DLNA audio quality test with 3 different media renderers same speakers


This may or may not be of any interest but I'll share it in case it is. I have been playing with various ways to stream my digital audio library across the house using JRiver. As long as I have a media renderer at the end point I am set with the digital music served. I am playing thru some KEF LS50 wireless speakers but have three media renderers handy at that location. I was curious what difference if any I would hear as all the audio is digital and the DAC conversion is taking place in the LS50s. I've yet to step into a full fledged streamer as it seems like so much can be easily handled with other means if the user is reasonable tech savvy. Maybe that is incorrect. 

When going direct to the KEFs (wireless) the audio is super clear and very well defined. I then connected to a Sony Blueray player which is DLNA compatible. The cabling from the Sony Blueray (BDP 3500) is via HDMI output to the TV and then TV optical output into the KEFs optical input. Here after adjusting for volume differences I'm not sure I could hear much difference and the sound was still quite good. I'd give the edge to the KEFs in wireless mode but did not listen for super long.

Then I was curious to try the 6 yr old Samsung TV as the DLNA device as pretty much all TVs have this option. Easy enough to get working and again the TV optical output goes directly to the KEFs. Here the sound became relatively muddy and lacking in the same definition. Even weirder is that the right and left channels were now reversed. Yes reversed from the above two tests. I tested this like 3 times as I always suspect user (i.e. me) error.  I'm really not sure why the TV based media rendering sounds relatively poor and certainly no clue how the channels are reversed. 

So my simple lesson is that yes the actual media renderer does matter to a decent degree in the digital domain. I know to most this is probably obvious but I am a "show me" kind of guy. Maybe someone else has tried something like this or understands why the differences are so dramatic. 


jbs
Good to check the options but the variables and obviously the variety of systems make this tough.

Using rebuilt tube amp mono blocks as the driver, I use a LG TV as an option for its digital audio toslink. 

In my system, I use Audirvana 3.5 and send out the DLNA music stream to the TV using a Firestick. This sound is astounding. It's astonishingly good.

When I test it versus a wired connection, it's almost always within a hair as good or almost as good for even more than casual listening.

So, can't speak to the Sony but most of these digital audio toslink connections will send out a 48K signal. For $40, you can test DLNA with your TV to your audio system too.

Vastly overlooked and IMHO, underrated.

Thanks for that info. I have a modern LG TV elsewhere where I was gonna go this route and see what transpired. This result was really worrisome. It has to be the media rendering inside the TV and who would have thought it was coded so poorly as to reverse left and right channels. 

I have a firestick connected here and it has the channels and sounds seems OK on a quick pass. I don't see it (yes) in JRIVER as a renderer but the firestick does see the music library which means it works in that direction. I just need to play with it more.  
Using for the DLNA connection from Audirvana 3.5 to the LG TV: 
AirPlay&UPnP.  

People will argue about upsampling but it's not going to make much of a difference as it comes out 48K on the optical output from the TV.

For a few dollars though, it goes a long way. 

Just a quick related point; if you're looking to upgrade your TV anytime soon, Sony and probably many other Android TVs(rebranding to Google TVs) work as Roon endpoints. I simply connect the tv via ethernet to a network switch and use the optical output of the tv to powered speakers and get great sound without Airplay limitations. Easy, peasy, instant 2nd system and lyrics on-screen are a bonus. This approach leverages my Roon core and eliminates DLNA completely. Cheers,
Spencer