Jump to content

jabbr

  • Content count

    5,431
  • Joined

  • Last visited

2 Followers

About jabbr

  • Rank
    Senior Member

Personal Information

  • Location
    Cincinnati

Recent Profile Visitors

9,645 profile views
  1. Oh c’mon this is nothing to do with computer audio. The person here who has actual experience with space system design is @marce so it’s funny that you are arguing with him. I’m just saying that commercial computer equipment hardly has a high error rate even with trillions of calculations. Yes you want low noise. So called “high end” audio designs use commercial off the shelf parts (digital) although Pass has commissioned custom transistors but regarding digital audio, and specifically the chain before the DAC none of the parts need to be “special” Real high end electronic design may use $80 - 100,000 software packages and the fact that audio companies can’t even afford mid-range measurement equipment suggests to me that practically none of the board design is done with the resources available to eg SuperMicro and certainly not that of Intel or Apple. If I’m wrong let’s see the data
  2. Uh no. The trend end these days is to put more function on an integrated CPU called a SoC ... ClearFog is an example that I often use as a specialized network I/O board... There is no one design for “audio” this concept is frankly absurd. The requirements for a streamer (eg ClearFog with essentially one chip) are vastly different from what is needed to run HQPlayer upsampling ... in my experience the video ouput has the most noise, but what about folks who want integrated video + audio ... similarly my NAS has terabytes of storage ... why do you think there are so many motherboards all made by SuperMicro as well as ASUS and then NVidia/AMD etc ... each have different options. Design the system that has the options you need. All off the shelf.
  3. Whaaaaat? I can tell you with certainty that both NASA and our national research laboratories (think nuclear weapons research etc) have been pioneers in the use of cheap of the shelf computers clustered together https://www.hq.nasa.gov/hpcc/reports/annrpt97/accomps/ess/WW46.html The concept that these cheap computers are designed to allow errors is not even wrong. Do you think NASA and the military have their own foundaries? Design their own CPUs ? ... maybe but just like a CPU becomes specialized with software, so are FPGAs.
  4. jabbr

    What is this jitter of which you speak?

    Ok, the 24 bits clearly have the overhead, and if the technique is fractional sampling, that does the antialiasing ... will handle at least a good class of PCM examples. DSD is different if for only the reason that there isn't a file that captures multibit DSD ... you aren't claiming to be analyzing DSD so that's fine...
  5. jabbr

    A proposal: the Objectivist Audio Review magazine

    I'd like to keep the hobby inclusive and welcome in a younger generation -- I try to give affordable recommendations for that reason -- sometimes (but certainly not all times!) there can be a certain snobbery whereby a very expensive piece of wire seems like the defining mark of a certain group of audiophiles -- I think that such jewelry does get the enthusiastic "SQ" mark from its proponents, but doesn't tend to be marketed with objective data. In that vein its good to have opposing voices so that folks who don't have excess $$$ for bling don't feel left out. @Teresa is a great example of someone who states she is on a budget -- she is clearly capable of making sensible decisions for herself but I wonder how many "lurkers" are looking for practical advice. No not bothered but willing to put up an alternate viewpoint That said, I'm personally a sucker for a rare, hard to get, no longer made chunk of silicon, or silicon carbide, or an exotic ...
  6. jabbr

    What is this jitter of which you speak?

    Round and round. Are you are resampling a 96kHz digital signal using random fractional offsets and storing it into a 96kHz file???
  7. jabbr

    Materials that absorb RF energy

    The first volume I quoted above discusses issues with CMOS logic families that result in RF/EMI and intrinsic jitter
  8. jabbr

    Materials that absorb RF energy

    Henry Ott: 1) https://www.amazon.com/Noise-Reduction-Techniques-Electronic-Systems/dp/0471850683/ref=pd_lpo_sbs_14_t_0?_encoding=UTF8&psc=1&refRID=W0612J5DE0Y2B3SW4YM2 2) https://www.amazon.com/Electromagnetic-Compatibility-Engineering-Henry-Ott/dp/0470189304 Both of these are invaluable resources for those folks interested in RF/EMI mitigation. If you pay yourself minimum wage for your time and want to avoid months and months of work get both of these books ... the second one is aimed more at cables, enclosures, connectors, transformers, and the first more at electronic components. @lmitche, @austinpop, @Cornan I know you guys (among others) are doing a lot of work with shields over cables ... the second book in particular goes into great great depth about cables, shields, geometry of connectors etc ... don't want to spoil the real fun of empirically verifying this for yourselves but if you've got the bug, this textbook will explain why what you are doing may be working ...
  9. jabbr

    What is this jitter of which you speak?

    Perhaps avoid "extract jitter from a file". Jitter is not in a file (now the file might contain a mathematical description of jitter i.e. a convolution/transform etc..) but the concept that jitter is embedded in the bits of a file is very problematic. This is my question: A digital bitstream is has samples that are intended to be clocked with a sample rate. At each sample there is a scalar value (e.g. PCM) as well as a phase error which is an analog value (yes @PeterSt we are treating the bitstream properly as an analog signal). scalar value is one axis phase error(jitter) is another axis So what you are proposing is that there is software which "embeds jitter" into a file ... ??? Where has this been validated? You are flattening phase info into scalar info -- obviously throwing out an entire dimension -- and you know that this is accurate? how does this work mathematically? I've suggested that by oversampling 10x you could "fit" the extra phase information alongside the scalar info -- yes treat the signal as analog (and its not strictly PCM encoded anymore)... but remember: jitter is a time based value, you cannot directly embed it in a file (unless you have a specific transform that "digitizes" the phase info) so what is this transform? how have you validated the transform? No problem if just fooling around but don't draw conclusions that "jitter sounds like "x".
  10. jabbr

    What is this jitter of which you speak?

    That's me... have you been talking with my wife? I do have an old HP 3048A phase noise measurement system: http://hpmemoryproject.org/technics/bench/3048/bench_3048_home.htm There is a reason that I could afford to get it, very cumbersome to use... That said the HP 3561a spectrum analyzer is special ... does sub Hz ... that's 640 microHz intervals ... yep! But there is a reason that its obsolete ... very cumbersome to use ... to the point where I had to find and rebuilt a PC with a certain Intel processor (had to be slow) to run the old HP software etc... There's a group called "time-nuts" which is into this sort of thing ... "Your doing this for audio??? why would they care about accurate phase error measurements???" Nope, no time for that. It real work to do and I've got too many unfinished projects in my basement ... I don't get anything out of reviewing other peoples stuff. Got my own stuff to work on...
  11. I’ve thrown Christmas toys out in the snow because I couldn’t find the “off” button... worth the grief
  12. http://www.diyaudio.com/forums/pass-labs/281520-official-m2-schematic.html EDCOR PC600/15K http://www.firstwatt.com/pdf/prod_m2_man.pdf
  13. jabbr

    What is this jitter of which you speak?

    Let's use the DSD512 example ... the clock is 22 Mhz. Let's assume a high level of 1 Hz offset jitter, e.g. 1/4 (22Mhz-1Hz), 1/2 22Mhz, 1/4 (22Mhz+1Hz) ... how does the software cause this and what effect on the signal do you predict? I'm using a very high (-3 dB) level of 1Hz offset jitter to reduce issues with quantization.
  14. jabbr

    What is this jitter of which you speak?

    Ok. Maybe I misunderstood. No doubt there is a category of jitter for which this would work. What testable predictions does your model make? For example (throwing this out): suppose you add enough jitter that you can hear a very clear difference. Now substitute a clock that has this level of jitter: does the audible difference between the constructed file played with a good clock, and the original file played with the jittery clock go away?
  15. jabbr

    A proposal: the Objectivist Audio Review magazine

    Ok I think I understand what you are getting at. Personally trying to "prove" that someone has a sensory experience is very hard. You could start sticking electrodes into monkey brains but that's really difficult and your question needs to be important. There's a lot of stuff that I put into an intermediate category where there isn't enough information to decide one way or the other. Each person should consider what they'd like to see to be convinced that their hypothesis is wrong: for example you hear a dramatic difference with a particular brand of shield ... what will convince you that this is an illusion? If you answer that you just don't care then you aren't being scientific nor objective. Conversely you may have a hypothesis that cables don't matter: what information will convince you that you are wrong? Everyone needs to decide how high their own bar for belief is. For me I don't want to throw out false negatives that have a reasonable chance of being real, nor do I want to chase down too many false positivies ... so I maintain a broad category where I'm skeptical but willing to listen to/view/read new information. For me anecdotes don't sway me one way or the other regardless of their number. Generally I like to see two independent confirmations: let's say two entirely different types of measurement, or a measurement and a controlled listening test.
×