Jump to content

Nenon

Members
  • Content Count

    483
  • Joined

  • Last visited

9 Followers

About Nenon

Recent Profile Visitors

1438 profile views
  1. I was checking something on his web site yesterday and saw them showing as in stock. I see other people commented on this already. Marcin's post was about the 2666MHz RAM. That was the "new" RAM he referred to. I don't really know but the ECC RAM sounded better as non-ECC on the last build I posted about. Perhaps it's not because of the ECC bit but the quality of the memory. I compared a 2666MHz Non-ECC Apacer with the same version of ECC 2666MHz Apacer running as non-ECC. I liked the ECC RAM running as non-ECC better. There is another member on this forum performing the exact same tests (2666MHz Non-ECC Apacer vs. 2666MHz ECC Apacer running as non-ECC) but may take him some time to report.
  2. It was in stock yesterday. Email them, maybe they have some. Not from my experience. The 2666MHz Apacer RAM running at 2666MHz sounds better than the 2400MHz Apacer running at 2400MHz or any slower speed. Also, the 2400MHz Apacer RAM sounded better than my 3200MHz G.Skill. The most important property for RAM to sound good is its quality. It's not the only factor but the most important one. It also matters how you tweak your server. If you are using a low-powered CPU, slowing down the RAM helps. I am not talking about those types of servers. Gave up on them a while ago but I do remember that slowing down the RAM helped there. My comment was about @StreamFidelity build, which is not aimed at low powered CPUs. With those servers, typically (given equal quality) the higher the speed, the better it sounds. And by better, I don't mean "technical sound".
  3. Let me be the first one to comment. Happy to see more builds documented, and this one looks promising already. Please see my comments below. Have you considered using Apacer RAM? I think you will like it. I am pretty much done trying consumer memory. No matter what speed or latency they have, the Apacer always sounded better in my system. My latest discovery was that you can use ECC RAM on those LGA 1151 motherboards. It works as non-ECC RAM but sounds a little better. You may want to hold off to see Marcin's new reference USB card. If your budget allows, try the HDPLEX 800W DC-ATX. It fits nicely in the chassis with the ASUS ROG motherboard: But you will have to drill holes on the bottom of the case. And yes, I will have a super short ATX cable, that's why it's lined up like this . I ordered some Jaeger connectors and a thick carbon fiber plate. Planning to replace the stock small power connector plate that comes with the case with a custom carbon fiber plate and two Jaeger connectors on it. I will post more about it when I make it. I think it will look really good. And those connectors are top notch. This is interesting. I may ping you on PM for more details.
  4. Really cool stuff Michael! Thank you for sharing. The biggest advantage of your build compared to the build I did as part of this post is that you have a full size ATX chassis that you can easily use to try out different motherboard CPU combinations. And I don't see you taking advantage of that ;). In the build I made, I optimized the power supply for shortest possible cables, lower output impedance, heat distribution, etc. But the disadvantage was that it is a static build and you can't do much once it's done. If I wanted to change the motherboard and the ATX connector on the new board is even an inch further, the cables would not reach... Your build looks like it was made for experiments with different motherboards. And I highly encourage you to do that, especially since you use HQPlayer and need more powerful CPUs. Sean typically oberbuilds his 5A power supplies and they can easily handle 10A non-continuous power. I can tell you from experience that his 5A LPS easily feeds an Intel i9 9900k without HQplayer. I don't know what will happen if you add the extra HQplayer processing. But maybe you can try an Intel i9 9900k or an AMD Ryzen 7 3700x without HQPlayer. You might be surprised how good a one-box Euphony sounds with those two CPUs (make sure you set CPU Isolations, ramroot, cache, and all the good stuff in Euphony). Speaking of Euphony, I noticed something really interesting a few weeks ago. My network is highly optimized now, Qobuz can sound better than local files if the recording is better, and my system became extremely transparent to any network changes. I noticed a difference in the sound quality when my browser was open at the Euphony interface and when it was closed. Apparently when there was traffic between my laptop/tablet and Euphony, that traffic was impacting the sound quality. The time lapse indicator for the track is updated by constant network communication between the Euphony server and the client interface. If I close the web browser I could hear an improvement. I open the web browser, and I could hear some degradation. This was only noticeable when streaming Qobuz. I could not hear any difference playing local files .Again, my network is highly tweaked now, my server is the only device on a dedicated VLAN... those are not differences I could hear before (and Qobuz did not sound so good before). I contacted Željko from Euphony, and he added a new setting in the latest Euphony version. It's called "Song pos. update" and you can set it to never. You guys give it a try and tell me if you can hear difference. I could not until recently... The track I was listening to for these tests was J.S. Ondara - Tales Of America - Lebanon. Back to Michael's post. There will always be improvements to make. Even if you have the Extreme, we will always be curious what a $5K power cord can do. Once we have that, we would be curious what a $10K power cord can do. The original Extreme was released with two Intel Xeon Silver 4114 CPUs. Those are 10 cores / 20 threads with 13.75 MB cache and 85W TDP. Shortly after that an Intel Xeon Silver 4214 was released with 12 cores / 24 threads 16 MB cache and the same 85W TDP. It's a drop-in replacement that gives you 4 more cores / 8 more threads, more cache, etc. Would that be an improvement? Who knows... Would I be curious to try it? Hell yeah! So @ASRMichael the other test I would do if I was you would be the Mundorf silver/gold wire for the CPU/EPS connector.
  5. Nice! Excellent choice of components. Don't forget you may need a new BIOS for the Ryzen 7 3700X. If I was starting a new build now, and if I did not have piles of parts I have been testing with just laying around, and if I was using USB output, and if I was not in a hurry... that's a lot of IFs... I would probably wait to see what JCAT is cooking. I have been testing the JCAT Signature LAN Cable GOLD. It's really good. I am using it from my etherREGEN to my server. And it immediately became a main part of my system. It's one of these things that you put with huge disbelief (like the Stillpoints Ultra 6 under my speakers), and you realize right away that this is not going out of there. And you can't un-hear what you have heard. Damage done forever :). I have been been trying to avoid high-end LAN cables. The most expensive one I own is AudioQuest Vodka. I also have some cheaper ones - Supra, Cable Matters, etc. The AQ Vodka was better than the rest. But this JCAT is on another level. Granted, I don't have anything more expensive than the AQ Vodka, so I can't compare it with other expensive LAN cables. I gave some feedback about the LAN cable to @Marcin_gps from JCAT, and he told me that he is planning to release a new USB card soon. He called it a "game changer". He also said that the existing USB card is not going anywhere. But if you are in a hurry, you can't go wrong with the JCAT USB Card Femto or the PinkFaun USB Bridge with ultraOCXO clock. I've written about both of these cards before.
  6. My guess is that the two EPS connectors are connected in parallel. It's easy to check with a multimeter. If the two connectors are in parallel, and you use two power supplies, you are essentially connecting two power supplies in parallel. It is hard to predict how well that is going to work. You can think of it as connecting two batteries in parallel. It would not cause a short (unless you do something wrong). But you may not get 20A if that's your goal. Or you can wait for my guide. I have done some good progress on it and have been trying to give it the usual level of attention to details. Short wires, good power, and vibration control are not easy problems to solve when you change the clocks as an afterthought, rather than implementing them on the motherboard as part of the original PCB design. All three are very important key items.
  7. Thank you for taking the time to write this review @gererick. I am glad you like the overall result. Give it another 3 months, and it would open up even more. Everything is still burning in and as you have already noticed it keeps improving. The big Mundorf caps, the big toroidal, the OCXO clock, and the silver/gold wires in teflon all take good couple of months to break in. This computer will shine at its best around around mid-April.
  8. He just told me he is about to post some feedback. Keep an eye for it.
  9. Big THANK YOU to @Blackmorec for taking the time to write this review. The best thing is none of us has any financial benefit to promote this cable. Hopefully that makes us somewhat trustworthy. We are just trying to spread some awareness. I have published everything you need to know to be able to make those cables yourself in this thread already. If you can't solder, you can probably find someone who can do it for you. I have tried many different wires, materials, geometries, shieldings, etc. This includes Neotech 7N pure copper, Gotham, VHAudio starquad and many more. The only other cable that might be very close to this is the AudioSensibility OCC Solid Silver. I have not had it in my system. But a member who has both reported he liked the Mundorf better I think adding 1% gold to pure silver would still have just a little bit less resistance than pure copper. But that's just on primitive calculation using resistance coefficients of 1.6 for silver, 1.7 for copper, and 2.4 for gold. With these coefficients, we end up with: 100% pure silver - 160 99% silver / 1% gold - 160.8 100% pure copper - 170 It seems like the resistance would still be less than pure copper. But of course we are looking at just one variable of something quite complex that no one really understands why it even makes a difference. @Blackmorec powered up his network devices with this cable, not his streamer. Granted, the effects of powering the CPU are even bigger. But I was able to hear differences between DC cables or even different shields powering up my cable modem. I have absolutely no explanation about that. The other thing I tried was different pure silver wires with different gauges and purity. None of those silver cables sounded good to my ears (as DC cables). I've had similar results with silver wires for analog signal. Way back in my audiophile journey I thought that I could buy silver wire for jewelry from ebay and make interconnects that would sound as good as some of the top cables out there. Wrong! Then I thought maybe I need to be buying better purity silver, and I bought 9999 silver. This did not sound good either. After trying many different silver cables for analog signal, I found one silver cable that sounded excellent (after 1000+ hours burn in time). And it was not from the eBay's jewelry department. It was a meticulously perfected wire by a high-end audio company. That tells me there is a lot more than the resistance of the wire. And I guess whatever Mundorf is doing with this wire is just right. I also think that the 1% gold might be just enough in the mix to "fix" some of the brightness associated with silver. But be careful how you read all these reviews and comments. So far everyone who tried this wire was impressed. Most of the people, just like @Blackmorec replaced their Neotech 7N copper cables from Ghent. But you need to have a resolving system to hear those differences. That wire is not going to fix other bottlenecks in your system, and unless you are at the point where you've done almost everything else and trying to squeeze out the last few bits of performance, your money can be better spend somewhere else. It's hard to quantify exactly how much difference DC cables make - in my system, I would say subjectively that the difference is somewhere between a power cord and a fuse. If you can't hear differences between power cords and fuses, don't waste your money on this cable.
  10. It depends on the application. The main clock on my AMD motherboard is 25MHz. It's 24MHz on my Intel motherboard. My I2S bridge uses the 24.576 MHz. The USB bridge uses a 20 MHz clock.
  11. Thank you. Never tried them, but I have been following the development of the C5 and waiting to be released. Those don't look like OCXO clocks, but that does not mean they are bad. A well implemented TCXO clock could be a lot better than a bad implemented OCXO clock. My guess is their clock has worse specs than the PinkFaun ultraOCXO, but that does not mean much. The reality is how I will implement the clock would make a bigger impact than how good the clock is. I may get the best clock available, and if I feed it with bad power and use long wires, it may be even inferior to the crappy oscillator my motherboard uses, with overall worse result regardless of how much I have spent. But I am aiming to use the best clock I can and implement it to the best of my abilities. Unless the clock is on the PCB right next to the chip, it's always a compromise.
  12. Yes, PinkFaun ultraOCXO. That's the best clock I could find on the market. If anyone knows anything better I can use, please let me know.
  13. Very interesting. Thank you for sharing. Can you share more details? What clock are you using? I see some external grounding (an external grounding box?)? What is the red board? A relay? Thank you.
  14. So you are thinking that Melco used the Buffalo BS-GS2016/A with better power supply and maybe a better clock? That would make the Buffalo BS-GS2016/A switch an excellent choice for tweaking. Do you happen to know the difference between BS-GS2016/A and BS-GS2016 (without the /A part)? The BS-GS2016 model is $210 on Amazon. Hmmmm.... not that I have spare time for more projects, but that is quite intriguing . How about EuphonyBridge ? I have been thinking about this too, but I think networking is completely different than processing music. So what works best in one may not work best for the other. I personally would not go that route though. But I am up for modding a Buffalo switch. I even have a 25Mhz PinkFaun ultraOCXO clock.
×
×
  • Create New...