With some filtering on the output to knock down harmonics and aliasing noise, and as long as the power output is low enough, this could actually be legal in the US. IANAL, etc, but this is my understanding:
It's actually ridiculously tricky to measure transmitter power output in the terms that the FCC regs are written in, but the rule of thumb is that, given the receiver sensitivity of a typical radio, if you lose the signal when you're more than 200 feet from the transmitter, you're probably in the clear.
This means a perfectly legal power level would be more than adequate to cover your house, office, hackerspace, or whatever.
If by "ridiculously tricky" you mean "requires specialized equipment and domain knowledge". Like say, skiing.
With a cheapish Rigol spectrum analyzer and homebrew receive antennas and DIY wooden/pvc tripods you could probably do 3m open-air measurements to within 3dB of what you would measure in an anechoic chamber at a test laboratory if you were very careful. I can 100% assure you that is possible with a commercial dipole kit and horn antenna, neither of which is that expensive if you're getting paid. This in in the FM band so you could use dipoles and a low-end analyzer to the tenth harmonic.
You can absolutely hook an antenna to a spec-an, but how do you calibrate the result? I can get a power measurement of mV or dBm at the instrument input, but that's not how the regs are written.
They're written as field strength in free space (microvolts per meter), which as I understand it, either requires a calibrated field probe, or a calibrated antenna. I'm sure there's a way to perform such calibrations from first principles, but that's where I'm invoking "tricky". If you've got a trick to it, please share.
In practice, Part 15.23 is pretty generous about home-built equipment, and I think since the "200 foot" thing came from the FCC's own mouth, if you can show that you walked the perimeter with a portable radio and confirmed that you're good, that probably shows good faith on the power issue.
...
Absolutely yes a spectrum analyzer is helpful to make sure your harmonic filtering is working. I would put that forth as "good engineering practice" under 15.5, and it's far more than any hobbyist could be reasonably expected to do even just a few years ago; this equipment has gotten mindbogglingly affordable and accessible.
You get a field strength the same way the pros do - from the antenna factors for your antenna and the cable loss. Look at a public FCC test report for an exposition of the method. DIY dipoles and even horns are not going to be very much different from ideal. Calibrated antennas are typically not more than a dB or two off ideal. Dipoles not even that unless they're broken. Horns and log-periodic antennas, enough that it can matter - but fine for hobbyist practice and even most pre-compliance work. I once discovered an RF engineer was using ideal gain numbers for a horn antenna instead of the calibration data for pre-compliance - it fortunately did not make a practical difference. The Roberts balun for measurement dipoles is standardized and I have instructions for building one around here somewhere. That is literally the reference for calibrating other antennas.
The primitive method for getting an antenna calibration is the three-antenna method. If you are particularly motivated, you can build three Roberts dipoles and check them. For a number of reasons, you might want to have three dipole kits anyway if you are doing a lot of these measurements.
"Tricky" in the sense that skiing is tricky. You will struggle learning from books and the internet. Starting from scratch with DIY equipment would be quite challenging. But it's all in standards and the open literature.
I have to admit that having a ham radio license that would be easy to take away is one reason why I wouldn't mess with anything questionable radio-wise.
I wonder though if you could trust the published specs of a commercial FM transmitter to calibrate a power meter on it from a distance of a mile or so...
Basically if anyone receives it that wasn't expecting to it's illegal. That is at the core of the FCC rules in the US, you can't interfere with the lawful use of the band, nor can you cause "interference" on unused bands. That said, the process for prosecuting you is long and somewhat convoluted as they have to file a report, then the FCC will confirm the report, Etc.
But this kind of project (and a similar I saw that broadcasts on every FRS[1] channel at the same time) is a great example of something you can do with SDR that you cannot (easily) do with transistors and boring old analog radio.[2]
It is one of the reasons I enjoy it (playing around with SDRs and DSP) so much as I keep learning new things.
[1] Family Radio Service
[2] I know, I know, building 20 modulators isn't all that tough yadda yadda.
If it's intended for personal consumption it's fine. It's not copyright infringement to hear my neighbor's stereo if it's a bit loud. It is if he's using it commercially for hosting a block party selling tickets it's a problem.
Ah, but if your neighbor's stereo is faintly audible in the distance while you record a screencast, your video will get taken down off Youtube very quickly.
Nah, don't kid yourself. Tech lik€ this is NOT fine. You're a douche if you deploy this. It may be technically interesting but just a dick move to use it.
Small, low-power FM transmitters have been around even prior to Mr. Microphone. Hooking to your Hi-Fi so you can get your music on the radio in the garage is as old as Jarts.
The main problem with this line of reasoning is that it neglects important parts defined elsewhere making it seem like this is legal when its not. IANAL, but I do hang out with a lot of hammies. You can't help but absorb a lot of technical stuff with them around.
It need not saturate the bands, you technically are in violation the moment a non-licensed entity has caused interference to a licensed entity.
Willful is a pretty low bar since general intent can be derived from any related negligent acts. Its pretty safe to say transmitting on these bands without a license is illegal, and yes its a fairly big deal.
Most of the public radio spectrum has been monitored, and archived for later retrieval if circumstances dictate since the 60s. There are services that aggregate this data and make realtime info available globally from satellite systems. Trilateration of signals is fairly trivial in most cases.
You may not get caught if you always stay below the noise floor, but you can't really have any real use of the spectrum in doing so, and algorithmic scanning of structured signals is always improving.
It is really negligent to make it seem like these things are legal, when they aren't, and the author should get in trouble for that.
A simple disclaimer isn't going to cut it when then primary purpose is effectively breaking the law in many civilized nations.
For what it's worth (which isn't much), all of that hinges upon the FCC's motivation and capacity to enforce its rules.
If amateur and citizens band radio shenanigans over the last few decades have proved anything, it's that you have to be an absolutely massive pain in the radio-wave-receiving public's ass to really get their attention.
If you had this device on 14.300, 7.200, or channel 6? Meh. No one's gonna care.
In my town there was a guy who was absolutely obliterating 88.5 as well as 88.3 and 88.9 broadcasting some Hispanic hip-hop with some grossly overpowered local transmitter. He overpowered the stations on the entire east side of town. It took over 6 months before it was shut down. I think the FCC penalties are so harsh because their enforcement arm is so small that they can only go after the most blatant offenders. Not that I would want to risk fate given how harsh those penalties are.
I hold a ham radio licence in a couple of countries, although what's happening here is far outside of the remit of ham operations.
I'm pretty certain this is not legal in my locale, but I don't think the author of the article is making this out to be legal - and I don't think they should get in trouble for any of this publishing [assuming you're adressing them, not the commenter]. Perhaps the equivalent of a script kiddy could get up to no good with it, nothing new from the last 20 years of computer hacking then.
I'd far rather see enforcement and ire directed at the proliferation of poorly shielded junk that spews noise all over the spectrum, at surprisingly high power, that seems to have no trouble being sold online or imported for sale in many countries - and is often, strictly speaking, not legal with regards to necessary EMC, and probably safety, standards.
... reminds me of the times I've taken my HT to Southern Cal and heard things on the radio (in the ham band and elsewhere) that I haven't heard any place else, like the ham who would hold court on the repeater with background music.
Eh, if they're so low power that they're not stepping on their neighbor's reception I'd personally say it's fine. Tree falling in the woods situation and all. That definitely varies based on where one is playing around with such things, and it's always important to understand how far your transmissions are actually going (which can be surprising if you don't have much experience with it!).
Someone in a shed on a few acres playing around with an SDR? Go ahead, I don't care. Playing around in an amateur anechoic chamber, have fun if you're sure you're containing things. On a desk in a random apartment in Manhattan? What an asshole, call the FCC.
Is it potentially against the law to do this? Yeah, probably. How bad is it? Somewhere around jaywalking on an absolutely empty street and dumping waste in a park, greatly depending on multiple factors.
I remember going to the drive-in movie theater in the 80's. You had to tune your car radio to a specific frequency to hear the movie audio. If this device had existed back then, you wouldn't have needed to tune your car radio at all. I doubt the neighbors would have liked that though. As an aside, I remember being jealous of the neighbors because they got to watch all these movies for essentially free. The irony is that today, we're bombarded by content, it's become essentially free.
> You had to tune your car radio to a specific frequency to hear the movie audio.
This was great because it allowed for stereo. This also replaced the earlier idea of taking the speaker from the pole next to the driver's window, and then hanging from the window inside the car. This was just a mono speaker, but it was also right next to the driver's head. It reduced the drive-in's repair bills from people driving away before placing the speaker to its host stand (and other hooliganism)
> you wouldn't have needed to tune your car radio at all
You still need to tune it but there are many "center frequencies" that can be tuned instead of just one. So, you wouldn't have to tune far but you would still have to tune :) If you had a knob like most radios in the 80's it would save you from cranking the knob several times to get the dial into the right range before fine-tuning.
I think the theory is that every car that has an FM radio has it tuned to something. So no matter what, as soon as you put on the FM radio, it would have the audio.
Yeah, until you have a 5 year old (probably me in this case in the 80s) that plays with random knobs on the dash when they get in the car. Is it volume? Is it the tuner? We’ll know once the car turns on!
The fancy Christmas lights on houses that sync with music do the same — instructing passersby's to tune to such and such in their cars as they drive past.
It's kinda refreshing that the "Legality" section is entirely concerned with power and frequencies, and not at all about the IP rights around Swift's music.
I was hoping this was a Tayloe Mixer[1] based SDR. The Tayloe mixer has excellent image and harmonic rejection.
<Oops... I missed the important detail about 100 different songs... sorry. I'd suggest putting the hit "Never gonna give you up" on loop for that version of the project>
In hardware, it would be easier to feed a 200 khz square wave into a diode and pull off the first 10 mhz of harmonics, low pass filter it, and mix that with 99.1 Mhz fm audio to get 88.1 to 108.1 Mhz fm at 200 khz spacing.
In software, you should be able to precompute a complete one cycle composite of -10 to +10 Mhz sine waves spaced 200 Khz apart. You could then loop this abitrary waveform and multiply it by an FM modulated source at 99.1 Mhz to get the same effect with a lot less processing power. This is the kind of thing GNU radio[2] is perfect for.
<Correcting the above>
Precompute the FM signal for each of the songs in it's channel from -10 to +10 mhz for the length of the song. Store them as 16 bit IQ samples. Each file will be different length.
Loop though all of the songs, adding all of the components into a 32 bit I and Q sum. Round those to 16 bit values, or as appropriate for your SDR transmitter. You should have a very easy to compute 100 channel FM station with each channel separately timed.
DSP engineer here:- the typical way to do this for better performance would be to do the FM modulation of each channel at a lower rate, say 100kHz, then use an FFT-based channeliser to mix each sub-band to the right carrier frequency and upsample in one step. That helps to break the n^2 problem mentioned in the article.
It would be an interesting exercise for someone to sit down with GRC and implement as many different approaches as possible, and benchmark them. This kind of silly-but-concrete goal can be solid gold for teaching.
The CRT enthusiasts on Reddit would love something like this, but for analog TV channels.
Something that people ask occasionally is how to use their vintage TV's tuner to receive digital TV. But they don't want to tune it to channel 3 and then change the channel with a digital TV box; they actually want to change the channel with the old tuner. The current advice seems to be to spend $hundreds to get an analog CATV head-end, and a separate digital tuner for each channel, but I've never heard of anyone actually following through and doing that.
It's definitely possible to transmit multiple TV channels at once, and I have experimented with that in the past. Since TV channels are 6 to 8 MHz wide (depending on what part of the world you're in) there are practical limits to how many channels can be transmitted at once. Consumer-grade SDR boards are limited to about 56 MHz total bandwidth, which would be enough for 7-9 channels.
This is something that would fit right into some dystopian, proto-cyberpunk movie.
Like, a bunch of activists need to disable comms of some nightshift guards in an area. So... someone figures out how to put more juice into this thing to pump out music on all frequencies they could have.
Bonus points because this wouldn't just jam their communications with white noise. It would be confusing as hell. Taylor Swift on the main channel. Switch to the backup. Irish Folk. Switch to something you and Bob used some time back. Power Metal! What?
Instead of struggling to generate this in real time, perhaps you could generate it in advance and loop it? The median Taylor Swift song is 3:52, then at 20M samples per second (and four bytes per sample) you need a little over 19GB [1] and I think streaming from SSD should be fast enough. Could probably also compress it.
I'm not sure if Stephen is on HN, but I asked him this question last week[1]. His response was that since the songs are different lengths, a file that was truly loop-able would be super long (if you have a 90 second song and a 2 minute song, you need 6 minutes of audio to create a file that perfectly loops both of them)
Seems like there is probably some set of clever hacks here that could get you around this (although I don't know enough about radio to propose any); I think I asked about pre-computing some state for each song on its own and he had a good response to why that either didn't work or didn't help much, but unfortunately I don't remember it!
I was thinking you could pick a group of songs that were close to the same length, and then make up the difference with fading, gaps, or a fake DJ? For this to work I think you want your overall loop to be just a few minutes.
This reminds me of a practical application of such tech that never really made it. A chap had his invention on (Australian) ABC's The New Inventors, 20 years ago, named Emergalert. It could announce the presence of nearby emergency vehicles. It seemed like such an obviously good idea.
> I'm not convinced that I'm operating anywhere close to peak efficiency. There may be some huge DSP-specific shortcut that I'm overlooking - I'm certainly no expert. But the current code works well enough.
Something along the lines of taking the FM at baseband, transform to frequency domain, copy result n times and shift the coefficients to the right indices, much wider ifft to RF should work, right?
Yes, modulate then FFT for each signal, do the upconversion and mixing in frequency domain, then a single large IFFT to give the final bitstream. Roughly the inverse of this (single large FFT, filter in frequency domain, and IFFT individual signals) is how the 'wideband sdr' at utwente.nl works to support hundreds of simultaneous receivers on a single gpu.
The amateur radio folks working with these techniques on HPSDR hardware were calling the technique "Direct Fourier Conversion" but I do not know if they ever got to a releasable state. They spent an awful lot of time prematurely optimizing their stuff to run on the original Jetson board which was not really adequate to the task, so I think that was likely frustrating and killed momentum.
The most efficient way is a filter bank technique. Key words to Google would be "reconstruction", "dechannelizer", "synthesis".
A power of two number (N) of equally spaced frequency channels can be efficiently combined using a short fir filter on each channel followed by an fft where one sample from each channel is input to the fft per frame. Then you get N samples of your output. There is a bit more nuance but the author's bandwidth and number of channels are trivial to handle with this method.
Since each channel is already assumed to be spaced in frequency, you are essentially already in the frequency domain and only so only one fft stage is required.
From phone phreaking 40 years ago to FPV RC, I've found it to be a good indicator you're focusing on interesting emerging technologies when fellow early adopters in your new hobby say "Technically, this isn't even illegal... yet."
The algorithm is an amusing nerd snipe. I’m not a DSP expert, but…
The idea of transmitting lots of low-bit-rate signals on different evenly spaced subcarriers is quite popular. OFDM is an example of exactly this, and the computation scales just fine. The same trick ought to work: compute in frequency space. You know what frequency-space signal you want to send on each subcarrier, so assemble the output in frequency space and iFFT it. I bet it can be done genuinely OFDM-style using transform lengths calculated to get the subcarrier spacing right with some additional care to get the boundary conditions right (conventional OFDM has a guard interval and doesn’t even try to transmit a continuous signal).
If all the songs are the same, there’s a much nicer solution. If the complex amplitude of the modulated song is A(t) (that’s just I + i⋅Q), then two copies are:
where n is the number of copies. The thing on the right is a geometric series and can be summed algebraically. So all that’s left is to upsample A, and it doesn’t need to be done especially precisely.
This is like a real life version of "I'm broadcasting this message on all channels," or when the aliens invade in a movie and you see the same thing on every TV channel. I love it.
> The Taylorator is a piece of software which allows me to flood the FM broadcast band with Taylor Swift's music.
You don’t need an SDR, just put on her song and most people at traffic lights will roll their windows up!
That being said, from my experience, power will be the deciding factor. I tried both LimeSDR and BladeRF, and when maxing out the power, they overheat really fast, so they probably will only work while driving/cycling so no obstacles. And then again, most people use Spotify or similar in their cars.
If it works on AM, however, that would be great to test.
> Stations will only appear on odd-numbered frequencies, like 88.1 MHz, 94.5 MHz, 107.3 MHz, etc. There's a technical reason for this - every FM broadcast takes up about 150 KHz of bandwidth, and spacing the broadcasts like this allows for an extra 50 KHz of wiggle room.
I believe the odd-numbered frequencies were chosen because if only even were allowed then there would be legal knock-down-drag-out fights over the round numbers.
I don't think it's anything so interesting. The FM band starts at the round number of 88.0 MHz and ends at 108.0 MHz. To divide that evenly into 200 kHz channels without bleeding over the ends of the band, they have to be centered at the odd decimal points.
I noticed a long time ago that the frequencies for AM channels are always divisible by 9 (at least in Australia). Never looked into why (those were the days before the internet...)
Must save it. Some day it might be needed to do some advertising regarding mutiny against government or to gather people in the post apocalypse time...
> an SDR is that a sound card takes real-valued samples, and an SDR takes complex-valued samples
Coming from a DAW background this concept was incredibly interesting. I am curious if stereo audio has ever been represented in this way, or if it is generally just used for the half sample rate?
Actually thinking on this now I guess that doesn't really make sense. But still really interested in this alternate view of sampling waveforms.
Testing has shown that humans can't hear phase differences, unlike radio demodulators, so there's no real use for giving a sound card complex numbers.
(Before someone replies about snare drums, 3:1 rule, etc. — of course if there are phase differences in the audio before you mix it, it will cause audible artifacts in the mixing process. Which is why there's an effect called a "phaser".)
I might have missed it, how does OP broadcast across 80mhz with a 20mhz SDR? I think my mental model about SDR capability is wrong, does sample rate not necessarily gate the range at which an SDR can TX/RX?
An SDR's sample rate only limits its bandwidth. Most devices include a mixer, which allows the transmitted or received signal to be shifted in frequency by an amount much larger than the bandwidth.
With some filtering on the output to knock down harmonics and aliasing noise, and as long as the power output is low enough, this could actually be legal in the US. IANAL, etc, but this is my understanding:
It's actually ridiculously tricky to measure transmitter power output in the terms that the FCC regs are written in, but the rule of thumb is that, given the receiver sensitivity of a typical radio, if you lose the signal when you're more than 200 feet from the transmitter, you're probably in the clear.
This means a perfectly legal power level would be more than adequate to cover your house, office, hackerspace, or whatever.
https://www.fcc.gov/media/radio/low-power-radio-general-info...
If by "ridiculously tricky" you mean "requires specialized equipment and domain knowledge". Like say, skiing.
With a cheapish Rigol spectrum analyzer and homebrew receive antennas and DIY wooden/pvc tripods you could probably do 3m open-air measurements to within 3dB of what you would measure in an anechoic chamber at a test laboratory if you were very careful. I can 100% assure you that is possible with a commercial dipole kit and horn antenna, neither of which is that expensive if you're getting paid. This in in the FM band so you could use dipoles and a low-end analyzer to the tenth harmonic.
You can absolutely hook an antenna to a spec-an, but how do you calibrate the result? I can get a power measurement of mV or dBm at the instrument input, but that's not how the regs are written.
They're written as field strength in free space (microvolts per meter), which as I understand it, either requires a calibrated field probe, or a calibrated antenna. I'm sure there's a way to perform such calibrations from first principles, but that's where I'm invoking "tricky". If you've got a trick to it, please share.
In practice, Part 15.23 is pretty generous about home-built equipment, and I think since the "200 foot" thing came from the FCC's own mouth, if you can show that you walked the perimeter with a portable radio and confirmed that you're good, that probably shows good faith on the power issue.
...
Absolutely yes a spectrum analyzer is helpful to make sure your harmonic filtering is working. I would put that forth as "good engineering practice" under 15.5, and it's far more than any hobbyist could be reasonably expected to do even just a few years ago; this equipment has gotten mindbogglingly affordable and accessible.
You get a field strength the same way the pros do - from the antenna factors for your antenna and the cable loss. Look at a public FCC test report for an exposition of the method. DIY dipoles and even horns are not going to be very much different from ideal. Calibrated antennas are typically not more than a dB or two off ideal. Dipoles not even that unless they're broken. Horns and log-periodic antennas, enough that it can matter - but fine for hobbyist practice and even most pre-compliance work. I once discovered an RF engineer was using ideal gain numbers for a horn antenna instead of the calibration data for pre-compliance - it fortunately did not make a practical difference. The Roberts balun for measurement dipoles is standardized and I have instructions for building one around here somewhere. That is literally the reference for calibrating other antennas.
The primitive method for getting an antenna calibration is the three-antenna method. If you are particularly motivated, you can build three Roberts dipoles and check them. For a number of reasons, you might want to have three dipole kits anyway if you are doing a lot of these measurements.
"Tricky" in the sense that skiing is tricky. You will struggle learning from books and the internet. Starting from scratch with DIY equipment would be quite challenging. But it's all in standards and the open literature.
I have to admit that having a ham radio license that would be easy to take away is one reason why I wouldn't mess with anything questionable radio-wise.
I wonder though if you could trust the published specs of a commercial FM transmitter to calibrate a power meter on it from a distance of a mile or so...
Basically if anyone receives it that wasn't expecting to it's illegal. That is at the core of the FCC rules in the US, you can't interfere with the lawful use of the band, nor can you cause "interference" on unused bands. That said, the process for prosecuting you is long and somewhat convoluted as they have to file a report, then the FCC will confirm the report, Etc.
But this kind of project (and a similar I saw that broadcasts on every FRS[1] channel at the same time) is a great example of something you can do with SDR that you cannot (easily) do with transistors and boring old analog radio.[2]
It is one of the reasons I enjoy it (playing around with SDRs and DSP) so much as I keep learning new things.
[1] Family Radio Service
[2] I know, I know, building 20 modulators isn't all that tough yadda yadda.
> if anyone receives it that wasn't expecting to it's illegal.
Are there any countries where that isn't illegal?
I don't think so, but it doesn't apply to international waters apparently?
> if you lose the signal when you're more than 200 feet from the transmitter, you're probably in the clear.
With the FCC maybe, but someone far worse will be interested, the RIAA. It's copyright infringement by illegal broadcast.
If it's intended for personal consumption it's fine. It's not copyright infringement to hear my neighbor's stereo if it's a bit loud. It is if he's using it commercially for hosting a block party selling tickets it's a problem.
Ah, but if your neighbor's stereo is faintly audible in the distance while you record a screencast, your video will get taken down off Youtube very quickly.
Nah, don't kid yourself. Tech lik€ this is NOT fine. You're a douche if you deploy this. It may be technically interesting but just a dick move to use it.
> It's copyright infringement by illegal broadcast.
And I'd bet they would make a good attempt at fining you for each channel that you modulated for.
Why do you think they would be interested?
Small, low-power FM transmitters have been around even prior to Mr. Microphone. Hooking to your Hi-Fi so you can get your music on the radio in the garage is as old as Jarts.
And a wonderful April Fool's prank, too!
The main problem with this line of reasoning is that it neglects important parts defined elsewhere making it seem like this is legal when its not. IANAL, but I do hang out with a lot of hammies. You can't help but absorb a lot of technical stuff with them around.
Like for example, https://www.fcc.gov/enforcement/areas/jammers
It need not saturate the bands, you technically are in violation the moment a non-licensed entity has caused interference to a licensed entity.
Willful is a pretty low bar since general intent can be derived from any related negligent acts. Its pretty safe to say transmitting on these bands without a license is illegal, and yes its a fairly big deal.
Most of the public radio spectrum has been monitored, and archived for later retrieval if circumstances dictate since the 60s. There are services that aggregate this data and make realtime info available globally from satellite systems. Trilateration of signals is fairly trivial in most cases.
You may not get caught if you always stay below the noise floor, but you can't really have any real use of the spectrum in doing so, and algorithmic scanning of structured signals is always improving.
It is really negligent to make it seem like these things are legal, when they aren't, and the author should get in trouble for that.
A simple disclaimer isn't going to cut it when then primary purpose is effectively breaking the law in many civilized nations.
How do you interpret Bulletin 63?
https://transition.fcc.gov/oet/info/documents/bulletins/oet6...
For what it's worth (which isn't much), all of that hinges upon the FCC's motivation and capacity to enforce its rules.
If amateur and citizens band radio shenanigans over the last few decades have proved anything, it's that you have to be an absolutely massive pain in the radio-wave-receiving public's ass to really get their attention.
If you had this device on 14.300, 7.200, or channel 6? Meh. No one's gonna care.
In my town there was a guy who was absolutely obliterating 88.5 as well as 88.3 and 88.9 broadcasting some Hispanic hip-hop with some grossly overpowered local transmitter. He overpowered the stations on the entire east side of town. It took over 6 months before it was shut down. I think the FCC penalties are so harsh because their enforcement arm is so small that they can only go after the most blatant offenders. Not that I would want to risk fate given how harsh those penalties are.
I hold a ham radio licence in a couple of countries, although what's happening here is far outside of the remit of ham operations.
I'm pretty certain this is not legal in my locale, but I don't think the author of the article is making this out to be legal - and I don't think they should get in trouble for any of this publishing [assuming you're adressing them, not the commenter]. Perhaps the equivalent of a script kiddy could get up to no good with it, nothing new from the last 20 years of computer hacking then.
I'd far rather see enforcement and ire directed at the proliferation of poorly shielded junk that spews noise all over the spectrum, at surprisingly high power, that seems to have no trouble being sold online or imported for sale in many countries - and is often, strictly speaking, not legal with regards to necessary EMC, and probably safety, standards.
... reminds me of the times I've taken my HT to Southern Cal and heard things on the radio (in the ham band and elsewhere) that I haven't heard any place else, like the ham who would hold court on the repeater with background music.
Eh, if they're so low power that they're not stepping on their neighbor's reception I'd personally say it's fine. Tree falling in the woods situation and all. That definitely varies based on where one is playing around with such things, and it's always important to understand how far your transmissions are actually going (which can be surprising if you don't have much experience with it!).
Someone in a shed on a few acres playing around with an SDR? Go ahead, I don't care. Playing around in an amateur anechoic chamber, have fun if you're sure you're containing things. On a desk in a random apartment in Manhattan? What an asshole, call the FCC.
Is it potentially against the law to do this? Yeah, probably. How bad is it? Somewhere around jaywalking on an absolutely empty street and dumping waste in a park, greatly depending on multiple factors.
Otherwise known as frequency swifting
Yeah, like how...
PSK = Phase Swift Keying
TDM = Taylor Division Multiplexing
Very good.
"... frequency swifting" Get out of here. lol
I remember going to the drive-in movie theater in the 80's. You had to tune your car radio to a specific frequency to hear the movie audio. If this device had existed back then, you wouldn't have needed to tune your car radio at all. I doubt the neighbors would have liked that though. As an aside, I remember being jealous of the neighbors because they got to watch all these movies for essentially free. The irony is that today, we're bombarded by content, it's become essentially free.
> You had to tune your car radio to a specific frequency to hear the movie audio.
This was great because it allowed for stereo. This also replaced the earlier idea of taking the speaker from the pole next to the driver's window, and then hanging from the window inside the car. This was just a mono speaker, but it was also right next to the driver's head. It reduced the drive-in's repair bills from people driving away before placing the speaker to its host stand (and other hooliganism)
*just to add some color
> you wouldn't have needed to tune your car radio at all
You still need to tune it but there are many "center frequencies" that can be tuned instead of just one. So, you wouldn't have to tune far but you would still have to tune :) If you had a knob like most radios in the 80's it would save you from cranking the knob several times to get the dial into the right range before fine-tuning.
I think the theory is that every car that has an FM radio has it tuned to something. So no matter what, as soon as you put on the FM radio, it would have the audio.
Yeah, until you have a 5 year old (probably me in this case in the 80s) that plays with random knobs on the dash when they get in the car. Is it volume? Is it the tuner? We’ll know once the car turns on!
The fancy Christmas lights on houses that sync with music do the same — instructing passersby's to tune to such and such in their cars as they drive past.
It's kinda refreshing that the "Legality" section is entirely concerned with power and frequencies, and not at all about the IP rights around Swift's music.
Who cares? She basically splits Spotify’s revenue with Joe Rogan, she does okay.
Ah, yeah. If this signal reaches your neighbor across the street, then maybe it's a "public performance" and you've got to get ASCAP and BMI licenses.
(And start taking requests.)
I'm going to need a device that allows requests to be made by calling any phone number.
This might help
https://en.wikipedia.org/wiki/Stingray_phone_tracker
Good luck getting anyone to answer the phone to take the request. I'd suggest not naming your station "Spam Risk"
I was hoping this was a Tayloe Mixer[1] based SDR. The Tayloe mixer has excellent image and harmonic rejection.
<Oops... I missed the important detail about 100 different songs... sorry. I'd suggest putting the hit "Never gonna give you up" on loop for that version of the project>
In hardware, it would be easier to feed a 200 khz square wave into a diode and pull off the first 10 mhz of harmonics, low pass filter it, and mix that with 99.1 Mhz fm audio to get 88.1 to 108.1 Mhz fm at 200 khz spacing.
In software, you should be able to precompute a complete one cycle composite of -10 to +10 Mhz sine waves spaced 200 Khz apart. You could then loop this abitrary waveform and multiply it by an FM modulated source at 99.1 Mhz to get the same effect with a lot less processing power. This is the kind of thing GNU radio[2] is perfect for.
<Correcting the above>
Precompute the FM signal for each of the songs in it's channel from -10 to +10 mhz for the length of the song. Store them as 16 bit IQ samples. Each file will be different length.
Loop though all of the songs, adding all of the components into a 32 bit I and Q sum. Round those to 16 bit values, or as appropriate for your SDR transmitter. You should have a very easy to compute 100 channel FM station with each channel separately timed.
[1] https://www.norcalqrp.org/files/Tayloe_mixer_x3a.pdf
[2] https://www.gnuradio.org/
DSP engineer here:- the typical way to do this for better performance would be to do the FM modulation of each channel at a lower rate, say 100kHz, then use an FFT-based channeliser to mix each sub-band to the right carrier frequency and upsample in one step. That helps to break the n^2 problem mentioned in the article.
It would be an interesting exercise for someone to sit down with GRC and implement as many different approaches as possible, and benchmark them. This kind of silly-but-concrete goal can be solid gold for teaching.
From the title I expected it to somehow be related to Taylor polynomials being used to approximate sinusoidal radio stuff.
https://en.wikipedia.org/wiki/Taylor_series
Me too.
The CRT enthusiasts on Reddit would love something like this, but for analog TV channels.
Something that people ask occasionally is how to use their vintage TV's tuner to receive digital TV. But they don't want to tune it to channel 3 and then change the channel with a digital TV box; they actually want to change the channel with the old tuner. The current advice seems to be to spend $hundreds to get an analog CATV head-end, and a separate digital tuner for each channel, but I've never heard of anyone actually following through and doing that.
It's definitely possible to transmit multiple TV channels at once, and I have experimented with that in the past. Since TV channels are 6 to 8 MHz wide (depending on what part of the world you're in) there are practical limits to how many channels can be transmitted at once. Consumer-grade SDR boards are limited to about 56 MHz total bandwidth, which would be enough for 7-9 channels.
It’s not that hard or expensive. CATV analog modulators were common in the past. I had 8 in-home TV channels until our cable company went digital.
there are cheap, new analog TV modulators/broadcasters available on ebay now
This is something that would fit right into some dystopian, proto-cyberpunk movie.
Like, a bunch of activists need to disable comms of some nightshift guards in an area. So... someone figures out how to put more juice into this thing to pump out music on all frequencies they could have.
Bonus points because this wouldn't just jam their communications with white noise. It would be confusing as hell. Taylor Swift on the main channel. Switch to the backup. Irish Folk. Switch to something you and Bob used some time back. Power Metal! What?
> The Taylorator is a piece of software which allows me to flood the FM broadcast band with Taylor Swift's music.
I suppose, you could call it a Taylor Series.
(I'm sorry)
A third voice who assumed it was about that. (there's another headline comment to the same point)
I suppose if I make this comment in sotto voce it IS a taylor sum..
Instead of struggling to generate this in real time, perhaps you could generate it in advance and loop it? The median Taylor Swift song is 3:52, then at 20M samples per second (and four bytes per sample) you need a little over 19GB [1] and I think streaming from SSD should be fast enough. Could probably also compress it.
[1] 20e6*(60*3+52)*4 / 1e9
I'm not sure if Stephen is on HN, but I asked him this question last week[1]. His response was that since the songs are different lengths, a file that was truly loop-able would be super long (if you have a 90 second song and a 2 minute song, you need 6 minutes of audio to create a file that perfectly loops both of them)
Seems like there is probably some set of clever hacks here that could get you around this (although I don't know enough about radio to propose any); I think I asked about pre-computing some state for each song on its own and he had a good response to why that either didn't work or didn't help much, but unfortunately I don't remember it!
[1] we are both currently at the Recurse Center / https://recurse.com
I was thinking you could pick a group of songs that were close to the same length, and then make up the difference with fading, gaps, or a fake DJ? For this to work I think you want your overall loop to be just a few minutes.
This reminds me of a practical application of such tech that never really made it. A chap had his invention on (Australian) ABC's The New Inventors, 20 years ago, named Emergalert. It could announce the presence of nearby emergency vehicles. It seemed like such an obviously good idea.
https://web.archive.org/web/20040923155023/http://www.emerga...
I got my laptop to play Taylor Swift on the AM band - https://www.youtube.com/watch?v=KH9yb1qFKDY by twiddling the memory bus (sounds a bit iffy ;)
Code: https://github.com/anfractuosity/musicplayer
Based on the work of https://github.com/fulldecent/system-bus-radio
> I'm not convinced that I'm operating anywhere close to peak efficiency. There may be some huge DSP-specific shortcut that I'm overlooking - I'm certainly no expert. But the current code works well enough.
Something along the lines of taking the FM at baseband, transform to frequency domain, copy result n times and shift the coefficients to the right indices, much wider ifft to RF should work, right?
Regarding wideband signals muxing/demuxing
Yes, modulate then FFT for each signal, do the upconversion and mixing in frequency domain, then a single large IFFT to give the final bitstream. Roughly the inverse of this (single large FFT, filter in frequency domain, and IFFT individual signals) is how the 'wideband sdr' at utwente.nl works to support hundreds of simultaneous receivers on a single gpu.
The amateur radio folks working with these techniques on HPSDR hardware were calling the technique "Direct Fourier Conversion" but I do not know if they ever got to a releasable state. They spent an awful lot of time prematurely optimizing their stuff to run on the original Jetson board which was not really adequate to the task, so I think that was likely frustrating and killed momentum.
The most efficient way is a filter bank technique. Key words to Google would be "reconstruction", "dechannelizer", "synthesis".
A power of two number (N) of equally spaced frequency channels can be efficiently combined using a short fir filter on each channel followed by an fft where one sample from each channel is input to the fft per frame. Then you get N samples of your output. There is a bit more nuance but the author's bandwidth and number of channels are trivial to handle with this method.
Since each channel is already assumed to be spaced in frequency, you are essentially already in the frequency domain and only so only one fft stage is required.
Ham radio has gotten me into more devious activities than I'd like to admit... but that's half the fun of experimentation :)
Well done :)
From phone phreaking 40 years ago to FPV RC, I've found it to be a good indicator you're focusing on interesting emerging technologies when fellow early adopters in your new hobby say "Technically, this isn't even illegal... yet."
Not even illegal. It’s the new not even wrong!
The algorithm is an amusing nerd snipe. I’m not a DSP expert, but…
The idea of transmitting lots of low-bit-rate signals on different evenly spaced subcarriers is quite popular. OFDM is an example of exactly this, and the computation scales just fine. The same trick ought to work: compute in frequency space. You know what frequency-space signal you want to send on each subcarrier, so assemble the output in frequency space and iFFT it. I bet it can be done genuinely OFDM-style using transform lengths calculated to get the subcarrier spacing right with some additional care to get the boundary conditions right (conventional OFDM has a guard interval and doesn’t even try to transmit a continuous signal).
If all the songs are the same, there’s a much nicer solution. If the complex amplitude of the modulated song is A(t) (that’s just I + i⋅Q), then two copies are:
A(t) + A(t) ⋅ e^(2πit⋅200kHz)
Lots of copies makes:
A(t) ⋅ [1 + e^(2πit⋅200kHz) + … + e^(2πitn⋅200kHz)]
where n is the number of copies. The thing on the right is a geometric series and can be summed algebraically. So all that’s left is to upsample A, and it doesn’t need to be done especially precisely.
This is like a real life version of "I'm broadcasting this message on all channels," or when the aliens invade in a movie and you see the same thing on every TV channel. I love it.
> The Taylorator is a piece of software which allows me to flood the FM broadcast band with Taylor Swift's music.
You don’t need an SDR, just put on her song and most people at traffic lights will roll their windows up!
That being said, from my experience, power will be the deciding factor. I tried both LimeSDR and BladeRF, and when maxing out the power, they overheat really fast, so they probably will only work while driving/cycling so no obstacles. And then again, most people use Spotify or similar in their cars.
If it works on AM, however, that would be great to test.
> Stations will only appear on odd-numbered frequencies, like 88.1 MHz, 94.5 MHz, 107.3 MHz, etc. There's a technical reason for this - every FM broadcast takes up about 150 KHz of bandwidth, and spacing the broadcasts like this allows for an extra 50 KHz of wiggle room.
I believe the odd-numbered frequencies were chosen because if only even were allowed then there would be legal knock-down-drag-out fights over the round numbers.
I don't think it's anything so interesting. The FM band starts at the round number of 88.0 MHz and ends at 108.0 MHz. To divide that evenly into 200 kHz channels without bleeding over the ends of the band, they have to be centered at the odd decimal points.
I noticed a long time ago that the frequencies for AM channels are always divisible by 9 (at least in Australia). Never looked into why (those were the days before the internet...)
Must save it. Some day it might be needed to do some advertising regarding mutiny against government or to gather people in the post apocalypse time...
We need a global deployment of this as a protective measure just in case aliens like the ones in "A Quiet Place" invade.
> an SDR is that a sound card takes real-valued samples, and an SDR takes complex-valued samples
Coming from a DAW background this concept was incredibly interesting. I am curious if stereo audio has ever been represented in this way, or if it is generally just used for the half sample rate?
Actually thinking on this now I guess that doesn't really make sense. But still really interested in this alternate view of sampling waveforms.
Testing has shown that humans can't hear phase differences, unlike radio demodulators, so there's no real use for giving a sound card complex numbers.
(Before someone replies about snare drums, 3:1 rule, etc. — of course if there are phase differences in the audio before you mix it, it will cause audible artifacts in the mixing process. Which is why there's an effect called a "phaser".)
I wonder if Rick Astley will be dismayed or delighted to learn his music has been displaced by Taylor Swift's as default prank audio.
Re: "I wrote a rational resampler which does this by upsampling, linear interpolating, and decimating to the target sample rate."
Where would you source Swifty music that wasn't already at 44.1kHz or 48kHz sample rates? Curiously, what sample rates were these at, and why?
This kind of thing is a perfect candidate for GPU acceleration. It would easily run 10x faster even on a low end graphics card.
I might have missed it, how does OP broadcast across 80mhz with a 20mhz SDR? I think my mental model about SDR capability is wrong, does sample rate not necessarily gate the range at which an SDR can TX/RX?
Really funny idea!
The FM band, from 88 to 108 MHz, is 20 MHz wide.
An SDR's sample rate only limits its bandwidth. Most devices include a mixer, which allows the transmitted or received signal to be shifted in frequency by an amount much larger than the bandwidth.
I like the idea of Rick Rolling the whole neighborhood. It'd be horribly illegal, but it would also be funny.
I am so sad this was not implement in Swift.
Someone please email him the link to Rick Astley’s greatest hit…
I love the concept, long live Taylor
Great insight into frequency modulation! Thanks for sharing.
Covering all frequencies? No Blank Space?
What, no FM Stereo?
And you Taylorate it…
any notable historical events or uses involving this?
Tunnel announcements in your car
[flagged]