Could Wireless Be Better or Worse Than Wired Ethernet?

[The Computer Audiophile]

Hi Guys, I almost always use wired Ethernet because it just works every time. However, I've always wondered what the possible and/or realized issues are with using wireless, mainly, is RF an issue with sensitive audio components?

 

Question: Has anyone seen tests of audio equipment sensitivity to WiFi and any effects on the output of the audio devices?

[plissken]

I haven't seen any measurements but I do know that one way to test would be to capture output into an ADC and airplane mode on/off during capture. Post the capture for download and let people evaluation and report their findings.

 

I did this years ago and posted a video here with Airplane mode on and music still playing (all over 802.11g).  AP's are often far enough away and equipment shielded that I've never seen it impact. As I type this I'm listening to Rachmaninoff in HD Naxos Stream over my $79 Pi3 B+ wireless to My DC-1 over JBL 308P MKII and matching JBL 310S subs.

 

Sounds phenomenal given the entire setup represents $1400 in spend. These JBL's are just killer. Shocking what they were able to achieve. Their only Achilles heal is breaking up when playing loud. But I have other speakers for that. 

 

Wireless are three $56 TP-Link Omada AC1350's with the Omada Controller for roaming and preventing client stick. Routinely get 300Mbps.  More than enough for any 24/192, 32/384, DSD 512.

[The Computer Audiophile]

Yeah, bandwidth isn't an issue at all. 

 

I'm thinking that if people use an access point behind their audio systems, where all the cables are, this could be an issue. I know this is also a popular place to put the AP based on my experience with other people's systems. People don't have wired Ethernet to their audio rack, so they use an AP behind it and it extends to the main AP. This is kind of a worst case scenario, but it would be very interesting to see if it matters. 

[plissken]

23 minutes ago, The Computer Audiophile said:

Yeah, bandwidth isn't an issue at all. 

 

I'm thinking that if people use an access point behind their audio systems, where all the cables are, this could be an issue. I know this is also a popular place to put the AP based on my experience with other people's systems. People don't have wired Ethernet to their audio rack, so they use an AP behind it and it extends to the main AP. This is kind of a worst case scenario, but it would be very interesting to see if it matters. 

 

I think we just had the 75 percenters talk about this privately :-) My AP's are tucked away in closets out of site and are PoE. I run them at -63dB to the supplicant location.  For VHD (Very High Density) designs it's an AP every 1500 sq foot. For low density office we spec at 3000 and 5000 sq foot. For low density warehouse it's 10,000. We service limit to 150 active connections per AP (this 1/2 the service rate of the AP's we deploy commercially).  Our average is something like 30 concurrent clients. This includes Wifi calling with E911 and emergency paging services.

[R1200CL]

55 minutes ago, The Computer Audiophile said:

Question: Has anyone seen tests of audio equipment sensitivity to WiFi and any effects on the output of the audio devices?

It’s attempting to ask why Sonore never added a wireless to the Rendu’s 😀

Or why other well know quality brads have done. Not sure if you can switch WiFi of. But thinking of some brands offering you to switch og leds and display for best SQ....., is it just a gimmick. 

 

Haven’t even you had equipment under test that offered both wireless and ethernet (or even fiber) ?

Shouldn't possible affects be possible to measure? How ? Challenge for Amir 😀

Now, one way people can test this is to place their wireless router next to the audio equipment and connect and see what’s happening to SQ. Well actually they need two I think. One acting as router. The other as access point. 
 

Or is your question purely EMI/RFI related and has nothing to do with WiFi connected to streaming ?
 

EDIT

Seems you answered while I was typing 

Also powerline adopters could be included in this topic. 

[The Computer Audiophile]

My questions are really about everything related to WiFi and HiFi. Sure RF seems to be a bad thing next to audio hear, but also what about components with built-in WiFi? I had a Devialet Expert series set of amps here many years ago and when it wasn't connected to a wireless network, the amps would place the wifi card and antennas internally into blast mode to search for networks. This was easily audible to even the most unlearned listener. 

[plissken]

5 minutes ago, The Computer Audiophile said:

the amps would place the wifi card and antennas internally into blast mode to search for networks. This was easily audible to even the most unlearned listener. 

 

Yikes. Amir blew one up in power sine testing if that's any indication. What people need to know is there are best practices when it comes to WiFi design.

 

If anyone wants they can request a demo of Ekahau and import a diagram of their house and play around with WiFi AP placement, power settings, radiation patterns, channel planning etc.

[R1200CL]

@plissken

My understanding is that you in any case just pull the wire or disconnect anything related to ethernet. 😂

(I just couldn’t resist).

 

I think like this. Whatever standards says about WiFi transmission, or 4G and now 5G and IoT, it may not apply to audio. Standards of hospitals etc may not fit audiophile requirements. And those audiophile requirements isn’t written down as a standard. More sometimes close to marketing BS in worst case. 

Some people, extremely few, react on cellphones. Whatever a standard says. 
I do believe mostly when people say they hear a difference, even not possible to measure.

 

This tread can be useful for those that cannot pull a ethernet cable.

 

Others goes to the extreme and have a separate network for audio and says it sounds better. Now I would like an explanation on that one. 
 

 

[plissken]

1 minute ago, R1200CL said:

may not fit audiophile requirements. Not that those requirements is written down as a standard.

 

That's an interesting perspective. RF design best practices should apply rather handily to audio given that it's not real-time and delay insensitive. Just need to meet the required throughput.

 

In general for WiFi calling I need to hit -63 to -68 but this is supplicant dependent. For Web traffic -72dB.  I try not to run more than 18 watts on an AP so if I have a failure self healing can take place and other AP's have headroom to up their power to fill the hole.

[R1200CL]

2 minutes ago, plissken said:

I try not to run more than 18 watts on an AP so if I have a failure self healing can take place and other AP's have headroom to up their power to fill the hole.

How many of us do you think understand this, or even are able to control their AP’s power settings ?

People is highly likely to upgrade to some sort mesh and WiFi 6 (or AIMesh), where I think it’s  less fiddling, and more automatic settings. 

 

If you could sometimes be a bit more educational in your post instead of being (showing off) superior in knowledge, you would helps us all a lot. Sometimes your language/posts is just to high tech to follow for most of us. It doesn’t help to be right or correct if you’re the only one. Graveyard is full of them. 

This isn’t meant as criticism of you, just a friendly request/advise. So excuse, if my language skills in framing things may sounds differently. 
 

[plissken]

15 minutes ago, R1200CL said:

How many of us do you think understand this, or even are able to control their AP’s power settings

It would be the same people that understand -/+ 3dB when it comes to their amplifier is a halving/doubling of power.

 

@jabbr has an entire thread based on optics and it's a learn/teach sort of thing. Every AP I've installed has a GUI. Doesn't matter if it's a $56 TP-Link or $1200 Cisco.

[R1200CL]

We could add PoE to this tread. Although it as been answered before when I raised the question.

 

32 minutes ago, plissken said:

RF design best practices should apply rather handily to audio given that it's not real-time and delay insensitive. Just need to meet the required throughput.

And this may be the key, how immune is your other equipment to various interference. 
 

I think, and probably you can confirm as well, WiFi equipment, and equal us under much more restrictive tests and regulations and requirements than any other equipment, so we should be quite sure that equipment meets standards. 
 

Not so with your audio items. 
 

So maybe the title of this tread is totally wrong 😀
It’s not about your WiFi, but rather about your equipment quality to resist EMI/RFI. 

[R1200CL]

16 minutes ago, plissken said:

Doesn't matter if it's a $56 TP-Link or $1200 Cisco.

It’s matters. A lot. But we can agree to disagree. Let’s give time to others to chime in. 

I have TP-link, Cisco, and Asus here. We can dive into various settings later. 
 

I guess people is more into super bowl tonight than AS treads. 😀

[plissken]

37 minutes ago, R1200CL said:

And this may be the key, how immune is your other equipment to various interference. 

 

Look for the FCC, CE, or other laboratory marks.

[Calvin & Hobbes]

I got a Pro-ject Stream Box S2 Ultra that I started using in WiFi mode. I got a WiFi extender to create an Ethernet access point and switched to an Ethernet connection and turned off the WiFi connection. Of all of the tweaks that I tried (USB cable, Ethernet cable, Jitterbug noise filter), switching to an Ethernet connection and turning off WiFi produced the largest beneficial effect on sound quality 

 

These were some of the sound quality improvements that I heard: Streaming via Ethernet has more presence & clearer. Better pace than when streaming via WiFi. Like a veil has been lifted off of the music.

[lmitche]

My listening room is on the floor above the networking gear room with no clean way to run cable between the floors. Therefore I am stuck with a wireless connection to my music server.  Here are a few things I have learned:

 

• 2.4ghz radios are noisier then 5ghz radios. I now only use 5ghz gear that allows disablement of any 2.4 ghz radios.
• Distance and isolation between the DAC and radio gear matters. My 5ghz USB wifi adapter is 15 feet from my DAC at the long end of a fiber run.
• Qobuz/Roon runs fine on a 867mbps connection. I can't hear a difference from local files.
• Much consumer network gear, not all, is shipped with stock 12 volt SMPSs, but will run fine and sound better at 5 volts.

Important - Stay away from any gear that includes embedded wifi unless the radios can be powered down.

[audiobomber]

For the past couple of months I've had an exaSound Playpoint Mark 1 and e32 DAC in my main system. Due to the mainly concrete construction of the house, I cannot reach this system with an ethernet cable. The Playpoint has built-in wi-fi capability, so I was using that. I had intended to test the onboard wi-fi vs. a "wi-fi to ethernet bridge" extender, but never got around to it because the system sounded so much better than the previous incarnation. This thread motivated me to finally do a test,

 

I played a few very familiar tunes using the Playpoint's onboard wi-fi, then connected to the Playpoint via Audioquest Pearl CAT7 ethernet from a TP-Link RE450 range extender, powered by 120V from a Furman power conditioner. The Playpoint's ethernet connection automatically disables wi-fi. The RE450 is configured to receive 2.4 0r 5 GHz signal. Wi-fi broadcast is disabled, the single ethernet port transmits 5GHz only (which I assume doesn't matter with a single client in bridge mode).

 

I found similarly pleasing results to @Calvin & Hobbes. The system lost a slight background hashiness that I had not noticed previously. The blacker background brought better instrument separation and imaging, smoother highs and a generally richer sound.

 

I need the RE450 for my secondary system, so I replaced it with an RE200 and listened again. The RE200 is a cheaper, simpler device; does not have capability to disable wi-fi network broadcast, and cannot follow the router when it selects a narrow channel band. The sonic benefits of ethernet vs. onboard wi-fi evaporated with this change.

 

I've ordered a TP-Link TL-WA901ND Wireless Access Point. I will power it with a Zero-Zone 12V 2A linear power supply, which from previous experience using a TP-Link RE580D, will provide a further sonic upgrade over the supplied wall wart.

 

The above does not fully answer @The Computer Audiophile's original question. I still wonder, how does a system with a wi-fi to ethernet bridge compare to a system connected using cables only? I suspect it is system dependent, i.e., how good is the implementation? A basic system will be limited, a system with a better switch and/or bridge and upgraded PSU will perform at a higher level. 

 

https://www.amazon.ca/gp/product/B002YETVXC/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1

[Calvin & Hobbes]

@lmitche Great post! My experience fits with what you've found.

 

7 hours ago, lmitche said:

Important - Stay away from any gear that includes embedded wifi unless the radios can be powered down.

 

That's a great suggestion. By nature, I'm open to figuring out what sounds better, but also skeptical. I got the Pro-ject streamer precisely because it could work on Wi-Fi, but also could shut down Wi-Fi as well. My thought was that the idea that Ethernet only might sound better, but didn't want to foreclose on the option to use Wi-Fi as it is just simpler to use if it wasn't too different in sound quality. I've realized that using a Wi-Fi extender with an Ethernet output via cable to the network streamer is just as convenient was having Wi-Fi built into the streamer, but with much better sound.

[Calvin & Hobbes]

9 minutes ago, audiobomber said:

I still wonder, how does a system with a wi-fi to ethernet bridge compare to a system connected using cables only? I suspect it is system dependent, i.e., how good is the implementation?

Return on effort is a big consideration for me in audio. Even if cables only sounds better (and I think it likely would), the perceived effort of dragging 100 feet of cable through my house dissuades me from wanting to try that. I'd rather try optimizing the output from the Wi-Fi extender.

[lmitche]

The Linksys RE9000 works well, at 5 volts, as a wifi network extender. I get 1300mbps  up and down. It has two 5ghz radios low, band and high band. I use the high band radio in the room.

[Rexp]

On 2/8/2021 at 6:42 AM, plissken said:

I haven't seen any measurements but I do know that one way to test would be to capture output into an ADC and airplane mode on/off during capture. Post the capture for download and let people evaluation and report their findings.

 

I did this years ago and posted a video here with Airplane mode on and music still playing (all over 802.11g).  AP's are often far enough away and equipment shielded that I've never seen it impact. As I type this I'm listening to Rachmaninoff in HD Naxos Stream over my $79 Pi3 B+ wireless to My DC-1 over JBL 308P MKII and matching JBL 310S subs.

 

Sounds phenomenal given the entire setup represents $1400 in spend. These JBL's are just killer. Shocking what they were able to achieve. Their only Achilles heal is breaking up when playing loud. But I have other speakers for that. 

 

Wireless are three $56 TP-Link Omada AC1350's with the Omada Controller for roaming and preventing client stick. Routinely get 300Mbps.  More than enough for any 24/192, 32/384, DSD 512.

Which software and power supply does your Pi have? Thanks! 

[plissken]

1 minute ago, Rexp said:

Which software and power supply does your Pi have? Thanks! 

RopieeXL and a Samsung 3AMP cell phone charger.

[audiobomber]

6 hours ago, Calvin & Hobbes said:

Return on effort is a big consideration for me in audio. Even if cables only sounds better (and I think it likely would), the perceived effort of dragging 100 feet of cable through my house dissuades me from wanting to try that. I'd rather try optimizing the output from the Wi-Fi extender.

I got into computer audio three years ago. At the time, the prevailing opinion seemed to be that wi-fi was better than direct ethernet cable connection. That has been a topic of debate since then, and not resolved as far as I can tell. 

 

If you want to optimize your wi-fi setup, I suggest you ensure your wi-fi access point does not broadcast wi-fi, and you improve the extender's power supply.

[klaus]

So far I am useing a TPlink wlan to ethernet bridge (TPlink) powered by a Farad super 3 and go with 5m ethernet cable to my Innuos SE. In comparison to my initial approach with the TPlink close to the Innuos and powered with aPowerbank the present solution brought a significant improvement.

Question is whether sound can be further improved by changing the SMP powersupply of my Router to a liniar PS?

A further observation relates to the control of the Innuos with Ipeng and Iphone. I am 5 meters away from my equipment but can hear when disabling IPeng and also when I switch off my IPhone.

This is a small effects but somewhat disturbing.

 

[jabbr]

Interesting question.

 

I have used both fiberoptic and Wi-Fi for years. Despite spending a lot of time testing out different things/devices etc, wired Ethernet is just rock solid stable, and Wi-Fi works most of the time. For example I can stream DSD512 always and effortlessly over wired and usually over Wi-Fi but sometimes it glitches. Similarly I can always connect HQPlayer/NAA wired using IPV6 but Wi-Fi broadcasts seem glitchy for me. 
 

I use fiber where possible and Wi-Fi where convenient.
 

So “SQ”? We aren’t dealing with apples to apples do who knows how much the theoretical RF issues are affecting SQ. Who knows maybe running a special 10Mhz clock upgrade will fix it 🤷🏻‍♂️