ricko01

They say the root of all evil is money (probably true) and it certainly is with objectivists verses subjectivists. Assume for a minute that a transparent component (DAC, amp, server, NAA etc ....) in the objectivists world can always be had for a max $500 and can solely be decided by measurements and the subjectivists say no... you need to spend a max of $1000 and the purchase decision can be purely based on listening or measurements as a first cut and then listening to the candidates. If this was the situation then objectivists wouldnt get huffy (cause the price difference is only $500) but they do simply because subjectivists are willing to spend (as per the example above), $40k on a music server when a RPI would/could be just as good. Or $10k on a power cord or $5k on some small round things you stick on the wall. They also despair when veil after veil are lifted, incrementally, after each new "thing" is added to a system. My comment wont further this debate any but to me, objectivists at least want to get value for money. Peter

Captured this from a qobuz stream...I did so cause the album sounded strange. The album was recorded in 1975 so predates a digital recording/master process but obviously has been digitized from tapes. what are these three lines at ~9.5k, ~18k and ~19k? Also not much energy above 20k... could be related to the digitizing process? Peter

Read most of this topic and based on my experience bench marking extremely large Unix and Linux system configs for customer workloads in pre-sales situations, the one issue I see with the method used is that a single monolithic process was used on the PC. Yes this use case would tend to match what our use case as audiophiles would be but a PC under 100% load with a single process wont cause excessive contention between competing processes. (i.e. massive amounts of context switching, disk fragmentation, network packet retries etc) To me, the test would actually need to be as destructive as possible (i.e. with extreme contention for all the PC'S resources) to show that the measured output stays within scope. Ideally the test would allow for increasing levels of concurrent processes to determine when audio performance degrades, if it does at all. This would then prove that the test rig is indeed immune to ANY level of noise or that potentially noise can indeed be an issue in a normally well behaved equipment stack (i.e. alluding to situations where noise could affect poorly designed gear) Peter

I am a low level Unix/Linux guy going back to 1980 so cant contribute much given the OS is windows. On LInux (as I am sure many people know) you can run "strace" against a process which will show you the system calls a process is making. Is there no similar process for windows? This way you would have an exact idea about the RAM memory requests/allocations, disk allocations etc. the app is doing. Using generic/high level monitoring on any OS (1 sec intervals is the lowest) wont have the resolution to allow you to trace with any accuracy the resources a process might use hence you need to track it at the call level. If someone with windows knowledge can provide the protagonists with how to do this and they then provide the dump files then we might advance the speculation. Peter

Thanks for all the replies and I will quote John as a reasonable summary of the thoughts. It still leaves me confused because if you read any objective review of a DAC, any graphing of filter response always(?) focuses on the Nyquist limit. And to frame this, I am only talking about most DAC's that oversample/upsample which is the most common scheme. However, if I look around the sox website I see this link; http://sox.sourceforge.net/SoX/NoiseShaping which shows the potential for this to affect the audio spectrum at amplitudes and frequencies we can comfortably hear. Not sure if noise shaping only relates to downsampling. Guess I need to do more reading. Peter

So my limited understanding of filtering is it all occurs at Nyquist (say ~22k for CD). Unless you are young and can hear close to 22k, most audiophiles that can afford a decent system are in the older age bracket with some degree of high frequency hearing loss. So why all this debate/preference around different digital filters (some DAC's run with 7 or more) when all the action happens at frequencies most older audiophiles cant hear. This question assumes the filters are good enough not to reflect back artifacts below Nyquist. The same question applies when using say "sox" to upsample... does the filter choice really matter. Peter

Audiophiles: Dead or Dying? Obviously older generations of audiophiles are in fact dead or dying so I assume the question relates to new blood In this regard, two phrases: disposable income / social media Looking back to when I started earning from my first job, the #1 goal was to save as fast as possible to get a decent system. The only distractions from this were basic living costs (including beer) and chasing the chicky babes. For any normal guy today (cause audiophiles are mainly guys) the same basic expenditure exists but what there is available to throw disposable income at has increased dramatically. A new blood audiophile's life will revolve around his phone and social media so that means cell plans and $1500 phones (probably updated every year or two) It might mean some kick arse gaming system as well... maybe a big screen TV to stream netflix. And travel... due to the reduced costs... gotta travel to pad out your social media profile. None of these existed in my day so raising the cash for a system was a single minded exercise back then. Given youngsters today are so phone centric then we have obviously seen the raise of hi-fi phone components but they arent buying the traditional, big arsed systems we did back in the day. As an anecdote, my two step daughters are attractive and when they hit their late teens/early 20's (early 2000's in Miami), we had a parade of boy friends through the house, where my system was on full display. Never once did one of these guys even show the slightest interest in the system. I know that if I had walked into my girlfriends place and her Dad had a kick arse system, I would have been all over it. Basically youngsters today dont care about big box systems and I dont think that portable hi-fi will sustain the industry when the current older audiophile generation dies. Maybe I am biased but I dont think that running around with a phone, a good portable DAC and headphones equals being a real audiophile... to me being a real audiophile, no matter equipment level, means not only a passion for the reproduction chain but also a devotion to carving out time to listen exclusively to the system and not as a shared task while using say social media. Social media killed the audiophile.

There are basically three types of posters/users: the die hard regulars, the casual information seekers and the lurkers (who dont post) Most of the issues I suggest tend to be with the die hards so why not allow a user to have a "no post" list on their profile so that any thread they start wont by "crapped on" by people they historically have an issue with. This wont be an issue for the casual information seeker as they wont have any experience with "thread crappers" and will generally get answers to their questions anyway. Lurkers (like myself) dont care (in fact we enjoy the back and forth) but it will allow the die hard regulars to have crap free threads. Peter

I have a client/server topology using LMS (server)/squeezelite (client) for music playback and nfs (server)/kodi(client) for music video playback. Because everything matters, I have looked at ensuring that the USB bus used for the USB connection into my DAC from the client PC is reserved only for audio traffic. As part of this, I drilled down into how much bandwidth is actually used... so thought I might share this. Ignoring the baseline hub devices, I have the four active USB devices: - the USB stick I boot ramroot off - the port that goes to the DAC - a USB wifi dongle (used to allow me to 100% control network bus utilization) - a USB dongle for mouse/keyboard The bus allocation is as below: ---->Bus 001 Device 003: ID 0951:1666 Kingston Technology DataTraveler G4 [BOOT USB] ---->Bus 001 Device 004: ID 20b1:2017 XMOS Ltd [DAC USB] ---->Bus 002 Device 003: ID 0cf3:9271 Atheros Communications, Inc. AR9271 802.11n [USB WIFI] ---->Bus 002 Device 004: ID 046d:c52b Logitech, Inc. Unifying Receiver [KEYBOAD/MOUSE] With the system "idle", I saw constant loads. Interestingly the keyboard/mouse dongle has no traffic when not being used: - BOOT USB does 0 KiB per second - KEYBOAD/MOUSE does 0 KiB per second - DAC USB does 4 KiB per second for both send and receive - USB WIFI does 20 KiB per second for both send and receive When moving the mouse, total bandwidth used (send+receive) was 10 - 20 KiB/s To test my worst case utilization I streamed a music DVD with a 16bit/48k sound track (worst case cause the network traffic includes both video and audio) Graphing the network load over a 15 min period we see cyclic transfers peaking at ~15MB/s (testing squeezelite shows similar cyclic transfers but peaks were ~11MB/s for ~4 seconds at the start of each track) Looking at the USB DAC traffic for the 16bit/48k sound track we see basically send and receive in a constant narrow band of utilization (to the DAC a constant 650 KiB/s and from the DAC a constant 300 KiB/s) For my purposes, this analysis showed the placing of the network dongle and USB DAC cable on separate ports did in fact provide separation and also confirmed that during playback the other devices that co-exist on the same buses as the network dongle and USB DAC use zero bandwidth. It was interesting to get metrics on actual bandwidth utilization as part of the analysis. Peter

yep... the digital outs are all limited to 96k.... which points to an internal max of 24/96. not a concern for me

however you may be right... that internally its all 24/96... I will check on this

Was true. Their most recent products (shown below) support high bit rates and ROON. The pre-mate has two analogue outs, the HDP-5 has three. We see that either supports the following (all PCM based, no DSD)

A bit late where I am to fully follow exactly what you are doing but it sounds like one of the DEQX products might be what you need. They have various processors but all provide preamp/dac/eq functions with multiple digital and analogue outs for different "zones". The processors support two or three zones and you can configure each as needed. They are well reviewed and I use an HDP-4 for a four channel (all full range) system (noting its 2 channel IN, four channel out) see: https://www.deqx.com/ Peter

Holy crap!! After reading some of the comments I was peeing my pants. So I rushed downstairs, pulled out the multi-meter,, powered up my HDPLEX 100W and.... 5.1V into my SSD drives 19.3V into my motherboard I also connected up my WattsUP! meter and load was ~35w at idle increasing to ~45 watts when running Kodi or reindexing via LMS...so no stress there Panic over. It would appear from the comments that the orginal 100W box historically was fine, the 160W not so much and the new 200W has had "birthing" issues but is now ok.

Thanks for that. I have an early 100W HDPLEX which has fixed rail voltages (not the dialup/dialdown of the newer models). works fine with powering the MB from the 19V and 5V for the two SSD drives. If the Iso Regen does help then probably not worth moving to a better LPS or maybe if it does then that might indicate power supply issues so a better supply might have additional benefit. Obviously one change at a time but with my simple setup and my recent move back to a memory resident OS, LPS quality and USB cleaning are all thats left. Thanks again, Peter