Does Pyra have a microphone?

[Grench]

The headset mic is also connected to the codec. Just disable it in the device tree or ALSA.

I think you may have misunderstood me - I didn't want to disable the headset mic - though the idea would do that too.

Are the case mic and headset mic fully independent paths to the SoC (both could theoretically run at once recording independent feeds)?
or
Do they share one feed? I.e.: Does plugging in the headset mic switch the mic feed from the case mic to the headset jack?

If the 1st, wish we could have included a 2nd case mic for stereo recording.

If the 2nd where they use the same connection path is true - and if plugging something into the headset jack switches the feed away from the case mic - then plugging a null-device-plug into the audio/mic/headset jack might effectively disable both mics.

Personally, the Pyra is way low on the list of devices I'm concerned about for spying on me.

 

[Xcl4m4t10n]

Does Pyra have puss puss?

 

[Letalis Sonus]

it will be much easier to mute microphone from pulseaudio.

You are very bad at being a tin foil head if you seriously consider trusting a piece of software with a network interface, only ALSA is able to satisfy the paranoia!!111

 

[ToastBucket]

I think you may have misunderstood me - I didn't want to disable the headset mic - though the idea would do that too.

Are the case mic and headset mic fully independent paths to the SoC (both could theoretically run at once recording independent feeds)?
or
Do they share one feed? I.e.: Does plugging in the headset mic switch the mic feed from the case mic to the headset jack?

If the 1st, wish we could have included a 2nd case mic for stereo recording.

If the 2nd where they use the same connection path is true - and if plugging something into the headset jack switches the feed away from the case mic - then plugging a null-device-plug into the audio/mic/headset jack might effectively disable both mics.

Personally, the Pyra is way low on the list of devices I'm concerned about for spying on me.

they have separate inputs to the codec. ALSA should be able to control them seaparately, though I haven’t started mucking around in this side of our OS yet.

 

[docbroke]

You are very bad at being a tin foil head if you seriously consider trusting a piece of software with a network interface, only ALSA is able to satisfy the paranoia!!111

BIG NOOO...! Nothing can satisfy paranoia...!

 

[rSl]

1. microphone is powered and operated through the twl6040 which is a quite simple and well documented audio peripheral chip. So it is easy to turn off the microphone by setting some registers.
2. it is probably easier to unsolder L1703 instead of the microphone module.
3. there is a transistor ("kill switch") controlled by an gpio that can cut off power from the modem. The modem firmware has no way to control it. Only the open source omap kernel and user-space. And there is a sense GPIO to check if the modem is on or off. For paranoia, there is even a current sensor (INA231) to monitor modem power consumption. All this is unlikely that some potentially rogue firmware ("closed os") inside the modem is aware of and can control. And nobody can have prepared for the monitoring software you will write.
4. there is no direct connection from the microphone to the modem - must be routed by software through ALSA or PulseAudio to modem or WiFi (for VoIP).
5. There is a version of the Pyra w/o modem...

Personally, the Pyra is way low on the list of devices I'm concerned about for spying on me.

i too think that pyra is as trustworthy as it can get these days. thanks again to hns for creating such cool design!

 

[Grench]

they have separate inputs to the codec. ALSA should be able to control them seaparately, though I haven’t started mucking around in this side of our OS yet.

Cool. So it is a hypothetical possibility that a person could plug a mic into the headset jack and set it up 3' or so off to one side and record off of both the internal mic and the wired mic and get a, likely out of balance, stereo recording of an event. Might be fun for someone to play with some day - just noting the hypothetical capability.

i too think that pyra is as trustworthy as it can get these days. thanks again to hns for creating such cool design!

Yes - it is clear that a a whole lot of thought and care went into the design. Thank you @hns!

 

[FBnil]

BIG NOOO...! Nothing can satisfy paranoia...!

Well, then I'm going to scare you more:

Weekend Project: MMA8451Q Accelerometer as a Microphone?

Last weekend was the Australia Day long weekend, so aside from chasing Shortwave Radiogram broadcasts and posting radiofax galleries, I also decided to build something I thought was rather interest…

goughlui.com

That accelerometer to know if you are shaking the device... well... you can also record sound with it Exactly what the Intelligence Agen&^$C^NOCARRIER

 

[levi]

Unfortunately, the air displaced isn’t quite enough to make an accelerometer vibrate, so instead, the accelerometer is probably best used as a contact microphone – kind of like an “inverse” bone-conduction speaker.

I think we're safe given the Pyra's accelerometers are internal.

 

[FBnil]

At least the Pyra is immune to this: https://alterslash.org/day/20191112#article-15199498

I think we're safe given the Pyra's accelerometers are internal.

And how about Electromagnetic eavesdropping?

https://www.cl.cam.ac.uk/~mgk25/pet2004-fpd.pdf

http://cryptome.org/emr.pdf

 

[levi]

I assume your joking, but in case you're not modern display links all use LVDS in which the differential voltage is 2.5V and the monitors don't contain a HV system any more. While sidechannel attacks are an increasing concern, they tend to concern software behaviours these days rather than RF emanations from grey boxen.

Edit: The HDMI output uses SCL rather than LVDS, but with it disconnected no actual current will flow. All you're likely to get off that is occasional 5V spikes on a pin as it tries to test for connections. The connection between mainboard and displayboard is MIPI-DSI if I recall correctly, and reading up on it I think that does use LVDS at least though.

 

[ToastBucket]

Read a research paper a few years ago where some people managed to get relatively accurate tracking of position and velocity of people just by analyzing EM disruptions using an off the shelf WiFi router.

Post automatically merged: Nov 13, 2019

Edit: The HDMI output uses SCL rather than LVDS, but with it disconnected no actual current will flow. All you're likely to get off that is occasional 5V spikes on a pin as it tries to test for connections. The connection between mainboard and displayboard is MIPI-DSI if I recall correctly, and reading up on it I think that does use LVDS at least though.

SCL? You mean the DDC clock? HDMI uses TMDS. MIPI uses LVDS at the physical level like you said.

 

[levi]

SCL stands for source-coupled logic, also known as current mode logic (CML). TMDS runs on top of that as I understand it (which is not much other than having started at the wikipedia page for HDMI for too long but there you go).

 

[Letalis Sonus]

LVDS in which the differential voltage is 2.5V

Isn't LVDS specified to use 1.2V?

modern display links all use LVDS

As an interface, LVDS is deprecated. Modern internal displays either use eDP or MIPI DSI.

MIPI uses LVDS at the physical level like you said.

MIPI uses a variable differential voltage, though. Its high speed mode only uses 0.2V.

 

[levi]

Isn't LVDS specified to use 1.2V?

That's Vcmo, I'm more interested in the signals that can be developed at range from specified voltages, although reading up on LVDS again, it says that the signalling is done using constant current either positive between one signal wire and the other or negative, such that is one wire is at a higher voltage the inverse is the other at the higher voltage, and they just swap over to signify a change from 0 to 1 or 1 to 0. At a sufficient distance those would be very hard to discriminate especially given the current is either -3.5mA or +3.5mA (developing about 350mV over the termination resistors), so I think you'd have a hard time reconstructing an image based on that personally.

Edit: In other words I was wrong to quote 2.5V as the differential voltage, but hopefully I'm more correct now.

 

[God Ginrai]

I hope the Drastic Port on the Pyra works this Time whit the and games that need the
I can remember on Pandora, if I load a Game where you need it, it’s dit crash
And the Pandora had a Microphone

DraStic had working mic support. I used it in Ace Attorney for the DS bonus missions that required it.

-God Ginrai

 

[matzesu]

My Pandora is a bit strange at some things, so I think it’s quite unique in its issues.. ^^
issues I have aren’t meant to be at other Pandorae..

But as an Early Pyra Preeorder, I don’t think I have to use my Microphone often now on Pandora..

 

[FBnil]

too bad. The quality of the mic is quite good. I've put lots of my musings to words on the Pandora.