Empirical Earth · go and get the evidence yourself

Stop taking anyone’s word for it. Go and receive the evidence yourself.

Every argument on this site points at something somebody else measured. This page does not. This page tells you how to point an antenna at the sky and collect the evidence with your own hands, on equipment that costs less than a pair of shoes. No agency serves you a file. No website can edit it. The photons come down and you catch them.

Read this part first. Some of what follows will not do what you want, and you should know which parts before you spend money. Hearing a signal proves almost nothing on its own: anybody can transmit from anywhere. What settles the argument is not the signal. It is the Doppler shift, the timing, and the geometry. Every level below has a box telling you honestly what it proves and what it does not. If a page like this only ever tells you good news, it is selling you something.
The five levels, by cost
  1. Level 0 — free. Hear the Space Station on somebody else’s radio.
  2. Level 1 — about $30. Take a picture of the Earth off a Russian satellite.
  3. Level 2 — about $30. Hear the ISS, and measure its orbital speed.
  4. Level 3 — about $100. Pull the whole round Earth down, live, from 35,786 km.
  5. Level 4 — a license. Bounce a signal off the Moon and wait 2.5 seconds.
  6. The antenna. Build it yourself. The length is the argument.
  7. The software. What it is, where to get it, and what to do with it.

00Free. Right now. On somebody else’s radio.

You do not need to own anything to start. There are receivers all over the world connected to the internet and left open to the public, called WebSDRs. You tune them from a browser. So:

  1. Find out when the Space Station is next over a receiver. Any tracking site will tell you.
  2. Open a public WebSDR near that receiver’s location.
  3. Tune to 145.800 megahertz (MHz), mode FM (frequency modulation).
  4. Listen.

You will hear the Space Station. Voice, sometimes. Data bursts, often. And you will hear it arrive and depart on a schedule that somebody published in advance.

What this proves, and what it does not. It proves that something is transmitting on 145.800 MHz, and that it appears and disappears on a predicted schedule. It does not prove where that something is. You are listening through a receiver you do not own, in a place you are not standing, run by a person you have never met. That is a real limitation and you should hold it in mind. Level 0 is for finding out whether this interests you. If you want evidence rather than a demonstration, you have to own the receiver. Which is what the rest of this page is for.

01About $30. A photograph of the Earth, taken by a Russian satellite, delivered to your garden.

This is the first level where you own the evidence. Buy a software-defined radio dongle. They cost about $25 to $40. Hang a piece of wire outside. Then wait for a Russian weather satellite to fly over, and take its picture off the air.

WhatRoughlyNote
RTL-SDR dongle$25–40The RTL-SDR Blog V4 kit is the usual starting point and includes an antenna.
Antennaincluded, or a few dollars of wireA “V-dipole” cut for 137 MHz. Two rods in a V. That is the whole thing. Build it →
SoftwarefreeSatDump, from satdump.org. Windows, Mac, Linux, Raspberry Pi, Android. What it is and how to use it →

What you are listening for

Russia’s Meteor-M satellites (2-3 and 2-4) transmit their images digitally, in the clear, on 137.9 MHz. Point the software at a high pass, let it run, and an image of the Earth builds up as the satellite crosses the sky. In color.

Check the frequency before you start. It changes. The Meteor-M downlink has moved between roughly 137.1 and 137.9 MHz over the years, and which satellite is healthy changes too. Look up the current status before you spend an evening on it. Any of the satellite-tracking sites will tell you.
And a correction, because a lot of guides are out of date. For twenty years the standard beginner project was NOAA APT on 137 MHz. Those satellites are switched off. NOAA-18 was decommissioned on 6 June 2025, NOAA-19 on 13 August 2025, and NOAA-15 on 19 August 2025. The analog APT service is over. If you follow an old tutorial you will build a perfectly good receiver, point it at a dead satellite, and get nothing. This is not a conspiracy. It is a fleet of satellites that were twenty years past their design life.
What this proves, and what it does not. You now have a picture of the Earth that nobody handed you. That is worth something. But be honest about the limits: the image is low resolution and the satellite is only 800 km up, so you will not see obvious curvature in it, and anyone who tells you otherwise is overselling. What you can see is the geometry. The swath is the right width. The pass takes the right length of time, from the right direction, at the time it was predicted. And the satellite that sent it belongs to a government that has no reason at all to help the Americans lie to you.

02About $30. Hear the Space Station, and measure how fast it is going.

This is the one. This is where you stop collecting demonstrations and start collecting a measurement.

Buy the cheapest handheld radio on the market. A Baofeng costs about $25 to $30 and comes with a stub antenna that is good enough. You do not need a license to listen. Not in the United States, not anywhere. A license is only needed to transmit.

SignalFrequencyWhat it is
Voice / SSTV145.800 MHz FMThe main downlink from the ISS amateur station.
APRS packet145.825 MHz FMData bursts. You will hear them as a harsh chirp.
Repeater uplink145.990 MHzFor transmitting. Needs a license.

Now do the part that matters

Find a pass that goes at least 30° above the horizon. Higher is better. Start listening two minutes before it rises. And watch the frequency, not the signal.

The carrier does not sit still. It arrives about 3.5 kilohertz (kHz) high, near 145.8035 MHz. As the station comes overhead it slides down through 145.800. And as it leaves, it keeps sliding, to about 3.5 kHz low. You have to retune your radio to follow it.

That slide is the Doppler shift, and it is not a curiosity. It is a measurement, and you can check it against physics in one line:

145,800,000 Hz × 7,500 m/s ÷ 299,792,458 m/s = 3,645 Hz
The frequency, times the speed of the source, divided by the speed of light. The number that falls out is 3,645 Hz. The number the ham operators publish is 3.5 kHz. You have just measured the orbital velocity of the International Space Station with a twenty-five-dollar radio.

Now ask what else could do that. A light on a dome does not Doppler-shift. A projection does not Doppler-shift. A stationary transmitter on a hill does not Doppler-shift. Only a thing that is genuinely moving at 7.66 km/s relative to you does that to a radio wave, and it does it whether or not anybody believes in it.

And on a good day it will send you a photograph

Periodically the Russian Service Module transmits slow-scan television, on the same 145.800 MHz, using the callsign RS0ISS, in a mode called PD-120. You do not need a decoder box. Hold your phone next to the radio’s speaker, run a free SSTV app (Robot36 on Android, an SSTV app on iPhone, MMSSTV on Windows), and the picture builds down the screen, line by line, sent to you from orbit. The software, and what to do with it →

What this proves, and what it does not. The Doppler shift is real evidence and it is yours. It cannot be handed to you, faked for you, or edited before you see it, because you measured it on your own equipment by physically retuning a dial. What it does not tell you, on its own, is the exact altitude: for that you need the pass timing and the geometry as well, and the tracking software does that arithmetic for you. But the fact of the motion, at that speed, is not in doubt once you have chased the carrier down the band with your own hands.

03About $100. The whole round Earth, live, from 35,786 km, on a repurposed WiFi antenna.

This is the one that ends the argument about photographs.

The American GOES weather satellites sit in geostationary orbit and broadcast full-disk images of the entire Earth, continuously, unencrypted, at anybody who cares to listen. It is not a press release. It is a radio transmission, and it is aimed at the ground.

WhatRoughlyNote
RTL-SDR dongle$25–40The same one from Level 1.
2.4 gigahertz (GHz) WiFi grid dish~$16Yes, really. The kind people bolt to a roof for internet.
Low-noise amplifier~$30–40A GOES-specific LNA. This one is not optional.
Computera laptop, or a PiFree software: SatDump (satdump.org), pipeline goes_hrit. How to drive it →

Tune to 1694.1 MHz. Point the dish. And the entire sunlit face of the Earth builds on your screen, over and over, all day.

And now the part nobody points at

The dish never moves.
You aim it once. Then you bolt it down and walk away, and it keeps working, for years, without ever being touched again. Ask what has to be true for that. An object must be sitting at a fixed point in your sky and staying there. And the only way anything stays over one spot on the ground is by going around the Earth once per sidereal day, at the one altitude where the orbital period matches the spin. There is no flat-Earth mechanism that parks a transmitter overhead and holds it there. A million television dishes on a million roofs are all aimed at the same empty-looking patch of sky, and not one of them has ever needed adjusting. That is not a photograph anyone can dispute. That is a bracket bolted to a wall.
What this proves, and what it does not. This is the strongest thing on the page, and it still has a limit worth stating. The image you receive is processed on board the satellite: it arrives as compressed data, not as raw photons hitting your dish, so a determined person can still say “the data was faked before it was transmitted.” The unfakeable part is not the picture. It is the geometry of the link. A fixed dish, pointed at a fixed spot, receiving continuously, from an object at an altitude you can calculate from the fact that it does not move. The picture is the bonus. The bracket is the evidence.

04A license, and some patience. Talk to the Moon and wait two and a half seconds for the answer.

This one you cannot do with a dongle, and it is worth describing anyway, because it is the cleanest measurement of the lot.

It is called EME, for Earth–Moon–Earth, and radio amateurs have been doing it since the 1950s. You point an antenna at the Moon. You transmit. The signal travels 384,000 km, scatters off the lunar surface, and comes back. And you hear your own voice, or your own morse, coming back at you out of the sky.

2.4 to 2.7 seconds.
That is the round trip, and it is not a constant, which is the point. It varies, from about 2.38 s when the Moon is closest to 2.71 s when it is furthest. And it varies on a schedule, because the Moon’s orbit is an ellipse and its distance changes in a way that was worked out three hundred years ago. You are timing the distance to the Moon with a stopwatch and a radio. If the Moon were a nearby luminous disc, the delay would be milliseconds, and there would be nothing to hear at all.

And there is a button for it. The free software that made amateur moonbounce practical is WSJT-X, written by Joe Taylor, K1JT, a Nobel laureate in physics who then wrote you some free software. It contains a mode called Echo, and the entire purpose of that mode is to send a tone at the Moon, listen for your own signal coming back, and measure it. It corrects for the Doppler shift on its own. Where to get it, and what the modes are →

It is not easy. The path loss is enormous, the antennas are large, and you need a license to transmit at all. But the barrier is effort and paperwork, not secrecy. Nothing about this is hidden. There are people doing it tonight, on published frequencies, and they will happily tell you how.

What this proves, and what it does not. The delay is a direct measurement of the distance to the Moon, made by you, and it does not depend on trusting a single institution. What it does not prove, on its own, is anything about the shape of the Earth: it tells you about the Moon. Its value here is different. It is that the sky is not a screen. There is a real object up there, 384,000 km away, made of rock, and you can bounce a signal off it and hear it come back.

ANThe antenna. You cut it yourself, and the length is the argument.

Everything above this point runs through a receiver you bought. That is a fair objection, and it deserves a straight answer. A dongle is a sealed box, built by somebody else, running software written by somebody else. If you do not trust the people who make the equipment, then the equipment cannot settle anything for you.

So build the antenna yourself. It is a length of metal. There is nothing inside it to accuse of anything.

And the length is not arbitrary. It falls out of one equation, and that equation has the speed of light sitting in the middle of it.

Wavelength = the speed of light, divided by the frequency.

At 145.800 MHz, the frequency the Space Station transmits on:
299,792,458 m/s ÷ 145,800,000 Hz = 2.056 m

That is the wavelength. Half of it is a half-wave dipole. A quarter of it is one leg of that dipole. Every antenna on this page is a cut of that one number.

Cut the metal to that length and it hears the Space Station. Cut it to any old length and it does not. That is a prediction, it came from physics, and you can test it on a kitchen table with a tape measure and a hacksaw.

Be honest about what this shows, and what it does not. Building an antenna does not measure the speed of light. You use the speed of light to predict a length, and then you find out whether the prediction was any good. That is still worth something. It is a number that came out of physics, handed to you, and tested against a piece of metal in your own hands.

And a receiving antenna is forgiving. A centimeter (cm) either way will not stop you hearing anything. Do not agonize over the last millimeter (mm). The tolerance is loose for listening and tight for transmitting, and nearly everything on this page is listening.

The lengths, straight out of the one equation

Two frequencies matter here. The weather satellites sit near 137 MHz, so take 137.5 MHz as the middle of that band. The Space Station is on 145.800 MHz. Divide the speed of light by each:

FrequencyWavelengthHalf-waveQuarter-wave (one leg)
137.5 MHz weather sats218.0 cm109.0 cm54.5 cm
145.8 MHz the ISS205.6 cm102.8 cm51.4 cm

Now look at the two numbers in the last column, and take them seriously for a moment.

137.5 MHz is the lower frequency, and it needs the longer element. Divide one leg by the other: 54.5 ÷ 51.4 = 1.060. Now divide the frequencies the other way round: 145.8 ÷ 137.5 = 1.060. The same number.

Length falls as frequency rises, in lockstep, because the two of them multiply to a fixed quantity. The fixed quantity is the speed of light. You did not have to take anybody's word for that. It is in the ratio of two pieces of metal on your workbench.

One caution, because it would be easy to overclaim here. That ratio works on the calculated quarter-wave lengths in the table. It does not work on the lengths you will really cut, further down this page, and you should know why before you check it and think we have made an error.

Real metal is not a mathematical line. A rod behaves as though it were slightly longer than it measures, so it gets trimmed a little. How much depends on how thick it is, what it is made of, and how it is fed. The two builds below use different trims, so their cut lengths do not sit in a clean ratio. The physics is in the wavelength. The workshop is in the last centimeter.

Build 1 — the V-dipole, for the weather satellites

This is the antenna for Level 1, and it is the easiest real antenna there is. Two metal rods and a terminal block.

WhatValueNote
Each element53.4 cmTwo of them. Welding rod, brazing rod, or an old telescopic aerial.
Angle between them120°A wide V, not a straight line.
OrientationhorizontalLay the V flat, open side up, and point the axis north to south.
Feed50 ohm coaxCenter conductor to one rod. Braid to the other. That is the whole wiring diagram.

Mount it outdoors with a clear view of the sky. The satellite passes overhead from pole to pole, which is why the V lies north to south: the antenna is aimed at the strip of sky the satellite will walk across.

Why 53.4 cm, when the table above says 54.5? That is the trim. The published figure comes from the people who sell the dipole kit most readers will own, and it works. If you cut 54 cm instead, it will also work. This is a receiving antenna.

What you give up, and what you get back. The satellites transmit in circular polarization. A dipole is linear. That mismatch costs you about 3 dB, and there is no way around it with this design. A quadrifilar helix would recover it.

What you get in return is that a horizontal antenna is deaf to the vertically polarized racket that fills the band around 137 MHz: broadcast FM leaking down, airband, pagers. On a cheap receiver, the noise you reject is often worth more than the 3 dB you lost.

Build 2 — the half-wave dipole, for the Space Station

Same antenna, different length, and straight instead of bent. This is the one for Level 2.

WhatValueNote
Each leg48.9 cmCut two. Wire, rod, or tubing.
Total, tip to tip97.8 cmPlus a small gap in the middle where it is fed.
Feed50 ohm coaxOne leg to the center conductor, one to the braid.

Why 48.9 cm, when a quarter of the wavelength is 51.4? This is the same trim as before, and it has a name: the velocity factor. A radio wave travels a little slower along a wire than it does through empty space, at roughly 95 percent of the speed. So the metal needs to be about 95 percent as long.

Hams have carried this around as a shortcut for a century: the total length of a half-wave dipole, in feet, is 468 divided by the frequency in MHz. Run it: 468 ÷ 145.8 = 3.21 feet = 97.8 cm. The same answer the physics gave. The shortcut is the physics, with the 95 percent already folded in.

For satellites, mount it horizontally. For repeaters and other signals from the ground, mount it vertically instead, because almost everything transmitted from the ground is vertically polarized.

Build 3 — the folded dipole, which is a different animal

A folded dipole is the same half-wave length, but the conductor runs out to the tip and then comes back, making a long thin loop. It looks like a small change. It is not.

Feeding a loop instead of a rod changes the impedance the antenna presents to the cable, and it changes it by a factor of four. A plain dipole in free space is near 73 ohms. Fold it and you are near 300 ohms.

Where the factor of four comes from. The current now has two conductors to flow along instead of one, so it splits between them. The antenna radiates the same power as before. But it is doing it with half the current at the feed point.

Power is current squared, multiplied by impedance. Halve the current and you have quartered the current squared. So to keep the power the same, the impedance has to go up by four. That is the entire trick, and it is one line of arithmetic.

So a folded dipole is not a drop-in for 50 ohm coax. You feed it with a 4:1 balun, or with 300 ohm twin-lead, which is the ribbon cable that used to run to a rooftop television aerial. In exchange you get an antenna with a wider useful bandwidth, and one that is mechanically stronger, because it is a closed loop rather than two rods hanging off a block.

That bandwidth is the reason it is worth knowing about. A plain dipole is sharp. A folded dipole covers more of the band without retuning.

Build 4 — the J-pole, in copper water pipe

This is the one that looks like a real antenna, and it is the one worth building if you build only one. It is copper plumbing. A hardware store has every part.

A J-pole is a half-wave element fed from its end rather than its middle. The end of a half-wave element is a very high impedance point, in the thousands of ohms, and 50 ohm coax cannot feed that. So the short arm of the J, the quarter-wave stub, is a transformer. It steps that impedance down to something the coax can drive. Tap it at the right height and you land on 50 ohms.

The payoff is that a J-pole needs no ground plane and no radials. It is self-contained. You put it on a mast and walk away.

PartLengthWhat it is
A — long side148.0 cmThree-quarters of a wavelength. Measure from the T-fitting to the top.
B — short stub49.3 cmA quarter wavelength. This is the matching section.
D — spacing4.6 cmBetween the two pipes, center to center. Keep them parallel.
Feed tapabout 5 cmUp from the bottom of the stub. This one you adjust. See below.

Cut it from half-inch copper water pipe. You need one T-fitting, one elbow, a propane torch, flux and solder. Center conductor of the coax to the long element, braid to the stub. Wind four or five turns of the coax into a coil just below the antenna, which stops the feedline itself from radiating.

These numbers were derived, then checked against a real antenna. The lengths above come out of the same equation as everything else on this page: three-quarters and one-quarter of a 2.056 m wavelength, trimmed by the velocity factor of bare copper.

A published copper J-pole, built and tuned for 146 MHz, used 1.47 m, 0.49 m and 0.046 m. The equation, run at that frequency, gives 1.478, 0.493 and 0.046. Eight millimeters apart on the long element, three on the stub, and the same to the millimeter on the spacing. The theory landed on the antenna somebody had already soldered together.

An aside on tuning, which matters less than you think

Antenna guides are full of talk about standing wave ratio (SWR). SWR measures how much of the power you send up the cable comes straight back down it, rejected by the antenna. A perfect match is 1:1. Below 2:1 is fine.

Here is the part the guides tend to bury. SWR is a transmitting problem. If you are listening, a mediocre match costs you a little signal and nothing else. Every antenna on this page will hear the Space Station without you ever owning a meter.

It matters the moment you transmit, which needs a license anyway. Power that will not go into the antenna has to go somewhere, and where it goes is back into your radio. On the J-pole, tuning is one adjustment: slide the feed tap up and down the stub, a few millimeters at a time, until the reading bottoms out. On a dipole, trim the tips.

What this level settles, and what it does not. On its own, an antenna proves nothing about the shape of the Earth. It is a piece of metal, and it would work on any world you like.

What it does is close the last door. The objection to everything above was that the hardware is somebody else's. Now the antenna is yours, cut to a length you calculated, and it still hears a station moving at 7.66 km/s with a Doppler shift that only makes sense if it is in orbit. You have taken one more thing out of the hands of people you do not trust, and the answer did not change.

SWThe software. What it is, where to get it, and what to do with it.

Everything on this page is free, and all of it is open source. Nothing here costs money, phones home, or asks you to make an account. Install it, point it at the sky, and it works.

SatDump — the one that does nearly everything

What it is. The decoder. It takes the raw radio signal from your dongle and turns it into a picture. It is the single most important piece of software on this page, and it handles Levels 1 and 3 on its own.

Where to get it. satdump.org/download. There are builds for Windows, macOS, Linux, Raspberry Pi and Android. Take the stable release, not the nightly, until you have decoded one pass successfully.

What to do with it, step by step. It has a pass predictor built in, so you do not need separate tracking software. Here is the whole workflow, because “install SatDump” is not an instruction.

1. Plug the dongle in first, then start SatDump. It looks for the hardware when it launches. Do it the other way round and it will not find the radio.
2. Open the Recorder tab.
3. Choose your device from the list. An RTL-SDR shows up as RTL2832U.
4. Set the sample rate to 2.4 MSPS (mega-samples per second: how fast the dongle digitizes what it hears). That is a safe default and you can leave it there forever.
5. Turn the gain right up, or tick AGC.
6. In the Processing tree, pick the pipeline: meteor_m2-x_lrpt for Level 1, or goes_hrit for Level 3, and the frequency box will jump to 1694.1 MHz on its own.
7. Press Start, then watch the signal-to-noise number, not the picture. You want about 6 decibels (dB) or better. Decibels measure how far the signal stands above the background noise, and below about 6 you are listening to hiss. Below that you will get noise, and no amount of patience will rescue a signal that never arrived. Move the antenna. Go outside. Wait for a higher pass.
8. The image builds while the satellite is overhead. When it is finished, look in the IMAGES folder inside your SatDump directory. That is where they land, and nothing tells you so.

And the one thing that decides whether any of this works. Everything else is fiddling. Get the antenna outside, clear of buildings, and pick a pass that goes at least 30° above the horizon. A high pass with a poor antenna beats a low pass with a good one, every time.

Cost. Free. It is written and maintained by volunteers, and if it works for you, consider paying them something.

An SSTV decoder — for the picture from the Space Station

For Level 2, when the ISS transmits slow-scan television, you need something that turns the warbling audio into an image. You do not need a cable, an interface or a soundcard adapter. You hold your phone next to the radio’s speaker.

PlatformUseNote
AndroidRobot36Free, open source, on F-Droid and the Play Store. Decodes PD-120, which is what the ISS sends.
iPhone / iPadAn SSTV appSeveral in the App Store. Any that support PD-120 will do.
WindowsMMSSTVThe long-standing free decoder. Pipe the radio’s audio into the soundcard.
LinuxQSSTVIn most distribution repositories.
What to do with it. Open the app before the pass. Turn the radio’s volume up and hold the phone’s microphone near the speaker. When the SSTV starts, the app will detect the mode on its own and the image will draw itself down the screen, one line at a time, over about two minutes. Do not touch anything while it is drawing.

WSJT-X — for bouncing signals off the Moon

What it is. The weak-signal software that made amateur moonbounce practical for ordinary stations. Written by Joe Taylor, K1JT, who won a Nobel Prize in physics and then wrote you some free software. It digs signals out from below the noise, which is the only way an echo off the Moon is ever going to be readable.

Where to get it. wsjtx.github.io/wsjtx. Windows, macOS and Linux. Free and open source. Take it from there and nowhere else: there are forks and “improved” builds floating around that are not the official project.

The modes that matter for Level 4. JT65 was designed for EME. Q65 is the modern replacement and is better at it. And there is a third one, and it is the one you want:

Echo. That is a mode in WSJT-X, and its entire purpose is to transmit a tone at the Moon, listen for your own signal coming back, and measure it. It handles the Doppler shift automatically. It is, quite literally, a button that pings the Moon and times the answer.

Also installed with it. MAP65 and QMAP, which correct for the polarization twist the signal picks up on the way (Faraday rotation), and apply the Doppler correction for you.

The odds and ends

You needUseWhere
A radio, with no radioWebSDRwebsdr.org — public receivers you tune in a browser. Level 0.
To know when to look upHeavens-Above or N2YOPass predictions for the ISS and every other satellite, for your exact location. Free, in a browser. (SatDump also predicts passes itself.)
To just listen to a dongleSDR# (Windows)
GQRX (Mac / Linux)
General-purpose SDR receivers. Useful for finding the signal and seeing the Doppler slide on the waterfall with your own eyes.
Nothing elseThat is the whole toolkit.
Watch the waterfall, not the meter. This is the single most useful piece of advice on the page. When you tune 145.800 MHz on any of the SDR programs above, you get a waterfall display, a scrolling picture of the radio spectrum. When the Space Station rises, its carrier appears as a bright line on that waterfall, and you can watch the line slide down the screen as the station passes over. That slide is the Doppler shift. You do not have to take anyone’s word for it, and you do not have to do any arithmetic. You can see it moving.

05What all of this is for

None of these experiments is difficult. None of them is expensive. And not one of them requires you to believe a word that anybody, including this site, has told you.

That is the whole point. The argument about the shape of the Earth is usually conducted as a fight about whose photographs to trust, and that is a fight nobody wins, because a photograph can always be dismissed. So stop having it. Buy a twenty-five-dollar radio. Point it at the sky. And find out what is up there for yourself.

And if you get a result that does not fit, say so. Publish it. Every entry in the main reference ends by naming what would prove it wrong, and this page is no different: a carrier on 145.800 MHz that arrives and leaves at the same frequency, with no Doppler slide, would break the argument in Level 2 completely. It needs a twenty-five-dollar radio and one clear evening. Anybody can run it. Nobody ever has.