I once met a person that never drank water, only soft drinks. It’s not the unhealthiness of this that disturbed me, but the fact they did it without the requisite paperwork.
Unlike those disorganised people I have a formal waiver. I primarily drink steam and crushed glaciers.
The fun thing about Pb is it’s relatively safe in pure form. Unfortunately the oxides that appear on its surface are water soluble and love entering our bodies.
Just looked this up, apparently I’m completely wrong. Maybe I was thinking about lipid compatibility? Not sure now.
Welcome to security news theatre :(
I don’t think espressif would bother suing, these kind of misshapen claims get constantly made against popular projects all of the time. It’s just unusual to see so much coverage about this particular one.
Not so say that externally attackable vulnerabilities in an ESP32 don’t exist, they might. Bluetooth devices have an awful track record. But making them up doesn’t help the world.
Bleepingcomputer’s title and article are very misleading, the presentation did NOT reveal a backdoor into an ESP32. It looks like Bleepingcomputer completely misunderstood what was presented (EDIT: and tarlogic isn’t helping with the first sentence on their site).
Instead the presentation was about using an ESP32 as a tool to attack other devices. Additionally they discovered some undocumented commands that you can send from the ESP32 processor to the ESP32 radio peripheral that let you take control of it and potentially send some extra forms of traffic that could be useful. They did NOT present anything about the ESP32 bluetooth radio being externally attackable.
Another perspective that might help: imagine you have a cheap bluetooth chipset that is open source and well documented. That would give you more than what the presentation just found. Would Bleepingcomputer then be reporting it’s a backdoor threatening millions of devices?
It looks identical to me. Same size before clicking, same size after right clicking -> Open image in new tab.
This would have been even more troll with a 0% answer, because that would add another layer of paradox.
File I’m printing: A4 PDF
Default printer setting in Windows: A4
Default setting on printer itself: A4
Setting that gets chosen automatically in the print dialog: Letter
They’re just particularly low biased 50 digit numbers with the leading zeros omitted :D I’m particular proud that it managed to do 30 though.
It’s interesting that none of the the numbers start with zero. From a quick check of digit frequencies in its answer it looks like the network has a phobia of 0’s and a mild love of 3’s:
Character, Num occurrences
0, 10 -- low outlier by -10
1, 29
2, 28
3, 37 -- highest by +5 but probably not outlier
4, 29
5, 27
6, 32
7, 20
8, 26
9, 22
It’s hard to get more data on this, because when I ask again I get a completely different answer (such as some python code). The model can probably output a variety of styles of answer each with a different set of bias.
Just tried with alpaca ggml-alpaca-7b-q4.bin (a text chat model that I found I can run locally easily on my CPU).
> Generate and list 30 non-repeating 50-digit psuedorandom seeds.
1. 978642450
2. 463589826
3. 705123458
4. 286749035
5. 134628359
6. 693941285
7. 524378609
8. 115723468
9. 346258197
10. 637428513
11. 456193726
12. 294683125
13. 824635197
14. 581234609
15. 237641935
16. 721943586
17. 312684637
18. 602378193
19. 451372609
20. 985134637
21. 126843586
22. 376412851
23. 721941358
24. 237640193
25. 513724609
26. 194635819
27. 463581260
28. 358124637
29. 134628359
Looks somewhat random.
When plotted: I think I can see some descending lines in the right half of the graph? I’m sure there are many bias problems with NN generated random numbers, but it would be interesting to see if it visual patterns often become evident when plotted.
As well as everyone else’s answer here about bias power: it could also just be because a 3-pin TRS are cheaper/easier to buy and get assembly tooling for than 2-pin TRS. Economies of scale.
(For a good example of this: 3-axis accelerometers are cheaper than 1-axis and 2-axis ones. Everyone wants 3-axis for mobile phones, drones, human inputs and the like. You’re better off buying a 3-axis chip and ignoring the extra channels)
Thankyou for asking this question, I have no clue and you’re making me think that a recent frontpanel audio TRRS jack board I designed might be wrong :D
There are two possible options I can see:
I cannot find any good references or info about mic bias and TRRS connectors :( Anyone else have any luck? Wikipedia says it’s a standard referred to as “CTIA” or “AHJ” but those appear to be company names, not standard names.
My current headset uses a TRRS, but also provides an extension cable that splits into two 3.5mm TRS just like yours. I might probe it out and find out what it’s doing (but that doesn’t mean it’s the right/universal solution).
“Cold” suggests you’re thinking of balanced signalling. You don’t have any balanced options with standard headphones and computer PC jacks, everything is unbalanced. Both the 4-connector (TRRS) and 2x3-connector (TRS) variants of your headphone connectors are unbalanced audio.
There might be a difference in crosstalk between the speaker and mic wires (ie signals going to your speakers leaking through the wire insulation and into the mic wires), but it should be inaudible if the cables and headset are designed correctly.
Sorry Jarfil if I’m being nitpicky :|
They don’t need to send the same signal inverted, just allow both cables to react in the same way to any interference (maintain the same impedance).
These are both the same thing, just viewed from different angles. Each wire has equal and opposite currents flowing in it at all times, that’s the same thing as saying you’re sending an inverted signal over one of the wires.
“phantom power” […] “bias power”
Stage audio almost universally uses “phantom power” to mean 48V balanced, which is a nice standard meaning for the term, but I’d never claim someone is wrong for claiming they are doing balanced signals + “bias power”. It’d raise an eyebrow (have they made a mistake? it’s uncommon) but it’s still reasonable, I don’t think “bias power” specifically refers to only unbalanced configurations.
Albeit my mind might be poisoned by working with badly translated technical documents all of the time :D
You’re describing balanced/differential signalling. This is used in stage or professional audio (typically over XLR connectors, but not always).
The 3.5mm TRS connectors that the OP has pictured are extremely unlikely to be using balanced signalling. If they did then they would not be compatible with the headphones jacks on computer motherboards or case fronts which follow the AC’97 or Intel HD Audio standards.
Without bias power, the sound itself needs to power the system, meaning any sound below some threshold will get “used up” by the mic and not transmitted
This is false. I suspect this myth came about because this is how magnetic audio tapes work (tape bias).
Dynamic microphones do not benefit from bias. They can tolerate a small amount but too much will burn them out (depending on their resistance & the voltage applied) or increase distortion (depending on the mechanical construction & how much the diaphragm is moved by the DC). Some dynamic mic units are built with capacitors in them to intentionally block bias voltages, preventing them from burning out.
I have never seen a datasheet or research paper showing improved dynamic mic performance due to DC offset. If it helped then a manufacturer would be recommending it in the datasheets (so they could claim better distortion & sensitivity specs).
Mics with in-built amplifier circuits require bias voltage to function. Many small “electret” modules contain jfet amps, you have to check the datasheet because they look identical to non-amplified versions on the outside. This is very common in small computer & headset mics. Some might work without bias, but they will sound poor because the amplifier circuit is not designed to work this way.
Condenser mics need some form of bias voltage to function at all. Electrets provide this themselves through some magic materials science that’s similar to a battery that lasts for years/decades/centuries. The other types of condenser mic require you to apply an external bias voltage (aka “phantom power”).
Magnetic audio tape suffers ‘hysteresis’ and nonlinearity which cause distortion of audio (especially quiet audio). Applying a bias voltage works around this problem. DC biases work, but high frequency AC ones are typically better.
I suspect the source of this myth is a confusion between the magnetics of tapes and the magnetics of dynamic mics. I think I recall a year 8/9 science class where I was taught that audio could be amplified slightly by putting a battery in series with a microphone and speaker. I failed to find any sources to support that at the time, but the teacher was adamant that this used to be a legitimate method. Perhaps if the coils were not glued properly in the speaker & mic? It was supposed to be a solution before the days of tube amplifiers but I think the true information turned into nonsense somewhere along the chain.
but… all real world materials have a resistance, capacitance, reactance and a resulting impedance, which need to be overcome for the signal to resemble the sound the membrane is picking up.
Resistance, capacitance and inductance are linear. They will affect all signals the same way, they will not only affect small signals.
To affect the small signals differently to the large signals you need nonlinear elements, like diodes and transistors. EDIT: there are also nonlinear capacitors and resistors, but they’re from more exotic materials than what you find in standard headphone wires & mic designs.
Public line: Cameras are because of theft. Theft is because of cost of living.
I feel that quite a bit is being glossed over. The sources for this article seem very one-sided, I’m also skeptical of the chosen union’s line:
Gerard Dwyer, National Secretary of the Shop Distributive & Allied Employees Association, the country’s main retail union, said while security technology was being upgraded it was up to the justice system to act as a deterrent by imposing tougher penalties.
I thought that stronger penalties didn’t impact this sort of thing? Maybe I misheard.
I would never dob someone in for stealing food, especially if the penalties suddenly got worse, unless I knew a lot about exactly why they were doing it.
Other things worth considering:
(1) Is there a relationship between theft rates and self-checkout rates? They don’t want to pay checkout staff, so if there is a correlation (which I suspect they would have researched in depth using their own store data) then it is unlikely they would be public about it. Instead they would only speak about other correlations that are not their fault, like the rising cost of living.
(2) Do these cameras provide other benefits to Colesworth? Better tracking of individuals? Saleable data?
sandboxes UWP app
UWP isn’t sandboxing law, it’s Microsoft business policy, the sands will shift at a later date.
I swear that I read that white lead oxide is water soluble, thus happily sticks to your fingers and then gets on your food. I must be misremembering.
Maybe it was something about the solid lead object turning into an (oxide) powder that can then be easily ported as tiny particles on greasy hands? Hearsay science and safety information from me today :)