Or I may have made a joke. I know what David is saying; I just found it funny that audiophiles spend thousands on “transparent” systems when a cheap solution like RPi (which is what I also use btw, not because of the “PC noise”) does the trick quite nicely.
Well, I did just that with my Topping D70 DAC, which is by no means an expensive piece of gear, and could not detect any influence of PC noise in the analog output. I did it for fun though, not because I think it’s reasonable to investigate the area any further. Filtering USB or PSU noise is a solved problem, and the fact that there are DACs out there that don’t properly do it doesn’t change that.
No doubt. But in the interests of fairness and out of due respect to the article and the proffered hypotheses, it’s important to compare eggs with eggs, so to speak. Hence I drew attention to the article’s content surrounding a multi-duty device incorporating streamer/DAC and decryption, DSP etc. Notwithstanding your point which I agree with in isolation.
That makes sense, @Marian - again, due to my own traits, I don’t always pick-up on nuance. You may have made a joke and it bypassed me completely.
For the record, I did not realise that my wife was flirting with me and giving me clear (to most people) signals of interest when we first met; I needed to have it spelled-out to me. Thankfully the penny-dropped for me eventually (she had to tell me) - for other people it would have dropped with the subtlety of an anvil - such is my lot in life, but at least that episode ended-up positively for me. Thanks for coming to my Ted Talk…
High-resolution streaming is sold as “bit-perfect,” but the way Connect modes from Tidal and Qobuz work can quietly chip away at that promise. These features make the streamer’s CPU handle encryption, decryption, and decompression on its own, which can create extra electrical noise in the listening chain.
That’s the warning from David Snyder, a principal engineer at Palo Alto Networks with more than 30 years in networking. While no bits are lost, he says the extra processing can blur the fine detail and openness audiophiles expect
Hmm! The article goes half way to suggest the “connect” services aren’t bit-perfect and then reverses that suggestion.
That “bit-perfect” promise falls apart once your endpoint starts crunching FLAC
Hmm, really. I think the article needed to be clearer and provide the data.
Playing FLAC files requires a minimal amount of CPU resources, and the load is generally negligible on modern hardware. The CPU usage for decoding FLAC is typically around 0.2% on a single core, which is comparable to the CPU usage for playing an uncompressed WAV file.
The CPU load during FLAC playback is primarily influenced by the compression level used during encoding. Higher compression levels (e.g. level 8) result in slightly more CPU usage during decoding compared to lower levels (e.g. level 0) showing a 4% to 20% increase in CPU time for the highest compression levels when I carried out some investigation into this several years ago.
The default for most software when ripping/converting to flac is level 5 because it’s a happy medium between storage and decoding.
I think the purest of audiophiles reading this should really buy a server farm and store all their music as .wav files.
Correct, but the linked article was making the case that in achieving bit perfect streaming the CPU activities necessary for doing so can introduce noise, thus linking the two together. The problem is that there isn’t much evidence for that so a bold assertion is unwarranted.
Note that it is certainly possible for that linkage to exist but the stark generalization remains in the crepuscular world of disinformation. It’s easily provable for different devices, of course. The claimants just need to do the measurements and back up the claims!
