Raspberry Pi vs more expensive Roon Streamers?

Thanks a lot for the long (and considerate answer).

On measurements, we indeed disagree. I don’t think there’s any magic left in audio, but I do think that one should demand from any manufacturer who claims magic not conceptual explanations (like a shorter signal path or this input sounds better), but bench-proven, independently verifiable explanations: they’re either engineers, or they’re magicians. If they can’t properly explain it by engineering, then the hypothesis must imperatively be that what they’re doing is magic, defined as a branch of (applied) psychology. Put rudely, no matter how suave and friendly they are, the hypothesis must be that they’re bullsh*tting unless they’re able to prove otherwise. “I tried and it also sounded better to me” is meaningless, because we can’t evacuate the hypothesis that we weren’t conned by the suave engineer.

The Gold Note example you gave is, I think, an interesting one: measurements would likely have told you if the analog out of the Gold Note is slightly hotter than the one you compared it to, or slightly better suited to your gain stages, or if its reconstruction filters were maybe more to your liking, all things that of course don’t make for a good piece or happy people at Gold Note, in part because they can be fixed for cheaper than a CD player (no matter if it can do double duties as an unusually competent DAC).

Appearances are interesting, because that’s where the “naïve friend-slash-spouse” fallacy shows its true colours. Remember the (in)famous Carver amp experiment, and imagine a HiFiBerry (or a USBridge Sig or any other sub-1000 streamer) in a dCS case, with dCS marketing behind it, while we’d put one of the higher-end solutions in a humdrum case with $99 board-level-copywriting. Would you be able to tell the difference, and don’t you think that your naïve friend-slash-spouse would fall for the “more expensive” solution ?
For those who don’t know, there’s a historical, bordering-on-the-con, example of this. Back in the day, Lexicon recased an Oppo (and there are the Goldmund, Theta and Ayre mods, but that’s another matter entirely), to interesting audio reviews: “(…) everything was presented in a smooth and detailed fashion, with good air around the various instruments, and made for great listening. (…) While this is not the best recording, the music is timeless and the Lexicon brought these aged tracks to life. (…) I was treated to an open and lush rendering of the track. Gone were the edge and glare often found in lesser players. The openness of the piano was excellent.” (etc, etc, etc).
This is by a reviewer who knew that he was dealing with what he indicated he though was essentially identical hardware… now imagine the same with a consumer who walks into the store and is treated to a bit of foot-tapping by a salesman who’s been sold pixie-dust improvements by a rep who’s been told the new product had this or that standout by the marketing team… I’d say Joe customer wouldn’t stand a chance not to get conned. Wouldn’t you agree ?

5 Likes

Oh, that… speak to mastering engineers / studio peeps if you haven’t. They’ll often blame the musicians themselves, where everyone wants to be louder than everyone else typathing (right, Metallica and Iggy Pop ?). I’d so, so love for labels to go back to the heavy vinyl days, and offer a mastering done without the egos getting in the way…

1 Like

I have a Raspberry Pi connected to a Audio-GD DI-20HE (DDC) with i2s and SPDIF out.
It beats my dCS Network Bridge , Aurender N10 and NAIM Uniti Core.

Same here. My RPi4’s have been utterly reliable.

Well said!

1 Like

I’ve got this issue with Adele. Great singer, can’t stand her albums because of the mastering. And while we tend to ‘blame’ the record companies/mastering engineers… I also think that there are a lot of artists out there that understand music but don’t understand sound.

In the mean time I’m going to sit quietly in my listening chair hoping/dreaming that the next Adele album is going to be produced by Steven Wilson :wink:

4 Likes

You can update the matrix element M to the latest firmware using just an android phone. I’ve updated mine twice this way since I bought it. No Windows OS required.

The Sabre DAC is only via USB and windows as its a pure DAC with no network connection. Matrix Streamers generally have network and internet access so can take over the air updates. It’s not an issue really.

1 Like

Correct Simon :+1:

Try a different usb cable. Makes a remarkable difference

I have been doing extensive tests with RPI4 and using a Windows surface tablet + fidelizer.
The only way Rpi comes close is using a powerbank (don’t have a linear psu) and even then it is missing the last bit of fidelity on the top end. This results in loss of realism,as it simply sucks the air out if the music.
Probably someone would say that this is a result from the darker background, meaning less noise, but the reality of it is that it sounds less convincing.
I am not sure a linear psu can help with that and not sure 200e is a price worth paying just to experiment.
I also consider that there may be a difference in the Denafrips driver for pc and rpi in which case it doesn’t matter.

Double blind? We’re talking about USB data, not the 3.5 mm audio output? I wonder if/how we can measure these sonic differences?

1 Like

Well, if they exist, they’re easy to measure. Just capture the output of the DAC in both cases and compare.

But it’s digital out from the Pi and Surface, so same bits. If the difference is real, must be a bad DAC.

3 Likes

I must be a nearly deaf and unsophisticated listener… for nearly five years I have been using a RPi with a wall wart as Roon endpoint, with USB out of the Pi into my DAC, and I have never noticed music ‘with the air sucked out of it’… I have compared it to USB and optical out from my MAC… and I don’t notice any difference at all…

But then again, I love my music too much and would enjoy it out of a transistor radio… wait… I did enjoy it out of a transistor radio, long ago…

2 Likes

I see no need for mocking.
If you are happy with what you have and don’t notice a difference I envy your bank account.
I do notice a difference between windows and Rpi and I don’t think myself a sophisticated “audiophile”. I have invited people over for a listen, friends who listen lossy music on lossy systems and they can pick up on a difference.
You don’t have the same system as me, so I don’t think it makes sense to deny something is real unless you can test it yourself.

BTW I listen on all kind of hardware too. Music is music and can be enjoyed even underwater (using headphones when I swim).

Not double blinds, or single blinds for that matter, just listening for certain queues over and over and over.
I don’t need to prove anything anyone, as I am not a reviewer, just doing this for my own benefit and listening pleasure.

The difference is only in signal delivery and driver. I am using wifi with the windows tablet and ethernet with the pi. Both are running essentially on battery, clean power shouldn’t be an issue (for those who even care about that concept).

And that’s what its all about :slight_smile:
I’m puzzled what can be causing these subtle differences though…

Well, the driver won’t matter – same bits. Got to be signal delivery, I think. But the DAC should render that irrelevant by properly cleaning the analog parts of the input signal. So, bad DAC.

1 Like

not true.

/tenchar