The Unspoken TRRS problem

[Psittac]

So I'm going down another rabbit hole regarding the TRRS connection that pretty much all wired gaming headsets use.  I will have to admit that I've only owned three headphone/mic combo's that use this solution and only two of them were used to any extent on comms.  But the problem I'm hoping to solve would be input/output bleed to where someone on discord will hear whatever you are listening to through your headphones.  What I'm wondering about: Is it the wire, the dac/amp/onboard or the very nature of the connection that is to blame and if it is inherent how to mitigate it.  From my limited pool of exposure, I know that the Fidelio SHP9600+Neego mic offers some  bleed over due to purchasing it for a friend.  But he had bleed over previously and just figured he had speakers too before his headset broke.  Another friend uses the Sennheiser Game One and I haven't heard a peep from him on discord, perhaps he listens to low volume and I know he doesn't have a second monitor and uses his phone for additional tasks.  Another uses a separate mic on an arm.  And for myself, I used the Game one for years but remember bleed being an issue, I chalked it up to being open back.  But I switched to a mod mic and no issues.

 

My plan here is to survey everyone I talk to on comms and do tests with them, asking them to watch something on youtube or listen to something on pandora at high volumes then turn it down little by little.  I'm also in communication with Hart Audio Cables to custom make me something for my NTH-100M so that I can rule out the cable as being the issue.  Hart Audio Cables is awesome, they custom make anything and don't mark up for doing so, it's just what the cable would cost if it was a regular item.

[RONOTHAN##]

1 minute ago, Psittac said:

Another friend uses the Sennheiser Game One and I haven't heard a peep from him on discord, perhaps he listens to low volume and I know he doesn't have a second monitor and uses his phone for additional tasks

I have used that headset and it did have bleed when using the tRRS cable, but not with the standard dual ended tRS cable though. Same motherboard and DAC used, it wouldn't bleed with the dual tRS cable, while with the tRRS cable I switched to with a splitter to dual tRS it started to have bleed issues. 

 

Cable being the issue would therefore make some sense. 

[porina]

Ran into this problem with a product where I used to work. In short, the problem stems from both mic and speaker sharing a common ground connection with that connector type. Also signal levels are bigger for the speaker than the mic, so a "small" signal for the receiver and appear to be a "big" signal for the mic.

 

On the headset side, you must use separate ground wires joined only at the connector which could fractionally increase the cost on a mass produced product compared to a shared ground. If you wanted to go audiophile on it, maybe you could try reducing the coupling further between the wires but in practice this wasn't a problem.

 

There could be another problem, that is how the "ground" connection is handled on the device you plug it into. If it isn't very low impedance, you're screwed. There is no fix other than get a better device.

 

Depending on the design of the headset, there can also be a mechanical and/or acoustic path from the speaker to mic. That's a fundamental design feature and not really something a user can control.

[AAJoe]

Well, I can save you a lot of testing: Before I start, I am not an engineer, but I do spend a lot of time speaking with engineers about this kind of stuff. So this is my interpretation of what I am told, it may be inaccurate (slightly or majorly), but I am pretty sure I've got it close enough for government work.

To some degree it is absolutely true, as soon as you share a ground between the headphone and the mic (as a TRRS cable does) some sound bleeds between the lines unless they are in "perfect" balance (like, lab environment perfect). Most of the time you don't hear or notice it because there is enough resistance on the lines that the audio is essentially too low to be audible.

 

The longer the ground is shared, the more this issue compounds. The lower the overall resistance of the lines, the more bleed you get. And lastly, the greater the resistance difference between the ground and hot cables the more bleed you get.

 

This is why "crosstalk" or "echo" or "audio bleeding" comes up so much more often when you use something like IEMs with a mic line, as they have long shared cables, low line resistance, and often large differences between grounding and mic/headphones. It is specifically why the Antlion Kimura has separate mic and IEM lines all the way to the end and we just include a Y adapter; to minimize and eliminate the problem.

So, there you have it :). Hope it helps! As for removing this effect, really if the issue is the mic line transmitting the noise that offending party probably just needs a noise gate and your problem will essentially be solved for most use cases (unless its SO loud that it can be heard when they are talking, in which case... probably active noise canceling will be better). A replacement cable may help, if the cable has some balance issues, but it's probably just the device's resistance.

[Heats with Nvidia]

14 hours ago, AAJoe said:

Well, I can save you a lot of testing: Before I start, I am not an engineer, but I do spend a lot of time speaking with engineers about this kind of stuff. So this is my interpretation of what I am told, it may be inaccurate (slightly or majorly), but I am pretty sure I've got it close enough for government work.

To some degree it is absolutely true, as soon as you share a ground between the headphone and the mic (as a TRRS cable does) some sound bleeds between the lines unless they are in "perfect" balance (like, lab environment perfect). Most of the time you don't hear or notice it because there is enough resistance on the lines that the audio is essentially too low to be audible.

 

The longer the ground is shared, the more this issue compounds. The lower the overall resistance of the lines, the more bleed you get. And lastly, the greater the resistance difference between the ground and hot cables the more bleed you get.

 

This is why "crosstalk" or "echo" or "audio bleeding" comes up so much more often when you use something like IEMs with a mic line, as they have long shared cables, low line resistance, and often large differences between grounding and mic/headphones. It is specifically why the Antlion Kimura has separate mic and IEM lines all the way to the end and we just include a Y adapter; to minimize and eliminate the problem.

So, there you have it :). Hope it helps! As for removing this effect, really if the issue is the mic line transmitting the noise that offending party probably just needs a noise gate and your problem will essentially be solved for most use cases (unless its SO loud that it can be heard when they are talking, in which case... probably active noise canceling will be better). A replacement cable may help, if the cable has some balance issues, but it's probably just the device's resistance.

I think you confused high and low resistance. To be overly correct, in this case, it should also be impedance. A low impedance ground is what you really want. High Impedance cables are more susceptible to interference.

[AAJoe]

6 hours ago, Heats with Nvidia said:

I think you confused high and low resistance. To be overly correct, in this case, it should also be impedance. A low impedance ground is what you really want. High Impedance cables are more susceptible to interference.

Probably correct, yeah. Don't let the marketing guy talk about engineering terms  - I guess the way I understand it or think about it is it's more like the difference between the ground line and mic or headphone line, and that the more that difference is as a percent of the total line the bigger problem it becomes; so for high impedance headphones a small difference is no big deal to the ground, while for low impedance IEMs that same difference is a big percent.

 

Again, not an engineer, just struggling to explain it roughly how it was explained to me when I was asking "Why didn't you use a TRRS cable on the Kimura?"  

[Heats with Nvidia]

29 minutes ago, AAJoe said:

Probably correct, yeah. Don't let the marketing guy talk about engineering terms  - I guess the way I understand it or think about it is it's more like the difference between the ground line and mic or headphone line, and that the more that difference is as a percent of the total line the bigger problem it becomes; so for high impedance headphones a small difference is no big deal to the ground, while for low impedance IEMs that same difference is a big percent.

 

Again, not an engineer, just struggling to explain it roughly how it was explained to me when I was asking "Why didn't you use a TRRS cable on the Kimura?"  

You are close. I was talking about the impedance of the wires, not the phones themselfes. A higher impedance means that there is a bigger difference between the voltage of the ground connection on the TRRS connector and the ground that the microphone and the headphone drivers "see", because the wires aren`t conductive enough, in other words have a too high impedance. Also how the cable is made, makes a huge difference. When both grounds only meet at the connector, thats a lot better. When then the signal wires are also separated and shielded from each other, thats basically perfect.