Since I plan to use mainly active speakers (Devialet Phantom), and feed them with Optical input, I wonder if there’s any benefit of going for something like the recommendation, or simply get a Raspberry Pi with a HifiBerry Digi+ Pro?
As I understand it, since the Streamer basically just gets the Roon Signal and translates it to optical bitstream, there should be no difference? The Raspberry Pi is like 1/30th of the cost
There should be no sound quality difference, let alone one that justifies the order of magnitude difference in cost. As long as you’re comfortable building a Pi+HAT, and they know that, you can safely assume whomever is trying to sell you on an element-x is either delusional, doesn’t know what they’re talking about or is trying to get money from you.
I’ve been using the optical from a Digi+ Pro for a few years now, I suspect it’s not just cheaper but a better fit for your needs. You just want the optical out right? It’s not even clear to me that the element X offers that, the specs show digital inputs but no digital outputs, just RCA, XLR and various headphone outs, so it would have to be analogue from that to the Phantoms.
Final thoughts if the Pi doesn’t work out you’ve lost very little, if the element X doesn’t suit, that’s a lot of cash.
It looks like the Matrix has a DAC so you don’t need that if your speakers are digital active (include their own DAC) - anything that outputs Toslink optical will do. You’ll need a HAT for the Pi as it doesn’t do optical out but you can get complete kits from the likes of Justboom and HiFiBerry - very simple to setup.
I don’t know of any other units further up the HiFi price ladder that focus on Toslink, that sort of gets sidelined in favour of coax SPDif or USB - which is pointless if your speakers want an optical digital input. I would have thought the Phantom would be more than happy with a Pi based Roon endpoint.
PS. It seems you can also go directly to the Phantom using Airplay without any other Roon endpoint being involved so that’s an option, if you aren’t interested in hires content - Airplay is limited to 16bit 44.1Khz. It also does SPotify Direct and UPnP so other music delivery apps can work directly with it.
Apple actually resamples to 48khz on airplay1 I believe which means it is not bit perfect. Still sounds pretty darn good though!
The other thing to try is upnp, while not officially supported by roon there are ways to make it work. This would provide for bit perfect sound, volume control from roon, and would be an optimal work around until devialet formally adopts roon, if they ever do.
Also, devialet used to require their “dialogue” hub to allow multiple phantoms to sing together in stereo, looks like this may no longer be required but it did offer a usb input and might be worth looking into. If it works with your phantoms it might be the easiest and cleanest solution for now to just use a pi to hand off usb to a dialogue.
So if OP is to follow your advice based on that post, they’d pay around 5x as much to get a better case, something they can switch on from the front, and, oh yeah: not the output they’re looking for. But it’d subjectively sound better according to some guy on the internets, with no credible explanation as to why, and no measurements to back that up.
Yep, you have it right. Just use a Pi. There’s no real point to these higher-priced streamers, other than providing a revenue stream to the manufacturer and their various advertising flacks. You might want to read this article:
Pair something as nice as Phantoms with a device that is nice to look at and use. Preferably Roon Ready for a guaranteed user experience. You may as well because you will end up there in the end anyway.
I never used the Element X, but for PCM material this (with a Pi 3 or Pi 4) is as good or better than >$1K streamers I’ve owned. It does not do DSD, though, which I don’t use anyway.
(Darko.Audio contributor, the dad in DadFi (coming soon))
That’s not what I said - I gave examples of several streamers that sounded different. In these cases the more you paid the better the SQ. Yes, diminishing returns applied. Whether the sound quality differences are worth paying for is a judgement for the individual.
I didn’t argue for or against buying the Matrix as I haven’t heard it.
You appear to be saying that there is no difference in sound quality between streamers. If so then we disagree. My comments were based on extensive experience with each of the options mentioned. Several months with the dCS Bridge, nearly two years with the S2 Stream Ultra etc.
No I don’t have measurements, it would help if I did as they can be useful. They are not, in my opinion, the be all and end all though. I’ve heard examples of things that measure better but sound worse (e.g. Allo Katana with the two different Op Amp boards - best measuring and best sounding).
None of which I suspect will change your views. If so fine, it’s not life and death
Well, the Raspi4 is a DIY product and you need to enjoy it.
Every now and then it will not boot as your memory card has an error. This can happen every month or every 5 years. It is better happening every month, otherwise one forgets how and where to burn a new one.
Depending on the software you use, it may occasionally forget that it has a HAT and then you need to teach it that again.
You can not cool it with a HAT, so the CPU will run at 60C. Not sure if it will last 5 years. You could easily buy a spare one for those Sunday afternoons when it blows up.
I am a great Raspi4 fan and have a couple of those attached to Roon. But there is hardly a week without a glitch in one way or another. For 1/100 of the price for a real Pro box I gladly fix it.
More nuanced than that, but pretty much, yes: I’ll totally buy audible SQ differences between streamers with integrated DACs, or when connected to truly shoddy DACs. I’ll also kinda buy there might be something audible with SPDIF in some cases, but I kinda think there’s very rapidly diminishing returns. I’m not entirely certain that the inherent jitter over optical is audible, for example, or that AES in is audible over SPDIF, but I’ll still rather use AES over the other two typathing.
Assuming you’re using USB in both cases, how do you explain the differences ? Your Pro-Ject review mentions very audible, but yet very subjective, differences - “air”, “tone”, you name it. These should be easily measurable, if not by you, by Ayre, who should be extremely concerned if you’re right. Same with Pro-ject and your USB drive experiment: you’re naming and, in this case, shaming their designer, who should be truly alarmed if you’re right. What did they have to say ?
I’m not certain if you’re intentionally… blurring the goalposts, or not. If you’re not, which I don’t believe, it’s not the best analog-y (ha!). If you are, let’s remind everyone that claiming analog world differences is a bit different than claiming USB differences, because digital. If you’d said “Allo has provided me with measurements at DAC output that differ between the USBridge and USBridge SIG”, then it’d be another matter.
It probably won’t, but it’s a worthwhile discussion, even though I completely agree with Archimago that the best use of their readership’s attention for mainstream reviewers and publications would be to start holding the record companies’ feets to the fire for what they’re doing to the source material. Probably less controversial and lucrative than advertising expensive cables and decrapifier do-dads, but the advantage is you don’t need access to an Audio Precision device, or to know how to use it, to easily explain why what you’re hearing is better or worse than the previous version.
Oh, and just to clarify one more thing: I completely believe that the likes of Pro-Ject, dCS, and Ayre bring value over Raspberry.
One is in industrial design and finish of the casework, which is an absolute disgrace for every single Raspberry solution maker I can think of. Of course, that stuff is extremely expensive, but that isn’t an excuse to not make it an option.
The second is support: there’s tremendous value there for some people, and we’re a crowd whose behaviour, if we’re honest, often justifies demanding extortionate amounts just for the aggravation of having to indulge us without having the option of telling us to stick it.
I’m curious about your experience with stability problems with the Raspi4. I have two of them that I’ve been running for several months with zero problems. I leave them powered on. Maybe that’s helping me reduce the chances of having a memory card problem. Do you think the difference in stability is the luck of the draw? I’m running Ropieee but have also played with DietPi, PiCorePlayer, and HiFiBerryOS. What are you using for OS?