• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

SROVA: Linux-based headless hi-res player beta for NUC/PC and Raspberry Pi

What if the streamer has a built in screen? Can the Srova GUI be run on it as a kiosk service?
It probably could, since the SROVA UI is browser-based, but I’m curious about the use case. One of the main ideas behind SROVA is that the streamer itself can stay completely headless and out of sight, while you control it from a phone, tablet or computer.

If the streamer already has a built-in screen, what would you want to use the local SROVA display for — full control, just Now Playing, or something else?
 
The screen would be used for everything, just like EverSolo A6 is for example. Most of the time it would be NowPlaying screen.
I can see the appeal of the Eversolo-style built-in screen, especially as a permanent Now Playing display.

Personally though, I’m probably too lazy to get off the couch and walk over to a small streamer touchscreen when I can have the full SROVA interface sitting in front of me on a 13-inch Android tablet in landscape. :-)

That was really the thinking behind SROVA being headless in the first place — keep the player and DAC where they belong, and put the control surface wherever you’re actually sitting and listening.

But I do understand the integrated-screen use case much better now.
 
Yes to leaving LMS alone, but of course you could add that possibility for an all in one future solution, just not needed afaic.

The render node being a slimproto client is critical of course.

The controller software I can't imagine wanting to keep unintegrated?

I mean I "suppose" a playlist queue could be set up in browser / Material Skin and then just put a start/pause control in your control UI

but that'd be a super lame kludge IMO
 
Yes to leaving LMS alone, but of course you could add that possibility for an all in one future solution, just not needed afaic.

The render node being a slimproto client is critical of course.

The controller software I can't imagine wanting to keep unintegrated?

I mean I "suppose" a playlist queue could be set up in browser / Material Skin and then just put a start/pause control in your control UI

but that'd be a super lame kludge IMO
Thanks — I understand what you mean now.

You’re really talking about keeping LMS as the backend/server, but using SROVA as the complete controller/front end as well as having a SlimProto-compatible render endpoint.

I can see why that would be attractive for an existing LMS setup, especially if you prefer the SROVA interface to Material Skin.

For SROVA though, I don’t think that’s the direction I want to take it. SROVA was built to be its own player rather than a new skin/controller for another playback ecosystem, so putting LMS underneath it would add a lot of integration work without really advancing the core SROVA architecture.

I’d rather keep focusing on SROVA-native playback and, one day, build things like synchronized multi-room in our own way.

Thanks for spelling out the idea though — I had misunderstood the controller part initially.
 
OK sure, but just to add datapoints

I'm not looking at "either or", the LMS content is just one input, how I serve my FLAC files across the network. I would never let the render / controller nodes serve up local files, too limiting.

I am looking to **integrate** that LAN/NAS library streaming with various service streaming (Tidal Qobuz et al), as Roon does.

I won't let Roon become central due to its added cost, paid services must always be ancillary, optional, never necessary infrastructure.
 
OK sure, but just to add datapoints

I'm not looking at "either or", the LMS content is just one input, how I serve my FLAC files across the network. I would never let the render / controller nodes serve up local files, too limiting.

I am looking to **integrate** that LAN/NAS library streaming with various service streaming (Tidal Qobuz et al), as Roon does.

I won't let Roon become central due to its added cost, paid services must always be ancillary, optional, never necessary infrastructure.
Ah, understood — that makes the distinction much clearer. You’re treating LMS primarily as a network library server for your FLAC collection, with that becoming another input alongside TIDAL, Qobuz etc., rather than wanting SROVA to become an LMS-only front end.

I can certainly see the logic, particularly when you want the library centralized on the LAN rather than tied to individual renderers.

SROVA approaches that slightly differently at the moment because its "My Music" can access the NAS library directly, so there doesn’t need to be another server sitting between SROVA and the files. That’s why adding LMS as a library provider still probably isn’t something I’d pursue — it would add another library/API layer around something SROVA can already consume natively.

Longer term, if SROVA ever needs a more centralized library architecture serving multiple SROVA endpoints, I’d be more inclined to build that as part of SROVA itself.

But I understand the architecture you’re describing now, and your point about keeping paid services ancillary rather than essential infrastructure is one I can appreciate.

One extra point I should have mentioned: I actually run LMS on the same machine as SROVA today. They coexist quite happily, and both use the same music library on the NAS.

So I can already keep LMS around for the things I still want from it while SROVA accesses the same FLAC library directly. That’s another reason I’m reluctant to make LMS part of SROVA’s own architecture — the two can already live side by side without either depending on the other.
 
Last edited:
I have not found a UPNP / DLNA based library browser / playlist builder as flexible as LMS + Material Skin especially for those doing everything with metadata tagging

So if you won't add slimproto support, I highly suggest you check it out in detail as a model to emulate, as you continue to refine your UX in future.

There are many reasons strong community and industry support live on 25 years later...
 
I have not found a UPNP / DLNA based library browser / playlist builder as flexible as LMS + Material Skin especially for those doing everything with metadata tagging

So if you won't add slimproto support, I highly suggest you check it out in detail as a model to emulate, as you continue to refine your UX in future.

There are many reasons strong community and industry support live on 25 years later...
Absolutely — fair point, and I agree. LMS has been around that long for a reason, and there is plenty SROVA can learn from mature systems like it.

The SROVA journey is only just starting, so I’m sure the library and browsing side will evolve a lot over time. The key for me is to keep learning from what already works well while still letting SROVA develop in its own direction.
 
SROVA V2.0 — a little post-launch technical update, and what comes next

Now that V2.0 has been out for a little while, I thought I’d share a short technical update on two of the bigger additions — Qobuz and Spotify Connect — because they actually arrived in SROVA by very different routes.

Qobuz became a native SROVA source.

It is integrated into SROVA’s own browsing, search, queue, radio and playback architecture rather than being handed off to another player. The aim was to make Qobuz feel like it had always belonged there alongside TIDAL, Local Music and Internet Radio, while still using the same SROVA audio path to the DAC.

That involved rather more work than simply adding another login button, but I’m pleased with where it landed in V2.0. Qobuz now behaves as part of the player rather than as a bolt-on.

Spotify Connect took almost the opposite route.

For that I use Soloist as the Spotify Connect endpoint. SROVA does not try to become Spotify itself; instead it manages the relationship between Soloist and the rest of the player.

The interesting part was therefore not the Spotify UI — the Spotify apps remain the controller — but DAC ownership and handoff. When Spotify takes over, SROVA has to get out of the way cleanly; when Spotify leaves, SROVA has to recover the DAC and return to normal native playback without requiring the user to manage audio services manually.

So V2.0 ended up with two useful examples of the way I’d like SROVA to develop:

integrate something deeply when that makes sense, and coexist cleanly when it doesn’t.

October / V2.1​

I’m now starting the V2.1 work.

This is primarily a hardening and refinement release, rather than another large provider expansion. The current working list includes things such as:

  • DAC hotplug/re-registration behaviour with ALSA MMAP;
  • tighter Bit-Perfect / Volume Locked semantics so the displayed state reflects the real DAC digital-gain state;
  • further Spotify/Soloist ownership and recovery hardening;
  • Qobuz artist-navigation and library-loading improvements;
  • screen-aware context menus and some smaller UI polish;
  • Cast session recovery and further Cast presentation work;
  • easier local-network discovery of SROVA Players from the Android Remote.
There are also a few things I want to inspect first rather than promise, including headless Wi-Fi provisioning and whether Auto Mix really needs further work. The rule for V2.1 is fairly simple: look under the hood first, then decide whether something genuinely belongs in 2.1 or should wait.

The full working plan currently has 17 consolidated points across the Player, UI/UX, Cast and Android Remote, with the emphasis on strengthening the suite rather than forcing a short release cycle.
SROVA is still very young, and I’m enjoying this stage enormously — use it, find the rough edges, understand them properly, polish them, and add new things only when they genuinely fit.

As always, feedback from people actually using it in different systems is extremely useful.

https://srova.music
 
Is Tidal and Qobuz also bit prefect?
Yes. TIDAL and Qobuz are both bit-perfect through SROVA’s native playback path.
Bit Perfect operation is automatic in SROVA — there isn’t a separate mode the user has to enable. For native TIDAL and Qobuz playback, SROVA uses the direct ALSA path to the selected DAC, bypassing the normal system mixer/resampling path and following the source format up to the capability of the DAC.
Spotify Connect is a separate playback path via Soloist, so I treat that independently rather than folding it into the same claim.
 
Assuming you mean the playback chain:
TIDAL / Qobuz → SROVA provider/stream layer → decode to PCM → SROVA native audio core → direct ALSA (MMAP where selected) → USB DAC

That native path bypasses PipeWire/PulseAudio and the normal system mixer, and SROVA does not resample it. The DAC is opened at the source rate, subject to what the DAC supports.

Spotify Connect is different and is not bit-perfect in SROVA. Its playback is handled through Spotify Soloist and a separate managed audio path, which is why I keep it outside SROVA’s bit-perfect claim.
 
Does srova fully compatible with pi 3b+?
I haven’t personally tested SROVA on a Pi 3B+, so I don’t want to claim full compatibility yet.

I have had a user report that SROVA was working successfully on a Pi 3, which is encouraging. The SROVA Raspberry Pi package is ARM64, so the Pi 3B+ would need to be running a 64-bit Debian-based OS.

Pi 4 and Pi 5 are the Raspberry Pi platforms I’ve actually tested myself.

If you have a Pi 3B+ and decide to try it, I’d be very interested to hear how it performs — another real-world datapoint would be useful.
 
SROVA V2.1 is now available.

Hi everyone,

A quick update on SROVA. Version 2.1 has just been released for Raspberry Pi 4/5 and compatible AMD64 Debian-based systems.

Unlike V2.0, which introduced Qobuz and Spotify Connect, this release is primarily about hardening the audio system and improving reliability.

The main changes include:

  • Improved USB DAC hotplug detection, saved-device recovery and player-bar availability reporting.
  • Stronger DAC ownership and safer handoff between native SROVA playback and Spotify Connect, including recovery after audio-output interruptions.
  • Faster Qobuz library loading and improved artist navigation.
  • Viewport-aware context menus and several interface refinements.
  • Improved Google Cast session recovery in SROVA Remote 1.6 for Android.
The underlying playback approach remains unchanged: native SROVA sources use a dedicated ALSA audio path designed for Bit-Perfect output, without intentional resampling or DSP. Spotify Connect remains a separate playback integration.

Much of the V2.1 work came from using SROVA daily and testing how it behaves when DACs disconnect, audio ownership changes or devices are restarted.

Downloads and full changelog:
https://srova.music/downloads/

As always, SROVA remains free and open source. Feedback from anyone running it on different Linux hardware or DACs is very welcome.

Thanks,
Sasha
 
Back
Top Bottom