You are listening to San Francisco
youarelistening.to
youarelistening.to
Listen to live Bitcoin transactions: http://btclisten.com/
Listen to airports: http://listentothe.cloud/
Listen to train ride: https://www.youtube.com/watch?v=9pVWfzsgLoQ
Listen to driving through NY: https://www.youtube.com/watch?v=3kV4zowtJC8
I'd love to get high-quality audio of a Subway train late at night when there are few passengers and little talking.
Also, check out https://somafm.com/missioncontrol/
(And another also - the original Grand Theft Auto had a CD audio track that represented the ambient police audio; that's what played when you were in a police car, ambulance or fire truck. There's also a City Ambient for when you're walking around. ( https://www.youtube.com/watch?v=87Ls5ImYEXA )
(Warning, slightly terrifying paper)
P25 is very vulnerable to jamming, decrypting encrypted traffic, traffic analysis, and geo tracking of individual radios.
-
Because of the design of P25, commonly used in America, if you enable encryption, a tiny bit of noise in the signal makes it unusable. One problem is that there are separate error correction codes for different parts of the packet, and very few error correction bits were allocated for the important encryption bits.
To get around this, sometimes just the dispatchers will have encryption on, since those packets are traveling over fiber to the transmission towers, where they are broadcast at high power levels. The other direction, personal radio to repeater tower, is almost never encrypted because of the much lower power of the personal radios.
P25 has some nice features - our local area has perhaps a hundred different public safety organizations, all of which can talk to each other at any time, and often coordinate for any major incident. But security is not one of them.
And that is just fine.
Police and firefighters/medics[1] are civilians and are in their role to serve civilians at the pleasure of civilians. There is no need for secret comms and, in fact, it is often quite valuable for the public at large to have visibility into their operations in real time.
There is no need to encrypt public safety radio comms and you should be quite suspicious of anyone who claims that there is.
Those radios belong to you. The frequencies belong to you. The people who use them are your own servants.
[1] I am a firefighter/medic in the SFBA.
However, the linked talk above notes that the researchers accidentally discovered that some "sensitive tactical law enforcement operations" were unwittingly operating in cleartext, and divulged to the open air "names and/or identifying features of targets and confidential informants, their locations, descriptions of undercover agents, [...] plans for forthcoming takedowns and operations," and a few other things you probably don't want to widely announce.
It also notes that radio locations could be easily discovered, even if they're not communicating. Bad guys could therefore build a map of possibly-federal-agents in the city, and learn things they're not supposed to know even if they can't hear what's being discussed.
Again, PD/Fire/EMT are great to have open and mutually communicating. But there are (life and death) reasons why you might really need a secure channel in other cases.
We expect block headed 18yo infantrymen not to send sensitive information over insecure channels. I don't think asking the same from our public servants is unreasonable.
There are false positives (e.g. you're speeding down the autobahn, and there's an ambulance in a town you're passing through), but it's probably good enough...
Most common thing heard on a police scanner around here: "Call me on my cell."
https://www.youtube.com/watch?v=7or-_gT8TWU
one of my favorite con talks in a long time. Highly entertaining and hugely informational.
That is incorrect. There is no difference in voice quality or degradation in quality of service between an encrypted P25 signal and an unencrypted one.
An analogy would be like claiming that there is a difference in quality of service between http and https.
The problem is that without encryption, P25 packets can handle a bit of radio noise and still be error corrected into usability. With encryption, in practice, certain bits get altered that cannot be error corrected and the packet cannot be decoded. P25 error correction ability is not evenly across the entire packet.
HTTP and HTTPS are riding on top of TCP, which mostly guarantees in order bytes, the correct bytes, and no missing bytes. If you loose the guarantees of TCP, and allow say 1 out of 2000 bytes to be altered randomly, HTTP would still mostly work, and HTTPS would be dead as a doornail.
The signaling, voice and error correction is then formatted into P25 speech frames. A transmission starts with a header data unit, identifying the destination of the call and if encryption was used or not.
It then continues with a series of logical link data units (LDUs), that carry the digital speech. The LDUs also repeat the signaling information and the synchronization sequence, so radios can join the call late, or reenter if they drop out due to bad RF coverage and then come back into the call. The call ends with a terminator data unit, indicating that the PTT’s been released, and this is the end of the call.
To get around this, sometimes just the dispatchers will have encryption on, since those packets are traveling over fiber to the transmission towers, where they are broadcast at high power levels. The other direction, personal radio to repeater tower, is almost never encrypted because of the much lower power of the personal radios.
This is also not correct. Encryption can either be manually turned on or off by the individual user (radio) and/or the dispatcher, or encryption can be "strapped" to a talkgroup in cases where full time encryption is used. THe scenario you describe is all based on the user's choice, not an operational need due to power levels.
http://www.crypto.com/papers/p25sec.pdf
"Because of the above-described property of the error correction mechanisms used, especially in voice frames such as the LDU1 and LDU2 frame types, there is no mechanism to detect errors in certain portions of transmitted frames. This was a deliberate design choice, to permit undetected corruption of portions of the frame that are less important for intelligibility."
Edit: Why am I being downvoted? It's a genuine question... I see listening to police scanners is a popular thing and I just want to understand why
I didn't mean to offend anyone's interest, I just wanted someone to maybe show me a side I'm not seeing. I should word my questions with more thought and empathy in the future.
In my world, posting a well-reasoned on-target opinion on a website is not isomorphic with a video game, where I need to score XP or some other adolescent incentive.
So, given the lack of specific feedback, I suggest you ignore the downvoters, and do what I do, which is to imagine them as ignorant youngsters who actually think anybody cares what they think. They lack power in their own lives, and so they get the illusion of power over others with their downvote button. We should feel sorry for such creatures.
Perhaps one could try Buddhism? ;-)
See how many are listening on Broadcastify as we speak:
http://www.radioreference.com/apps/db/?sid=60 looks like motorola type II is still preferred, although AP25 is...available?
Open radio also makes it more likely the news media is there when the action is still happening instead of giving cops time to paper over their mistakes.
Police getting to operate in secret is a dangerous road to go down. Same reason that the names of suspects who get arrested are public. It might suck for their future reputations, but the alternative is a world where the police get to make people disappear.
Worth noting, though, are the SomaFM links to similar stuff over there. I’m a big fan of Mission Control myself.
Ambient music mixed with the sounds of San Francisco public safety radio traffic: http://somafm.com/sf1033/
I also recommend their DEF CON channel, which has (usually humorous) snippets from various talks between tracks: http://somafm.com/defcon/
Also here is cool simple shell script - "noise.sh"[1]