Jump to content

barrows

  • Posts

    6663
  • Joined

  • Last visited

  • Country

    United States

Retained

  • Member Title
    Senior Member

Personal Information

  • Location
    High above Boulder

Recent Profile Visitors

19632 profile views
  1. My understanding was that the signal chain was: May-Integrated amplifier. So there is not the option of lowering the level in the analog domain before the signal hits the input stage of the (integrated) amplifier. Worrying about problems caused by a few dB of attenuation in the digital domain is a non-issue (audiophile myth)-note that virtually all recordings produced in the last 20 years are mixed in the digital domain, and the levels are adjusted a lot more than a few dB. Digital volume control operating at 24 or 32 bits when does well at all is transparent. Also note that analog volume control is not transparent, as the resistance adds both noise and distortion.
  2. The above is not correct in the case where the output voltage of the DAC in question is above the level at which the integrated amplifier's input stage overloads. For example, say one has a DAC with a peak output level at 0 dB of 6 V, and the integrated amplifier's input stage maxes out at 4 V, when the voltage of the signal from the DAC goes much over 4 V, the input stage of the amplifier overloads, creating distortion.
  3. I would suggest that better performance may be available from a third party Renderer anyway. For example the Sonore opticalRendu, and Signature Rendu SEoptical, both have the advantage of isolating the Network noise via their optical connections, and can send DSD 512 (1024 MAY be possible as well, but has not been tested) to the new T+A DAC. When using HQPlayer my preference is almost always for DSD 256 to take advantage of the amazing ASDM7EC modulator, but I understand some are enamored with the higher rates. In any case, pick the Renderer of your choice, as far as "synergy" goes, that would apply more in the analog domain than in the pure digital domain. BTW, I think this new DAC from T+A deserves its own thread, maybe someone who has one should start a dedicated thread for it, I personally find it very interesting with its nice DSD converter and discrete HV output stage, and is in what I feel is the sweet spot for DACs price wise.
  4. SFP+ transceivers will not work with SFP cages. You need SFP modules for 1GBe SFP cages. If you want to go with single mode fiber, try these:
  5. Here's the consideration: Define "many", and how many of those determined the results in a statistically relevant blinded comparison? One additional consideration: what measured parameter of performance has been determined? I am highly skeptical of such changes making an actual improvement. These kinds of claims are extraordinary, in that there is no technical reason yet described or even speculated for how linking these asynchronous clocking elements to single clock source could make any improvement, hence I would suggest that extraordinary evidence is necessary to actually support any real improvement. I do not believe such an improvement is entirely impossible, just very unlikely.
  6. Chris, I would suspect also that if they do, they use a sine wave clock signal, square wave clock signals are subject to much degradation of cables and connectors due to the "infinite bandwidth" requirement of passing a square wave. External clocks make the most sense for times when multiple devices must be synchronized to the same clock (as is often the case in pro audio recording applications): in Ethernet applications this is not necessary, as each Ethernet interface operates asynchronously with its own clock domain. At the absolute outer reaches of clock accuracy, especially when longer term accuracy is necessary (not an advantage for audio), external clocks can have a small advantage as they can be well shielded in a separate enclosure away from other electronics-of course inside a DAC or Renderer they can also be separately shielded and powered to achieve the same gains, although few manufacturers do this because of the expense (EMM/Meitner clocks appear to be in a shielded sub enclosure, for example).
  7. That is only your opinion, there is no actual evidence (meaning confirmed measured improvement) to support such at this time. Although I am open to the possibility, I also am quite skeptical that any external clock would provide a benefit beyond the already "Femto" level clock which is internal in the ER. Adding clocking externally generally comes with compromises: the cable, and the additional connections add additional phase noise, especially to the square wave clocks used in these circuits. In order to eclipse the performance of an an already "Femto" level internal clock, the external clock would have to be many times better than the internal one, to overcome the additional noise added by the cabling and two connectors. And even then, it is certainly unclear at what level of phase noise there are no further improvements to be gained-this is really the real question here. My original statement in this thread RE the internal clocks in opticalRendu and Signature Rendu SEoptical was meant to be comparing them to the clocks present in typical commercial 10Gbe switches/NICs etc, as that was what the discussion was referring to. A switch such as the Microtik referenced in this thread does not contain a clock with nearly the performance as what is internal to the oR and Signature Rendu SEoptical. Although, to settle this matter for myself as to applying a better (Ethernet) clock to the Signature Rendu SEoptical for my own satisfaction, I am engaged in a project where I will be adding a state of the art external clock with a sine wave output (sine wave outputs do not accumulate significantly more phase noise when running over cables and connections, unlike square wave clock signals) as an experiment. A Sine to square converter will be added right at the point where the clock signal enters the circuit, to avoid any additional phase noise. The clock I will be using has been measured and confirmed to offer state of the art performance, including the sine to square converter. I expect to hear no difference, and I will be very surprised is there is any improvement, as the internal clock is already very good, and I suspect entirely adequate to achieve best possible performance. But I will do the experiment anyway, and see if there is any audible difference. I do to know when this project will be completed, as I am very busy, and the project is quite complex.
  8. Just to note: The Signature Rendu SEoptical, opticalRendu, and opticalModule all use very good (Femto level) onboard clock oscillators with dedicated ultra clean power supplies. It would be very difficult (nearly impossible) to improve upon the performance of their onboard clocks by adding an external clock, especially considering the added clock phase noise introduced by the additional connectors and cabling.
  9. Note to anyone doing these kinds of tests: many DACs mute their outputs when no signal is present, so one cannot test the noise floor accurately with no signal. Make sure to be giving the DAC a signal to be sure the output is not muted. Of course if you are going to be testing this at full gain of your system, be absolutely sure that your signal is digital silence!
  10. Superior measured performance, measurements by Jussi. Less artifacts via DSD 256 playback vs. any rate of PCM.
  11. Agreed with the first point, on the second, well the subjective POV would be system dependent. Jussi has shown that DSD does perform better than PCM on Holo DACs technically via measurements. My own subjective point of view is that DSD on a good set up sounds more like music, and PCM sounds more hyped up, or edge enhanced if you will, of course different systems may suit one approach over the other, but the superior measures of DSD suggest that on a neutral/transparent system DSD should be the more accurate to the source playback. Again agreed on the DAC-I have been waiting for someone to make available a really good commercial DSC-2 style DAC variation for DSD only, optimized for use with HQPlayer oversampling to DSD. A simple DAC like this, with close attention placed to all the details of power supplies, clocking, output stage, and a single well implemented USB input-optimized for DSD 256 (considering how superior many find the EC modulators) woudl be a niche product, but for some folks would be an excellent option.
  12. My point was that Holo Audio DACs perform best with DSD, not sure what you are saying about PCM. Point being it is better to (assuming one is using HQPlayer with optimum settings) to convert all PCM to DSD before conversion, this is what I was referring to: not saying anything about PCM rates at all.
  13. Sonore Rendu products will do up to PCM 768 and DSD 512 confirmed, and may be able to do PCM 1.5x MHz and DSD 1024, but this has not been tested (very few DACs support these rates). Also note that is is unlikely that any other source can deliver a cleaner USB output than the Signature Rendu SEoptical. @Miska has noted that Holo Audio DACs perform "best", as I recall, with DSD 256 input, and HQPlayer achieves very high performance with DSD 256 via the EC modulators (higher rates of DSD are not possible with the EC level modulators).
  14. For cost no object designs, you guys might consider the connectors from Lemo in Switzerland. These are very high in quality and with low contact resistance, and are available in many different pin configurations and sizes.
  15. Right, it is way better (IMO, DSD is a superior format to any kind of PCM, generally, it is worth it (again IMO) for example to convert all data rates to DSD via HQPlayer before conversion)... Point being is that the Bricasti M21 and M3 DACs use a discrete, single bit, converter for DSD with no DSP of any kind applied to the incoming data stream. Hence it is NOS for DSD input. For example, if one just inputs a DSD recording at any rate from DSD 64 to DSD 256, then it is NOS, the recording is converted at its native data rate with no manipulation. It is a separate concern from the DAC if one chooses to apply oversampling in software. Of course that is not the same as PCM, and I never suggested that. The Bricasti M21 has a couple of R2R chips for PCM conversion, but they do use the proprietary oversampling running in the SHARC DSP of the DAC, and there is no way to run this R2R DAC as NOS, as there is no bypass of the OSF available.
×
×
  • Create New...