Jump to content


  • Content Count

  • Joined

  • Last visited

About lmitche

  • Rank
    Senior Member

Personal Information

  • Location

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. I believe you can add a second regulation stage at a later time.
  2. John, there are indeed controls in Audiolinux to do what you want. I used them in my builds for many years until recent versions of Roon. SQ significantly improved without any CPU pinning or dedicated cores in the January release. My guess is that Roon took control of resource allocation in an improved way, and AL just sounds best without constraints. While I no longer use hqplayer, we do know that Jussi manages resources as well. Given all this, I'd recommend staying away for these controls. There are plenty of other tweaks that you can explore that will raise sq with Audiolinux.
  3. Hi Alexey, can you please repost your pictures from February 2018. I can't find them.
  4. According to your screen shot you are running your memory below 2666MHz. You should enable the DOCP choice to pull the memory timings from the SPD. In a modern Asus bios it is usually on the first pulldown on the "Tweaker" page.
  5. Look in your bios for Vt-x or SVM, on Intel or AMD respectively, and disable it.
  6. Nenon, Have you tried the Monoprice Slimrun Fiber USB 3 extension cable found by Roy? Larry
  7. Fans make both vibration and ambient noise, which can often be tolerable, but worse make electrical noise even with inline fan filters.
  8. John, many thanks for the quick response. Having done many experiments, one detects patterns and your paper is consistent with observations made here. I am grateful for your efforts, and I look forward to learning more.
  9. Hi guys, Here is dumb question number one and two! The paper states that the clock timing is the same regardless of "noise". Understanding that, if the state change threshold is at 2 volts and the ground pane noise pushes the threshold voltage to 2.1 volts, won't a 2 volt "bit" go undetected? Likewise, assuming the noise can also lower the threshold, to say 1.9 volts, could not "bits" be detected that don't exist? Just curious. Larry
  10. Sounds like it's time to dust off the Adnaco. Maybe even roll a few SFPs.
  11. I second that suggestion. Even better try adding another (4th) layer of shielding with the ends tied to the other three. Yesterday a MuMetal braid was added to my .4 meter Lush 2 cable. It was initially configured as two jssg360s. That was an improvement, but something was missing. Tying all 4 ends together fixed that, and the results is more image weight and clarity across the spectrum.
  12. Hi Bricki, Another option is to use the two bridged USB NICs on the server, one to the ER, the other to the NUC, with direct to cpu ports if possible. (A multi-port USB card like the 2 channel 4 port Startech externally powered with 12 volts and installed in a direct to cpu pcie slot works well in Intel machines. On AMD motherboards there are often several direct to cpu USB 3 slots on board already). TheEthernet port in the i7 NUC sounds better than an USB NIC here, so you only need two server side USB NICs not three. Be sure to disable Energy Efficient Ethernet (EEE) on the server to NIC connection with ethtool. Once the latter is done, give 1gbps another go.
  • Create New...