I Declared Victory. I was SOOO Wrong…

[James]

Check Out Metageek Software: https://lmg.gg/qZDu2

Buy a Wi-Spy Wi-Fi Spectrum Analyzer: https://lmg.gg/OrxHy

Buy a Edimax Wi-Fi Adapter Dongle: https://geni.us/EGbdaH

Check Out Wireshark: https://lmg.gg/Dk9uN

Check out a Wireshark Tutorial: https://lmg.gg/dGWDf

 

Thank you for watching! It can become challenging to communicate online and work from home. With the help of Grammarly, the right tone can move any projects forward! Sign up for an account and get 20% off Grammarly Premium: https://grammarly.com/LTT

 

 

 

[ahuckphin]

at what point will Linus install curtains/blind for all his windows to save editors time in censoring the sunlight/outdoors

[darwin006]

Turns out wired is just better....

[cmndr]

9 minutes ago, darwin006 said:

Turns out wired is just better....

Most people don't get sponsorships to install EVERYTHING imaginable under the sun in a large house. 

[Bitter]

Wires...who knew! Should be able to snake them behind the trim at the bottom of the walls then up the inside of the wall out to a plate for the speaker to connect to. What's nice about wires is they don't care what brand the speakers are so you can upgrade without changing everything out at once.

[mynameisjuan]

Without having the capture to review and going off screenshots, I don't believe the interpretation that the beacon is bouncing between channels "more than once a millisecond" (12:42) is the correct assumption. Looking at what's presented, I think Sony is utilizing multiple beacons and balancing the speakers over these channels. AKA, it's not changing channels multiple times per millisecond, just some speakers are using channels 44 and some using 48 during some periods of time.

 

While my career is in network engineering, wireless engineering is it's own subset and of which my depth of knowledge is limited when it comes to the 802.11 spec, so take that as you may. But looking at Wireshark's output you can see that the two MCAST streams are being flooded on both channels in this capture simultaneously. "IPv4mcast_4[2,3]" is the multicast MAC address for a dst. IP of [224-239].0.0.[66,67].

 

I would not imagine a scenario where each packet of a flow would average a packet of data every 1-2 ms to then send one, quickly change channels following another packet within a ms.

 

 

Later on in another glance you can also see the Seq. #s which you can also see are equal for a given flow on each channel at the same time. Though I would think the Seq. # are independent per AP as it's part of the 802.11 header and just down to flooding occurring over both APs.

 

Finally, the spectral shows the individual channels with the SSID covering both. Now what is not certain, is the BSSID which I would suspect if @LinusTech was able to go back and check in Wireshark, would see two different BSSIDs in the headers.

 

While there is clearly an observable issue, even I would not have the confidence to call out Sony and say they would have to re-engineer their hardware over these tests and captures. PCAPs don't lie, but their results can be meaningless unless you know how the client software is handling or is expecting the data

[coonwhiz]

Why do shots with windows in the background if it's going to result in 50% of the shot being blurred like at 1:40?

[brandishwar]

THANK YOU, THANK YOU, THANK YOU for not just the content warning about the flickering colors and text with the Wireshark demo, but also the audio chime for when it was safe to look at the video again. Definitely keep doing that in the future! It's something I wish more content creators took into account.

[Citizen5391]

Small issue with the description, your lorem ipsum was not replaced!

Dunno who to ping for such an issue, but its there.

[Vishera]

21 minutes ago, brandishwar said:

THANK YOU, THANK YOU, THANK YOU for not just the content warning about the flickering colors and text with the Wireshark demo, but also the audio chime for when it was safe to look at the video again. Definitely keep doing that in the future! It's something I wish more content creators took into account.

Was useful for me as well, though there was a recent video that did have flickering without warning:

My original reply to it:

On 1/16/2023 at 5:26 AM, Vishera said:

@LinusTech @Plouffe What's going on with the "No Fans" flashing prompts in the video?

I suffer from migraines that are in the same family as epilepsy, flashing like in the video can give me nasty migraines that will disable me for a few hours, some times for the entire day.

So it's important that the flashing issue will be addressed, especially for people with epilepsy.

 

[Vishera]

30 minutes ago, coonwhiz said:

Why do shots with windows in the background if it's going to result in 50% of the shot being blurred like at 1:40?

Creeps and other unwanted people can use that to locate where Linus's house is located.

You would be surprised but people do those kind of things.

[harry.cpp]

Surprisingly, and also interestingly enough, all the xbox wireless headsets that connect directly to the Xbox have the same issue with audio randomly cutting off, or even worse, headphones fully disconnecting and reconnecting.

[Blade654]

5 hours ago, mynameisjuan said:

Without having the capture to review and going off screenshots, I don't believe the interpretation that the beacon is bouncing between channels "more than once a millisecond" (12:42) is the correct assumption. Looking at what's presented, I think Sony is utilizing multiple beacons and balancing the speakers over these channels. AKA, it's not changing channels multiple times per millisecond, just some speakers are using channels 44 and some using 48 during some periods of time.

 

While my career is in network engineering, wireless engineering is it's own subset and of which my depth of knowledge is limited when it comes to the 802.11 spec, so take that as you may. But looking at Wireshark's output you can see that the two MCAST streams are being flooded on both channels in this capture simultaneously. "IPv4mcast_4[2,3]" is the multicast MAC address for a dst. IP of [224-239].0.0.[66,67].

 

The behavior and capture look fairly normal. The AP is using a 40mhz channel meanwhile Linus is capturing the 2 20mhz channels that make up the 40mhz channel(Channel Bonding). Normally the packets would be split between the 2 channels but I'm guessing the duplicate packets are from the stations (stereos) only supporting 20mhz and thus the AP has to duplicate the packets between the 2 channels to ensure the stations see everything regardless of what 20mhz channel they are connected to.

[dogwitch]

1 thing sony known for. Garbage tier wifi chips.

since ps3 era till now.

 

[mynameisjuan]

10 minutes ago, Blade654 said:

The capture looks fairly normal. The AP is using a 40mhz channel meanwhile Linus is capturing the 2 20mhz channels that make up the 40mhz channel(Channel Bonding). Normally the packets would be split between the 2 channels but I'm guessing the duplicate packets are from the stations (stereos) only supporting 20mhz and thus the AP has to duplicate the packets between the 2 channels to ensure the stations see everything regardless of what 20mhz channel they are connected to.

While I am familiar with channel bonding, I was not aware what it looked like in a spectral graph (assumed it utilized the entire range) but looking at some examples, it does appear to be as expected like you said. This would also clear up my assumption with the duplicate frames on the individual channels.

 

However, while my explanation was incorrect, overall it still aligns with my initial argument that the stations are not rapidly bouncing between the channels yet the stations are balancing individually between the two channels. When Linus asked why is auto channel switching being done, there is nothing else on that frequency...well yes there is, the individual speakers. During low traffic/signal noise periods, they may all switch to the same channel, then during high traffic/signal noise periods, some change to the least congested of the two channels which is reflected in the captures.

 

Nothing that was presented stood out as a clear issue. I'd be more interested in going through the PCAP to look at instances containing bad ACK frames and what was going on at that time. If I recall, when he demonstrated the issue, audio dropped out entirely for all speakers vs what would be expected in the case of the stations ACS, I would expect individual speakers experiencing an issue briefly.

 

Either way, I would not push back and say this is clearly an engineering problem without concrete evidence, I would lead more towards software or another external source.

[Blade654]

Also worth mentioning, the AP is sending 802.11 broadcast/multicast frames which are not acknowledged or retried. This has the benefit of allowing the speakers to mainly listen and not generate that much traffic but the major downside is the packet loss. That 23% retries could be 23% packet loss. Although of course the radios, antennas, tx power, etc in the AP are probably different from the speakers so the packet loss isn't necessarily the same.

3 hours ago, mynameisjuan said:

 When Linus asked why is auto channel switching being done, there is nothing else on that frequency...well yes there is, the individual speakers. During low traffic/signal noise periods, they may all switch to the same channel, then during high traffic/signal noise periods, some change to the least congested of the two channels which is reflected in the captures.

I don't think that's what the pcap is showing. Maybe the stereos are switching channels but I don't think you could tell that from the limited data that was shared.

 

Also I think another issue with this pcap is that it's 3 wifi adapters and we don't know how they are hopping channels. It's possible that at 10:52 the adapters are all spread out hopping wildly different channels but at 12:44 they are more synced up and hopping closer channels which would explain why you see 44 and 48 interleaved.

Edit:

The Tonic user-guide, https://support.metageek.com/hc/en-us/articles/4545963398167-Tonic-User-Guide, outlines it's hopping behavior at the bottom. The behavior depends on what view you have selected. It seems like at 10:52 they had the channels view or clients view open so only 1 adapter was hopping channels 36-116 and at 12:44 they had the bssid view or client view open which uses 1 adapter to monitor the current channel and another to hop the remaining 5ghz channels.

 

3 hours ago, mynameisjuan said:

Either way, I would not push back and say this is clearly an engineering problem without concrete evidence, I would lead more towards software or another external source.

Ya, I don't think there is enough information to determine where the problem lies but I'm leaning towards multicast + dropped packets or a firmware issue. Maybe a bit of both since I don't think packet loss would explain all of the speakers dropping out. 

Edited January 29, 2023 by Blade654
Looked at the tonic user guide

[bluehawk]

I like how this followed his video about going wired for those room sensors. 

[wanderingfool2]

I could be wrong, and I hope they did so in the attempt to trouble shoot it but for myself I'd be pulling out the base device, putting it lets say in front of the TV (leaving the subwoofer in the cabinet) and then trying the system to see if anything is improved. (Like the manual says you should)

 

Especially if you are getting retransmission's of 23%...that to me screams that there is something going on with the wireless communication.  I can't help but wonder if the very close proximity to essentially a moving magnet could be creating an issue or being in the enclosed cabinet is creating an issue (I say this as someone who had a cheaper speaker system that would hum when some standing next to the speaker received a text).

 

So a few things I would ask to see if it was done by @LinusTech.
- Following the user guides suggestion of putting it on top of a shelf, instead of inside the shelf done?

- Was it moved further away from the subwoofer (or having the subwoofer removed) to see if there was any interference with it so nearby.

- Was there an option to switch it to 2.4 ghz?  [the manual for the A9 says there should be a way to under the wireless setting in the options menu]

 

 

Side-note, Is the mesh covering on the cabinet a metal mesh?

 

Also there is a warning about putting metal objects near the speaker (to prevent interference), so I'd imagine putting a magnet and a coil of wire that is being energized might not be the best thing.

 

https://helpguide.sony.net/ht/a9/v1/en/contents/TP1000196690.html

Quote

Do not install the control box in the inner position of a shelf. Install the control box in or on a shelf so that the front panel display is not hidden.
Move the control box, change the direction of the control box, or install the control box far away from a TV.
Interference of wireless bandwidth between the network connection and speaker connection may occur. Change the wireless LAN connection of the speaker system to the wired LAN connection, or switch the band of the wireless LAN connection to 2.4 GHz.

I know there is a mesh paneling on the front, but depending on what the paneling is (like if it's metal) it could create issues.

[Techrocket9]

The callout about broadcast traffic was just straight wrong.

 

In wireless networking, every transmit on a given frequency shows up on the antenna of every receiver in range. That's an RF property -- doesn't matter what your higher-level protocols want.

 

Broadcast/multicast vs unicast signaling is just a difference in the "to:" label on the transmission. A unicast message says, "Unless your mac address is xx:xx:xx:xx:xx:xx, ignore this message". An open broadcast says, "Everyone should pay attention to this message".

 

That filtering happens after receive though (and it's trusting the client to honor it -- hence why your WiFi sniffer works), so the same airtime is spent whether the message is unicast or broadcast.

 

Logically, WiFi devices using the same frequency are equivalent to devices on the same ethernet hub (hub, not switch). When you ponder the implications of that, the fact that WiFi is terrible starts to make intuitive sense.

[hormelspam23]

Anyone know the background music from 1:39 to 4:22? I'm pretty sure it's a Hyper Potions track but I'm confused because there was no music credit in the description. 

[DJKaotica]

Hey @LinusTech and @James,

 

I posted this on the YouTube video but figured it'd probably get more traction here.

 

I am not affiliated with AnthemAV, but I do own a receiver from them.  (caveat: when I first set up my AVR I was still setting up other stuff around it / determining how the room would be setup, so I never did use their room correction technology)

 

Would love to see you do a collab or at least review of their products.  They are a Canadian company which makes Audio Video Receivers, Processors, Amplifiers, etc.  They have a technology, Anthem Room Correction (ARC), which it sounds like would be right up your alley for a follow up video, or Lab work, since Linus sounded super interested in the room correction technology.

[GodAtum]

TLDR the sony is a piece of shit

[maplepants]

On 1/28/2023 at 9:57 PM, mynameisjuan said:

going off screenshots, I don't believe the interpretation that the beacon is bouncing between channels "more than once a millisecond" (12:42) is the correct assumption. Looking at what's presented, I think Sony is utilizing multiple beacons and balancing the speakers over these channels. AKA, it's not changing channels multiple times per millisecond, just some speakers are using channels 44 and some using 48 during some periods of time.

I don't have the same hardware engineering background, but that's how I read the screenshots too. I was expecting him to bring this up at some point, but then he just didn't.

 

The saga here isn't done (obviously). So hopefully they tackle this in the next instalment. 

[JEskandari]

wonder why bother with wireless when your speaker are pretty much unmovable because of their power cables that is fixed because of your cable management which is fixed .

I assume they have fiber optic connection , why not just bring and electrician and ask him to pass some fiber optic cable between those speakers , and sound system . problem solved , no interference , better sound quality , no delay no drop ...

and honestly don't kid yourself your setup is not like you can move it anywhere in the room after first setup its pretty much fixed

if you need for some reason to have sound in some other place for a period of time just buy some anker speaker they hook with each other (if I'm not wrong you can hook together more than 100 of them and they have 12 hour of battery life , so you can have the sound any place in your home when there is party . if that is so important  but to be honest I never had good experience with wireless audio even on a simple headphone and rather use a wired one)

[James]

On 1/29/2023 at 11:28 PM, DJKaotica said:

Hey @LinusTech and @James,

 

I posted this on the YouTube video but figured it'd probably get more traction here.

 

I am not affiliated with AnthemAV, but I do own a receiver from them.  (caveat: when I first set up my AVR I was still setting up other stuff around it / determining how the room would be setup, so I never did use their room correction technology)

 

Would love to see you do a collab or at least review of their products.  They are a Canadian company which makes Audio Video Receivers, Processors, Amplifiers, etc.  They have a technology, Anthem Room Correction (ARC), which it sounds like would be right up your alley for a follow up video, or Lab work, since Linus sounded super interested in the room correction technology.

 

You can see us use their room EQ software, Anthem ARC, in my Intel Home Upgrade