As long as your entire chain works on “thresholds” aka digital levels all you need to meet is a specific margin for noise and timing and all is fine at the end result (there can still be issues, which is why error correcting codes like reed Solomon codes exist). The moment you want to include a continuous amplitude analog or mixed signal circuit all sorts of noise can manifest and affect the device in different ways (ground plane noise being most common), some can be captured well by using analysers, some not at present.
Even the audio library (no dsp changes) can change the sound. If you’re comfortable with command line, maybe give wtfplay-live a try to hear for yourself. It is free. Regarding the audiophile ssd, I don’t have much clue, the controller is not an issue if you’re not looking for massive io speeds (which is generally obtained through complex algorithms hiding latencies) and a low complexity controller can be more predictable for adding further filter stages if they planned to.
Not sure who is missing the point.
The precise atomic clocks sending the data are analogous to your perfect source. And they need to be precise at a level that is orders of magnitude beyond any audio system ever created.
However, the local noisy device is still in charge of capturing , processing and interpreting said data..
Data that is way more precise than the cleanest music file can ever aspire to be..
And then performing its calculations on the ultra minuscule deltas on the arrival time of RF travelling at the speed of light, no less.
To determine your exact position on the surface of the earth down to a meter of wiggle room.
Now the argument on this thread was that electrical nasties on an audio processing device can cause time domain and other drifts on audio signals .. (Correct me if I am going on a complete tangent)
So if that were indeed the case, surely the same would show up in this scenario too ?
It would. But you have the option of checking and re aligning things multiple times, run complex digital filtering algorithms, and also have support from other parameters like your cellphone internet location, etc. No such option in audio so far.
Realtek has no history with PC audio - it makes garbage audio chips that sound like $hit and garbage network chips that drop connections randomly. Any internal sound cards worth their salt use CMedia chipsets. Realtek's NVMe controller is equally bad - loads of people complaining about sudden disconnections. Only thing realtek does is sell things for cheap - ultra cheap. Probably that was the only motivation here.
Realtek dacs sound great as long as the motherboard manufacturer has not screwed up the signal line or interpolation algorithms (or if you don’t run the output wire inside the computer for the front socket causing all em noise inside the chassis to catch on to the audio signal, use the rear socket). You clearly have a bias against Realtek.
It’s not a chip to compete with dedicated dacs anyway but within onboard devices, Realteks dacs decently outperform most other portable devices including iPhones/macs audio output in my listening experience.
And that is the reason many have experienced that playing song from hard disk (which has motor) sounds less enjoyable than ssd.
Ssd is a noisier source than hdd (and there’s more factors on top of this, the device density, slc vs mlc vs tlc, the controllers, even the firmware). If I were to build an audio server, I’d likely use a hdd and/or properly formatted sd card, preferably slc as cache.
Alas, I find convenience to be a big part of the experience so I’ve went Bluetooth, despite all fidelity oriented trade offs!