• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Do Audiophile Network Switches Make a Difference (video)?

phoenixsong

Senior Member
Joined
Nov 17, 2018
Messages
348
Likes
263
Slightly unrelated, but does anyone know what happens to distortion and other measurements if I play two audio tracks simultaneously? Say, two youtube tabs or two different audio players (Windows Media Player and Groove or something). Also, would it matter if the sampling rate is altered under such conditions?
 

voodooless

Major Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
1,267
Likes
1,829
Location
Netherlands
Slightly unrelated, but does anyone know what happens to distortion and other measurements if I play two audio tracks simultaneously? Say, two youtube tabs or two different audio players (Windows Media Player and Groove or something). Also, would it matter if the sampling rate is altered under such conditions?
What does any of that have to do with network switches? The two sources just get mixed. Usually there is one system sample rate, everything gets converted to that.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
32,939
Likes
110,889
Location
Seattle Area
Thread Starter #187
Slightly unrelated, but does anyone know what happens to distortion and other measurements if I play two audio tracks simultaneously? Say, two youtube tabs or two different audio players (Windows Media Player and Groove or something). Also, would it matter if the sampling rate is altered under such conditions?
Windows audio stack resamples every audio source into one constant (set in Advanced properties in your sound card control panel), converts the streams to floating point, mixes them, and then converts back to integer with dither and sends it out to the sound card. So yes, the bits are transformed. What impact it has is complex but results are very good most of the time.
 

Billy Budapest

Addicted to Fun and Learning
Forum Donor
Joined
Oct 11, 2019
Messages
672
Likes
754
This is a good watch. “You don’t know what you don’t know” accurately describes all of Hans Beekhuyzen’s videos, not just the ridiculous one torn apart here.

 
Joined
Jan 27, 2021
Messages
12
Likes
5
Thanks Amirm!
Almost believe & buy this one https://nuprimeaudio.com/product/omnia-sw-8/?v=69e1aafeccc5 , as it claims that jitter is much lower (396 ps) against others making at the level of 705 ps though I don't even sure I can hear the difference. Now I know that we aren't listen the signal of a Network Switch but a DAC, then whatever Jitter or noise it can reduce is just white elephant~~~ Appreciate your scientific lecture again!
 

Billy Budapest

Addicted to Fun and Learning
Forum Donor
Joined
Oct 11, 2019
Messages
672
Likes
754
Thanks Amirm!
Almost believe & buy this one https://nuprimeaudio.com/product/omnia-sw-8/?v=69e1aafeccc5 , as it claims that jitter is much lower (396 ps) against others making at the level of 705 ps though I don't even sure I can hear the difference. Now I know that we aren't listen the signal of a Network Switch but a DAC, then whatever Jitter or noise it can reduce is just white elephant~~~ Appreciate your scientific lecture again!
It might or might not do what it says, but as Amir demonstrated, noise and jitter on an Ethernet network do not matter as long as the cables and switches are performing to spec. Any dropped packets will be re-sent, and noise and jitter are not issues as they are dealt with at the input of the music server and/or DAC.

Quite simply, “audiophile” network switches are a solution in search of a problem.
 

somebodyelse

Major Contributor
Joined
Dec 5, 2018
Messages
2,025
Likes
1,487
Any dropped packets will be re-sent
People keep trotting that one out without looking at the details. It depends on which streaming protocol you're using. For protocols using TCP it will happen, but may be too late. However many of the streaming protocols end up using RTP over UDP under the skin, and that doesn't do retransmision - it was intended for low latency use so the re-sent packet would arrive too late to be used anyway. In practice dropped packets are rare enough to be a non-issue unless you've got bigger networking problems.
 

Jinjuku

Major Contributor
Forum Donor
Joined
Feb 28, 2016
Messages
1,129
Likes
946
People keep trotting that one out without looking at the details. It depends on which streaming protocol you're using. For protocols using TCP it will happen, but may be too late. However many of the streaming protocols end up using RTP over UDP under the skin, and that doesn't do retransmision - it was intended for low latency use so the re-sent packet would arrive too late to be used anyway. In practice dropped packets are rare enough to be a non-issue unless you've got bigger networking problems.
How many music services use UDP though? I setup networks for Shoretel, Avaya, Vocera, GetWell. I setup IGMP / PIM Sparse, DSCP, most of this type of traffic uses UDP due to the realtime nature.

Music isn't realtime and the data is at rest somewhere which means we can resend it. You don't get that with voice or video conferencing because real-time is real-time.

Depending on the buffering it isn't too late as long as the error is detected and corrected at less than the buffer thresh-hold. Roon has a 10 second buffer, Tidal will buffer entire 10 minute tracks, JRiver will buffer up to a GB (entire 16/44.1 albums).

I don't think RTP is used as much as you imply. Also there is RTP/TCP.
 

Mnyb

Addicted to Fun and Learning
Forum Donor
Joined
Aug 14, 2019
Messages
956
Likes
1,204
Location
Sweden, Västerås
LMS ( the server software for squeezeboxes ) uses tcp and have 30s of buffer at 16/44 flac .

I helped archimago to try some small test some years ago regarding network and audio .

I hosted the test files on my LMS server in Sweden and he is in Canada and gave him remote access so he could both test and listen to music and browse my collection .
Lag and ping was a bit terrible and the LMS system is not made for this . It's designed for local playback on your Lan. so there whee some issues for him.

But it worked anyway and performed the same when the bits where coming in.
 

somebodyelse

Major Contributor
Joined
Dec 5, 2018
Messages
2,025
Likes
1,487
I found UPnP using RTP/UDP as transport when I was looking into using it some time back. I know it also supports TCP based transports too - perhaps it's become more common since I looked, or it was just outliers that I looked at deeply enough to find it, but UPnP seems to be in common HiFi use still.
 

Jinjuku

Major Contributor
Forum Donor
Joined
Feb 28, 2016
Messages
1,129
Likes
946
I found UPnP using RTP/UDP as transport when I was looking into using it some time back. I know it also supports TCP based transports too - perhaps it's become more common since I looked, or it was just outliers that I looked at deeply enough to find it, but UPnP seems to be in common HiFi use still.
From wiki:

UPnP assumes the network runs Internet Protocol (IP) and then leverages HTTP, on top of IP, in order to provide device/service description, actions, data transfer and eventing.
 

Similar threads

Top Bottom