Serento: The audiophile UPnP controller app for iOS/macOS

[simbun]

(27-08-2026 09:33)Donuk Wrote:  As I understand it, the playlist on one's streamer is simply a list of addresses for the streamer to read from, say, a NAS. But I am willing to be educated....

Some control points have the ability to transcode streams when the target renderer lacks support e.g. playing DSF to Sonos.

Quote:The one principle I refused to compromise on: it stays out of the audio path. Serento finds your server and your renderer, tells them to talk to each other directly, and steps aside.

Quote:No audio processing - zero DSP, zero touch on the signal

Translated, I presume these mean that Serento won't/can't transcode; instead it'll either pass the stream URL to the renderer as-is (which will probably fail silently) or display a compatibility error in the UI. In other words, from a playback perspective it's no different to mconnect, Lumin, Linn e.t.c.
I use BubbleUPnP which does have the ability to transcode, but I have it disabled as I'd prefer MinimStreamer to perform any conversions.

If you're on Windows, TuneBrowser is an excellent tool to view the communication between control point and renderer.

I captured the following events using BubbleUPnP and the Linn app (I'm on Android so can't test Serento but the commands will be the same):

Code:

Action                   Control point command
---------------------------------------------------------------
Enqueued an album      = Insert (per track)
Play the first track   = SeekId
Opened Linn            = ReadList (populated the queue in Linn)
Advanced to next track = Next
Stop playback          = Stop
Cleared the queue      = DeleteAll

Had I not been interacting with the OpenHome queue, playback would have continued uninterrupted with the control point only polling/listening for state changes (track progress/transition) so they could be reflected in the UI.

Here's a screenshot of the OpenHome renderer in TuneBrowser with the UPnP log showing the first Insert call from BubbleUPnP.

   
As you can see, in OpenHome the renderer holds the queue. In standard UPnP-AV only the current and next track are known to the renderer with the control point holding the remainder.

All of this is to say that if the control point isn't doing any transcoding, the renderer pulls the audio track straight from the server, meaning the control point can't affect the sound quality in any way.

There looks to be a lot of other features to like, but sound quality isn't one of them.