Re:How to Make a $19 Police Scanner
by Maxious on Wednesday October 16, 2013 @01:00AM
(#6558094)
It does function simultaneously by switching away from the control channel. More advanced software like SDR-Studio or GNU Radio can extract multiple channels/signals.
Re:How to Make a $19 Police Scanner
by superkuh on Wednesday October 16, 2013 @01:38AM
(#6558191)
According to my tests (with many dongles of E4K and R820T tuners) it takes about ~50 milliseconds for everything to happen. Use keenerd's experimental branch for fast tuning with R820T tuners. I'm not sure the main branch has fixed the driver yet. My best guess is that something in libusb is slowing down the process which in theory should only take ~8 milliseconds.
Re:How to Make a $19 Police Scanner
by FiloSottile on Tuesday October 15, 2013 @11:02PM
(#6557793)
Yeah, it is said at the end of the article that the system is suboptimal because the receiver will have to switch back and forth between signal and control.
Re:How to Make a $19 Police Scanner
by mikeash on Tuesday October 15, 2013 @08:37PM
(#6557321)
You really only need to memorize it if you're going to speak it. If you're only listening, then you merely have to know that each word represents the first letter used to spell that word.
Re:How to Make a $19 Police Scanner
by balabaster on Tuesday October 15, 2013 @09:44PM
(#6557537)
Growing up in the outdoors and sailing, I learned to use this as second nature when relaying information - it's crazy that when you start doing this on the phone how few people understand it and you have to resort to A like Apple, F like Frank... it feels so childish after the clarity and precision of the NATO phonetic alphabet.
Re:How to Make a $19 Police Scanner
by freshhawk on Wednesday October 16, 2013 @03:15PM
(#6561519)
Interesting, I've never had people on the phone not understand it. If anything I sometimes get the opposite, they start talking faster and treating me with less of the "yet another moron on the phone" attitude that I'm sure you can't help developing working a call centre.
I always thought it was a shibboleth, like the tech support reaction I'm sure people here get occasionally where you can hear them realize they aren't going to have to explain how to click on a menu entry to you and switch scripts after you correctly use a technical phrase. Come to think of it, maybe they assume I was/am in the military. That never occurred to me before.
Re:How to Make a $19 Police Scanner
by andyjohnson0 on Wednesday October 16, 2013 @06:08AM
(#6558831)
I work on software that is used by, among others, the vehicle repair trade in the UK. The phonetic alphabet is commonly used by them to relay vehicle registration numbers over the phone. I've noticed that they rarely use it with customers though.
Re:How to Make a $19 Police Scanner
by jevinskie on Tuesday October 15, 2013 @09:22PM
(#6557451)
I want to continuously monitor for 10-codes and record incidents in a DB. Any ideas for the speech recognition portion? I figure it should be easier than full blown text-to-speech because I'm just looking for a small set of 10-codes: 10-4, 10-55, etc.
Re:How to Make a $19 Police Scanner
by nknighthb on Tuesday October 15, 2013 @10:39PM
(#6557711)
In my area, the two city PDs and FDs, the sheriff's office, and the rural fire districts all use LAPD[1], not NATO. I gather LAPD is most common throughout the US.
Re:How to Make a $19 Police Scanner
by devenson on Tuesday October 15, 2013 @10:53PM
(#6557767)
Practice the phonetic alphabet when you're on a long boring drive by spelling out the words on signs. Extra points if you spell the whole word before you pass the sign.
Re:How to Make a $19 Police Scanner
by ajays on Tuesday October 15, 2013 @10:25PM
(#6557674)
That page you linked to is a barebones guide to RTL-SDR. Any specific tips on finding and listening to interesting frequencies (or using it as a police scanner)?
Re:How to Make a $19 Police Scanner
by pantalaimon on Tuesday October 15, 2013 @10:47PM
(#6557738)
Hm, I missed the fact that UniTrunker is required for frequency hopping, there is OpenTrunker, but it doen't seem very active.
However, there have been reports [1] of UniTrunker running with wine.
Apart from that, you should be able to follow the directions in the OP.
You can use SDR# or more straight-forward, rtl_fm, (forget about the later, it doesn't allow to switch frequencies without restarting, but this should be possible with rtl_tcp) now with PulseAudio it should be possible to redirect the audio from those applications to dsd, but that seems to expect something like /dev/audio - I don't know how to provide that with PA.
Re:How to Make a $19 Police Scanner
by voltagex_ on Tuesday October 15, 2013 @08:53PM
(#6557363)
I remember running into this as well - there doesn't seem to be any better software on Windows to route audio like this (AFAICR the DSD decoder software expects a "sound card" input).
Re:How to Make a $19 Police Scanner
by voltagex_ on Tuesday October 15, 2013 @11:22PM
(#6557832)
Actually the "easiest" way around this problem would be to modify DSD to talk directly to SDR# or the rtlsdr DLL. I'm just not sure if the Windows patches to DSD were ever made public.
Obviously you can do a lot more with it than just listen to cops, and even this tutorial makes it look a LOT more complicated than it is.
Plug the device in, and select a frequency. It's pretty much that easy.
I use it for listening to everything from ham radio operators to airplanes. Seriously have had every penny worth of fun playing with this.
Here is everything you need, available on amazon prime: http://www.amazon.com/Receiver-RTL2832U-Compatible-Packages-...
(Although I recommend a better antenna. Dig one out of your garage, that is what I did.)
http://www.ebay.com/sch/i.html?_sacat=0&_from=R40&_sop=15&_n...
Can the same USB receiver function simultaneously as the signal and control receiver? If not then the HW cost is $19*2.
They do sometimes seem surprised that you are using the "offical" words so I gather knowing this is pretty uncommon.
<nostalgia time>I learned it from The Pocket Ref book as a kid</nostalgia time>
I always thought it was a shibboleth, like the tech support reaction I'm sure people here get occasionally where you can hear them realize they aren't going to have to explain how to click on a menu entry to you and switch scripts after you correctly use a technical phrase. Come to think of it, maybe they assume I was/am in the military. That never occurred to me before.
CMU Sphinx and Julius both also have options for continuous speech recognition, but it's quite a feat to train your own [good] language model.
[1] http://en.wikipedia.org/wiki/LAPD_phonetic_alphabet
For flying, and for ham radio, we use NATO.
English > NATO may not be easy without memorization, but there's no need to memorize anything if you want NATO > English.
Apart from that, you should be able to follow the directions in the OP. You can use SDR# or more straight-forward, rtl_fm, (forget about the later, it doesn't allow to switch frequencies without restarting, but this should be possible with rtl_tcp) now with PulseAudio it should be possible to redirect the audio from those applications to dsd, but that seems to expect something like /dev/audio - I don't know how to provide that with PA.
[1] http://forums.radioreference.com/trunking-control-channel-de...
Ok, turns out there really is some additional research/instructions required to get this running on Linux…
As long as you have a $25 software package.
That runs on your $100+ Windows installation.
On your multi-hundred-dollar laptop.
Asshole.