Maybe an odd question.
But do you ever notice that the Streamer sometimes sounds better after a restart?
This weekend I was listening for a long time and noticed a choppy soundstage (distortion?).
I rebooted server and player and the restlessness was gone.
1 Like
I will give it a try. Thanks.
1 Like
Others have reported that. I think I noticed this a few times but it’s hard to say.
1 Like
I suggest turning off and restarting your system approximately monthly. At times it makes a noticeable difference. It is the least costly and most effective tweak I have found. Many scoff but most devices have some sort of cpu and memory which can benefit from a restart.
3 Likes
Quotes from above review:
One of the more controversial reasons JPLAY was developed was to improve sound quality. People are quick to say that the television remote doesn’t change the picture quality, so a UPnP control point also can’t change the sound quality. However, this is a very primitive point of view and shows a lack of understanding of how applications and networks function. Most apps are “constantly” communicating with either a mothership, the network, online services, or a host of other entities.
JPLAY has always said to improve sound quality a main goal was to eliminate or majorly reduce network activity from the control point. I wanted to see how successful the JPLAY team was at this, so I setup a test and recorded video for everyone to see. Within the JPLAY settings is an option called Update Time. This is how often the control point application queries the audio endpoint for a progress update. The allowable settings are between 1 and 12,000.
1 Like
In short: “Leave the UPnP renderer (player) alone.”
It appears that the streaming communication between a UPnP media server and a UPnP renderer closely resembles the process used by online streaming services. The UPnP media server acts as the HTTP server, while the UPnP renderer functions as the HTTP client. Playing music primarily involves HTTP streaming rather than UPnP communication. As a result, this leads to improved SQ, similar to UPnP renderers streaming music tracks directly from online servers of platforms like TIDAL and Qobuz. To further enhance sound quality, consider using a UPnP control point (smart device) with a huge amount of processing power that manages all the processing without burdening the UPnP renderer, ideally doing so with minimal queries.
Anyone feel free to jump in here 
I’ve been experimenting lately after my Roon/HQP configuration had a bit of an audio meltdown and was sounding a bit rough. I tried MiniDLNA with mconnect and felt it was an improvement on what I had before. Since then Roon/HQP has returned to what it was (no cause identified) but MiniDNLA still bests it.
I then tried Jplay, at first because the UI looked better and not realising that improved SQ was an integral part of its design. Fast forwarding a bit, MiniDLNA and Jplay is the best sounding combo in my set up and I have to say, by some margin.
Now for the hard bit (for me at least) describing the sound: First off, the overall sound presents in a way that feels much more within the limits of the system. I’ve avoided using relaxed or calm to describe this because that might suggest a lack of dynamics or too soft a sound, which isn’t the case. With MiniDLNA/Jplay I can turn up the volume to create a more realistic facsimile of the music being reproduced. Other combos make me want to turn it down as all the well known digital traits are forced to the fore.
In a bit more detail, most obvious improvements are with the top end which is now free from any shrillness or scratchiness. The low end is leaner and better defined and remains controlled when the volume increases.
More subtly, performance across the range seems more even, resulting in a more realistic soundstage and separation is a little clearer with everything holding its own space a little better.
Although I stated that I felt MiniDLNA/Jplay was better by some margin, I mean that in audiophile terms where we tend to wax lyrical about small gains. I could live with any of the combos but as Antipodes servers give us the choice, we can experiment and pick the one we like best.
1 Like
Do you get a delay when you start playback or skip a track when using hqp/roon?
Yes, around 3 or 4 seconds… even when HQP is not upsampling.
1 Like
There are protocols for streaming over HTTP, but they’ve emerged more recently to support mobile usage. I believe the UPnP uses UDP on top of HTTP to actually send the stream. HTTP itself is used to post to or get data from a server. This can include servers on the internet, which in the case of Roon can be a lot of traffic.
Having to respond to repeated bombardment over HTTP can cause a lot of activity on the server. Reducing this will usually result in improved sound quality.
One thing that I believe helps to give Squeeze an advantage is that closing the browser tab display the Material interface brings an end to the traffic providing that no other controller is running. Apps are usually kept alive so JPlay offering control of how chatty the app behaves should give it a sound quality advantage. But having no control app running should be even better.
I didn’t like that JPlay didn’t have gapless turned on by default, but from a sound quality perspective that was the right move. Gapless support requires the next song be retrieved while the current song is playing. That additional activity can degrade sound quality. I’m not sure that alone will be noticeable on most systems but when combined with other attempts to reduce activity, benefits would be more likely to be heard.
2 Likes
…and don’t forget also, potential benefits like when increasing Update Time. Marcin (JPLAY founder) prefers the maximum value of 12000. YMMV but the idea behind it exactly addresses this chattiness of control apps.
To quote Marcin regarding JPLAY for IOS from Audiophile Style:
Still, despite reducing network traffic to absolute minimum, the impact of network can be heard simply, because the network connection is active whether we send data or not. (…)
In my personal system I’ve now eliminated wifi completely. I’m running the JPLAY app on M2 Mac Mini which is connected to XACT S1 (switch) and the XACT S1 (switch) is conected to XACT S1 (server/player).
Even though there is almost zero network traffic (few bytes of data every couple seconds) the difference in sound between control point connected by wifi and wired ethernet is clearly audible. And adding a network switch is, again, clearly audible which is quite significant in terms of SQ - even for files played directly on the server which is baffling.Everything in hi-fi ‘plays’ even though we may have no answers why.
But when I asked him why, to listen to music, he would do that from behind his desk (M2 Mac Mini), he replied:
My listening ritual is on a sofa with an iPad in my hands. To be able to ‘see’ XACT S1 in the home network again I have to connect one of the ethernet cables back.
It’s just for more critical listening I isolate my ‘audio’ network from my main home network by disconnecting the cable going to router from XACT S1 switch (this is possible once XACT S1 server/player and Mac get their IPs assigned by DHCP server running on the main router). (…) It’s not a big inconvenience.
(link update edit)
Thanks for sharing the insightful quotes. It reminded me that I should eventually experiment with creating a segmented network for just audio.
1 Like
Marcin and Emile’s efforts combined definitely seem to point in separated and isolated audio networks.
I imagine a completely isolated audio network will deliver a nice sound quality improvement. I’m not ready to go there yet or maybe ever. I’m happy with where things are now with my audio system.
My system reached a new level of realism/transparency/believability recently thanks to the addition of a single cable. I reported my findings here and here.
Last night I returned to the same reference track that caused the tears to flow as reported in my second link. This time around, a not-yet-burned-in extension cable was used. In addition, instead of Squeeze, I tried both HQPlayer and MPD/MinimServer/JPlay.
Neither of those two other applications were able to match my recollection of what I heard from Squeeze that night. Both these apps smeared the vocals and that harmed believability. She just didn’t sound like she was in the room with me as she was with Squeeze. The not yet burned in cable could be a factor in this, but I think it’s more of the case of the software as this was consistent with past comparisons.
What was surprising though was that I found myself unable to fight back the tears when listening to her sing to me through MPD/MinimServer/JPlay. HQPlayer did not bring about this reaction. It’s tough to put into words exactly why this song moved me so much with Squeeze and MPD but not HQPlayer, but I think it came down to the reproduction microdynamics. The singer’s inflections are more realistically reproduced when microdynamics are better conveyed. HQPlayer, even with processing set to minimum as possible. I should caution though that emotions are never predictable, so on another listen HQPlayer might also be able to bring me to tears with this song. I intend to repeat this in a few weeks after that DC extension cable has more time on it.
Kenny, thanks for posting those thoughts. By coincidence I have this morning been comparing
Squeeze + Squeeze
MPD + MinimServer (controlled by the JPLAY ios app)
Squeeze + Roon
The track I was using was the first track ‘Biscuits’ on the Fink live album ‘Wheels Beneath My Feet’.
With Squeeze + Squeeze I find the track to be absolutely immersive and I just want to keep on listening. The lower frequencies are tight and deep, the percussive notes are clear and detailed and the voice is full of character.
MPD + MinimServer (controlled by the JPLAY ios app) is much softer and everything just seems a bit smeared to me from the low notes to lovely cymbals that are now rather less real and ‘live’. The bass notes are softer and less detailed. The whole track is less immersive and stops being as enjoyable. With Squeeze + Squeeze I found myself letting the music run on to the second track. Not so with MPD + MinimServer, I frequently stopped the track less than half way through having heard enough.
Squeeze + Roon sounded better than MPD + MinimServer to my ears but was not as good as Squeeze + Squeeze, not by quite a margin. The criticisms were more or less the same as for MPD + MS but I felt them to be not quite so prominent. However it was still nowhere near as enjoyable and as immersive as Squeeze + Squeeze. With my Oladra I see no point in having a near 30k streamer and then degrading its sound by listening via inferior apps so until I find anything different it is back to Squeeze + Squeeze for me.
As an aside, a big additional factor in favour of Squeeze + Squeeze is that I can listen to the BBC Sounds app with this combination. This not only gives access to live Radio 3 but also gives access to the past BBC library of broadcast music and speech programmes. In fact I am listening to Radio 3 now using Squeeze + Squeeze. I frequently listen to previously broadcast live concerts on the BBC Sounds app.
1 Like
“Degrading its sound”: yup that’s exactly what the other software options do relative to Squeeze + Squeeze.
You guys make me feel I must be missing something, are there any Squeeze settings you know of that impact SQ that should be set a particular way?
Cult of Squeeze. Long live roon.
I think maybe Nick and I are a bit more sensitive to smearing and harm done to the time domain. We both upscale our music using filters that significantly reduce smearing and significantly improve timing precision (though we use different technologies to achieve that). I think that leaves us both with systems that spotlight the harm these other apps can do. I made tremendous progress this year with improving my system’s transparency but I think it’s now made it impossible for me to really enjoy listening to music unless I use Squeeze.
1 Like