lines, neither of which is at ground.  So there's no ground "issue" with properly implemented balanced signaling.
     
    2.  If you still insist on using RCA connected gears, stop reading now, you are on your own, and please feel free to purchase as many overly priced audiophile grounding accessories as you might choose to.  But your money would be better spent on balanced components in the first place.
     
    3.  Use properly designed XLR cabling: a proper XLR cable has signal conductors for the pin 2 and 3 connections, surrounded by a shield which is connected at both ends to pin 1.  Nothing else should be connected to pin 1.  Some audiophile cable companies do not do this correctly, so check.  There should not be any wire connected to pin 1, pin 1 is shield.  Inside your components the XLR shell, and pin 1 should be connected directly to the chassis, this creates a continuous shield of the chassis-cable shield-and the subsequent chassis.
     
    4.  If one pays attention to the above things, there is likely to be no "grounding issue" in the system.  But what about internally, and noise on ground planes, perhaps, especially in mixed signal components (specifically, DACs).  This depends highly on the skill of the DAC designer, and how they handle their signal routing, PCB design, etc.  Suffice it to say, that in a well designed DAC, done by a competent mixed signal designer, implemented in a balanced system context, there will be no issues.  Although some audiophile DACs appear to be designed by analog based engineers who may not have the requisite high speed signaling experience to implement proper PCB layout and grounding to avoid problems.  If your DAC responds to grounding tweaks, be very suspicious of the skills of its design.  Perhaps, in that case, take a look at the low level noise floor of the DAC in published measurements and see if it shows problems, and if it does, consider replacement with a better implementation by a competent mixed signal designer.
     
    5.  If you play vinyl, insist on a phono stage (and turntable wiring) which can take advantage of the balanced output of phono cartridges.  It is absurd that phono stages are not always balanced, as with their high gain they likely have more to gain from balanced connection than any other component.
     
    And finally, if you still insist that audiophile grounding boxes make a difference in your system, go back and figure out where the problem is instead of applying a very expensive "audiophile bandaid".
  12. Thanks
    jabbr got a reaction from sledwards in HQPlayer Linux Desktop and HQplayer embedded   
    Aha,
     
    The highest driver level reported in
     
    $ ubuntu-drivers devices
     
    is nvidia-driver-455
     
    Its possible to add 
    sudo add-apt-repository ppa:graphics-drivers/ppa  
    and get the experimental/beta drivers
    alternatively the driver can be downloaded from the nvidia site and installed.
     
    (just so folks know)
     
  13. Upvote
    jabbr got a reaction from blue2 in Subjective comparison of Software Music Player   
    HQPlayer is more of a system than simply a player per se. It would be a project to fully employ its capabilities:
    1) filters and modulators which run on a high powered PC and output to either a local device (ALSA) or a network device (NAA)
    2) loadable convolution kernels usable for room correction, digital crossovers as well as arbitrary convolutions
    3) stream input either HTTP or NAA.
    4) CUDA offload
    5) NAA remote/network ALSA protocol including OS distribution (intended for low power devices)
     
     
    Presumably a NAA (networkaudiod.xxx.deb) package could run on wtfplay OS.
    Aside from @Miska's custom linux distributions, the scheduling on an NAA player device is determined by the underlying Linux OS. I have run NAA on custom compiled/patched linux kernels.
     
     
    That is an entire area of investigation and to one extent convolution kernels can be deployed and tested. I think what is unique to HQPlayer in this area is that the kernels are running or can be running in SDM space, not PCM, which is a really unique capability!
     
    Custom filters and non-linear transforms that cannot be expressed as a convolution kernel are not, to my understanding, deployable with HQP.
     
     
    It would be really terrific to see further work incorporated into SOX. I understand that Jussi has done a tremendous amount of work and has developed the NAA protocol for his purposes, but I see that as the best remote ALSA implementation available. If an open source lightweight remote ALSA protocol were developed that would be terrific, and then, for example, you could use SOX or HQplayer on a workstation, and then send the audio to a low powered (or underclocked) audio machine, running wtfplay and then you could compare wtfplay against NAA which would be a great comparison!
  14. Upvote
    jabbr got a reaction from R1200CL in Help creating Ethernet and fiber Optic network for Innuos   
    I have installed LC-LC Single mode fiber in my walls.
    The reason: Single mode fiber works with all speeds 1G,10G,25G,100G also the fiber is cheap.
  15. Thanks
    jabbr got a reaction from Duke40 in Optical Network Configurations   
    I received the Mikrotik CRS305 yesterday. $130 shipped on Amazon, 4 SFP(+) ports ... yes sports fans, that 4 10Gb Ethernet ports at an unheard of crazy cheap price.
     
    Why oh why would an audiophile need 10 Gb Ethernet? ... I mean audio can fit into 100 Mb right? Well it’s the standards folks that essentially require a femtosecond clock and extensive testing on these puppies. That’s why they have been so expensive until now. 
     
    The IEEE 802.3ae-2002 standard specified 10Gb Ethernet. Clause 52 specified for the first time, a really tight eye pattern that severely limits jitter on the signal. Prior Ethernet standards (100m, 1G) did not limit jitter in this way, so who knows what you get in a cheap 1G switch, but all conformant 10G devices must meet this low jitter standard. 
    Testing:
    1) https://www.keysight.com/upload/cmc_upload/All/669-10GigEthernet-08-01-02.pdf
    2) http://www.ieee802.org/3/10GMMFSG/public/jan04/lindsay_1_0104.pdf
    3) https://download.tek.com/document/85W_16988_0.pdf
     
    You see that the 10GBase-X standard requires an entirely different level of jitter control. That’s one reason why 10G devices have been so expensive. Yet femtosecond clocks are a dime a dozen these days, particularly for the big network boys that purchase by the crate of rolls.
     
    Let’s take a look at the setup:
     
    Unboxing
     



     
    Couldn’t be easier. It’s silent. It’s tiny. It’s low power. You plug a 12V DC supply in the back. Plug in SFP modules and you are good to go. No need to configure out if the box.
     

     
    You can use singlemode or multi mode 1G SFP modules ... you can also use 1G RG-45 modules with copper Ethernet. 
     

     
    Here is the device configured as an “FMC” ie fiber media converter. 
     
    How does it sound? Completely silent. Dark black silent. The only thing you can hear is the music. $130
     
  16. Upvote
    jabbr got a reaction from R1200CL in Best way to get a fibre connection to SFN8522   
    Just use SFP modules on both sides, as SFP(+) NICa are generally compatible with SFP modules ... ore use a dual 10g/1g SFP(+) module e.g. Finisar, just be sure they are both either SX (multimode) or LX (singlemode) and you use either a multimode or singlemode cable.
  17. Upvote
    jabbr got a reaction from Gavin1977 in Best way to get a fibre connection to SFN8522   
    Just use SFP modules on both sides, as SFP(+) NICa are generally compatible with SFP modules ... ore use a dual 10g/1g SFP(+) module e.g. Finisar, just be sure they are both either SX (multimode) or LX (singlemode) and you use either a multimode or singlemode cable.
  18. Upvote
    jabbr reacted to Miska in HOLO Audio MAY DAC   
    Problem with all those I2S interfaces is that clocking is wrong way. From the I2S source, while it should be the opposite direction, from DAC towards the source...
     
  19. Like
    jabbr reacted to Miska in HOLO Audio Spring DAC - R2R DSD512   
    Yes, working fine here on my HP Z4 G4 (Xeon W-2245, 64 GB of ECC RAM) workstation using it's on-board USB ports:

  20. Upvote
    jabbr got a reaction from motberg in Where should I try better ethernet cables first?   
    True but the question is "where to try the cables first"
     
    For your first time, I can't think of a better place than Haleakala at sunrise.
  21. Upvote
    jabbr got a reaction from motberg in Network Playback, Buffers, Continuous Stream, etc...   
    I use HQPlayer. @Miska has programmed buffering as well as a network protocol on top of Ethernet. There is also buffering at the Ethernet layer. HQPlayer does not use file transfer, rather a stream of bits. Other protocols might transfer files, and buffer the entire file at the endpoint. Different protocols will show different patterns of network utilization. 
     
    Moreover, different NICs plugged into different motherboards/CPUs with different drivers will use different patterns of buffering. For example DirectIO can transfer network bits directly into the CPU cache. 
  22. Like
    jabbr reacted to mav52 in Where should I try better ethernet cables first?   
    The wife and I go every year to Maui, and Covid sure screwed that up for us this year.,  We and 2 other couples bought a home in Haiku on the north shore not to far from Jaws. Haleakala at sunrise a must for a newbie. Pack a darn warm jacket. At the top, temperatures during dark hours can be 30 degrees less than sea level.
  23. Upvote
    jabbr got a reaction from mav52 in Where should I try better ethernet cables first?   
    True but the question is "where to try the cables first"
     
    For your first time, I can't think of a better place than Haleakala at sunrise.
  24. Upvote
    jabbr reacted to pkane2001 in Archimago on Greene vs Harley   
    Eyes provide a very strong input into the audio perception of position. Ears have a very coarse sense of direction, and we usually use them to help look in the generally right direction to pinpoint the source of sound. Once sighted, the ‘audible’ position becomes much more precise in our mind. Without sighted position, soundstage and instrument locations are much more vague.
  25. Upvote
    jabbr got a reaction from Confused in Archimago on Greene vs Harley   
    There is a pervasive belief among certain audiophile “factions” that measurements do not explain how certain components such as cables “sound” and that measurements never will.
     
    So then we get into DBC trials etc etc etc .., we have all heard these perma debates.
     
    In an attempt to bridge the gap it has been suggested that fMRI might show in an objective fashion that Person A is actually perceiving a difference between Component X and Component Y. That fMRI can show the differences in audio perception at the audio cortex by  measuring audio cortical brain activity. 
     
    Yet it turns out that when you *imagine* that you are hearing something, that the same cortical centers light up as when you hear something with your ears — well assuming that the above referenced study and line of research were replicated for audio, as opposed to touch.
×
×
  • Create New...