Why do DLNA apps sound differently (or do they)?

[Mike Rubin]

I don’t know enough about how network protocols work even to hazard a guess about this, but it seems to me that there are differences in what I hear when I use one of two Android DLNA streaming apps.  My music is on a NAS, the streaming software is JRiver on a Dell XPS 8930, and the renderer is an optical Sonore Signature Rendu SE.  The signal travels via copper to an etherregen and then via optical to the Rendu.

 

I have been controlling the streaming with an Android tablet or phone via wifi.  The controller apps I use are JRemote2 and HiFi Cast, both streaming from the JRiver library to the Rendu using DLNA.  To my ears, they sound different, with HiFi Cast having more “presence” in the midrange and JRemote2 sounding better balanced from top to bottom.  On some material, one sounds better and, on other material, the other sounds better.  On the other hand, the HiFi Cast stream usually sounds louder and that might be the difference I am hearing.

 

I don’t understand what might account for sonic differences when the only variable is the software on the tablet or the phone.  Am I just imagining things or is there a reason why these apps, which I understood to serve just a traffic cop function, actually sound different?

 

Also, for what it is worth, I had a trial of Audirvana for Windows on the Dell, but ended up uninstalling it because it was highly unstable on my system, seemed too primitive to use for tagging and library management, and had issues keeping its remote app connected.  All that said, while I used it, the sound quality was superb, noticeably better than what I was streaming from JRiver in many ways.  As I understand things, though, Audirvana isn’t actually a DLNA server like JRiver, so there must be some other explanation for the difference in sound character despite the hardware being identical.

 

I also have noticed SQ differences, albeit not large ones, between the DLNA apps and Logitech Media Server with the same hardware.  To my mind, different protocols indeed sound different from one another, but, as I said, I can’t guess why two different iterations of the same protocol would sound different from one another.

[mansr]

The app on your phone (control point in DLNA parlance) normally simply tells the renderer what URL to play. The renderer then fetches the data directly from the server. In some cases, the app acts as a proxy between the renderer and the media server, and then it could of course be altering something (some such apps have the ability to do format conversions). An easy test to see whether the app is proxying data is to disable wifi on the phone while playing a track. If playback continues, the app is obviously not touching the data.

 

Besides the data delivery, the control app can also set the volume (and possibly other things) on the renderer. If your renderer has a volume readout, I'd check and see if it shows the same level with the two apps.

[Mike Rubin]

7 hours ago, mansr said:

The app on your phone (control point in DLNA parlance) normally simply tells the renderer what URL to play. The renderer then fetches the data directly from the server. In some cases, the app acts as a proxy between the renderer and the media server, and then it could of course be altering something (some such apps have the ability to do format conversions). An easy test to see whether the app is proxying data is to disable wifi on the phone while playing a track. If playback continues, the app is obviously not touching the data.

 

Besides the data delivery, the control app can also set the volume (and possibly other things) on the renderer. If your renderer has a volume readout, I'd check and see if it shows the same level with the two apps.

Thanks. You have confirmed what I thought, i.e., the tablet and phone apps just route and manage the traffic.  Your suggestion to disconnect the tablet from wifi was a good one.  Disconnection of the DLNA made no difference in the stream, so it is probable that I am just hearing volume differences.  I don’t have access to the volume information, though.

 

Also, just for grins, I reinstalled Audirvana, which I controlled using remote access to my desktop.  It still sounds pretty exquisite to me when it streams,  I may keep it installed this time, even though I will continue to need JRiver for library management and metadata editing. 

[manueljenkin]

All data transfers take finite time and have a series of fetch, buffer and transfers using lot of intermediate logic. Usb audio protocol doesn't mandate proper transmission and doesn't guarantee that it'll fail when data is missing.

 

Not to mention different timings of data can have different noise patterns.