Preamble
Introduction
Glossary
Mistakes and pitfalls
SDR specific information
1. RTL-SDR
2. HackRF One
Preferred software
Using SatDump
VHF APT/LRPT reception guide (137MHz)
Common issues
Colors of received imagery
Miscellaneous stuff
1. Pass rating scale
2. Why does the horizon appear as a straight line during skew tests?
Epilogue
1. Contacts

Preamble

NOTE: If you’ve been linked here and expect to see anything related to L-band reception, it was moved and can be found here. This was done to make this page seem less daunting, limiting its scope to only include VHF reception.

Before anything, I have to start with some credits:

Lego11’s articles
The Dereksgc Discord
The SDR++ Discord

I learned almost everything you can read here from the places linked above, I can’t express enough gratitude towards everyone who helped me start out with this niche hobby. If you are ever curious about this topic and wish to learn more, be sure to visit these, the people you can meet are incredible.

Introduction

While you have certainly seen imagery gathered by satellites before, did you know that you can receive it by yourself with very little effort? The purpose of this guide is to show you how the broadcasts work, how you can receive them in practice in an easy to understand and approachable way.

How can I receive data off of computers when they’re flying hundreds of kilometers above my head?

Satellites use the same method as your regular old FM station to transmit their data, radio waves! Of course you aren’t able to just tune to a frequency on your car radio and start getting images, this is because it lacks the hardware & software needed to decode and process satellite signals.

To receive satellite signals (among other things, an SDR has a wide range!), you can use a Software Defined Radio [SDR]. This is a device, that - unlike conventional radios - uses software to perform radio-signal processing. You usually plug these into a computer and use software like SDR++ to operate them.

If you ever used a DVB-T tuner, it might have even had SDR capability - some RTL chipsets come with an SDR mode!

The actual reception process is also not as daunting as it might seem! In essence doing so is no different from listening to an FM station, it only happens with different hardware and at a different frequency… with the transmitter flying 800 km above your head. The most difficult part is gaining knowledge about this hobby, which is the reason I wrote this guide.

Why should I do this?

There is no definitive reason for trying this niche hobby out - you can do it for research (The satellites send much more data than just imagery after all), to kill boredom, or just to share the images online - they are quite pretty after all!

What is covered by this guide?

This guide covers weather satellite reception in the easier to receive VHF satellite band (~137 MHz), which contains imagery at a 4 km/px and a JPEG-compressed 1 km/px quality. You can also find a guide on how to receive a more difficult but more interesting L band satellite band (~1.7 GHz) here. This band contains imagery from not just low-earth-orbiting satellites, but also geostationary satellites providing full disk Earth imagery. The prevailing resolution is roughly 1 km/px but can get as high as 500 m/px with American GOES satellites.

These aren’t the only bands that weather satellites broadcast in, but are by far the easiest to receive imagery in both due to physical constraints and commercial availability. Other bands range from requiring a simple LNB (C band) to year long endeavors that require a lot of dedication and individual research. Looking at you, X band!

Glossary

Understanding the terminology used in satellite reception is essential, especially if you aren’t familiar with radio. Below, you can find a list of new terms you are likely to encounter while doing this hobby. Any other unknown terms will have a definition pop up when hovering over them.

Hardware terms

SDR - Software Defined Radio → A device used to translate radio waves into digital data
LNA - Low Noise Amplifier → A tool used to amplify radio signals

SMA → Type of connector used by most SDRs

Abstract terms

Pass → Refers to the time when you can see a satellite passing overhead, used with orbiting satellites
Elevation → Height of a satellite above the horizon
Azimuth → Compass heading
AOS - Acquisition Of Signal → The moment when you start getting a signal from a satellite
LOS - Line Of Sight / Loss Of Signal → Depending on the context this abbreviation is used in either describes your ability to see the satellite, or the moment when you stop getting a signal from a satellite

LEO - Low Earth Orbit → Refers to objects orbiting the earth at an altitude of less than 2000 km

RX → Shorthand for “Receive”
TX → Shorthand for “Transmit”

Software terms

You’ll likely meet these terms when using SatDump and other SDR programs.

SNR - Signal to Noise Ratio → A calculated value describing how strong a signal is, usually expressed in dB
AGC - Automatic Gain Control → Automatically sets the gain based on your reception conditions
FFT Spectrum → The slice of the radio spectrum being sampled by your SDR
FFT Waterfall → A visual representation of the spectrum throughout time, almost always found right below the FFT Spectrum
Noise floor → The bottom of your FFT Spectrum, excluding any radio signals

FEC - Forward Error Correction → Code that tried to fix errors in data, most common type is Reed-Solomon
BER - Bit Error Rate → How many bits were incorrect within one frame of data. Usually expressed in a decimal point (i.e. 0.01 = 1% of data has errors).

Interference → Commonly referred to as RFI (Radio Frequency Interference), is an umbrella term for unwanted signals produced by erroneous sources such as cheap power supplies, HDMI cables, and devices such as laptops (USB RFI @ 480 MHz)
Overloading → Occurs when your gain is set too high and/or you are near a very strong broadcast. Presents as any of the following:
- Your noise floor jumping or being unstable
- Spurs of interference throughout your spectrum
- Signals that shouldn’t be there (i.e. broadcast FM in satellite bands) that don’t move consistently with your tuning direction (If you tune right [up], the signal moves left [down])

TLE - Two Line Element set → A format used to describe the location of objects in space, is used to calculate the position of satellites

DB - Direct Broadcast → A signal that is transmitted constantly, directly from the instruments (Without processing from a ground station)
Rebroadcast → A signal that is transmitted down to earth in a different band/time period, processed by a ground station, then uplinked to the satellite to be rebroadcasted. An example is xRIT.
Dump → A signal that is transmitted to a specific ground station at a specific time - has a beginning and an end

Mistakes and pitfalls

This is a list of mistakes I made that ended up in wasted time, avoid them for the sake of saving you a headache or two

Using old, outdated guides and software → Radio is very niche, the majority of guides you can find online are heavily outdated giving you bad advice and suggesting deprecated software, leading to confusion and subpar results. Please make sure that any sources you use are up-to-date.
Compass tracking with directional antennas → Due to inaccuracies in your phone’s compass and differing real world conditions, you should always try to go by signal strength instead of tracking by elevation and azimuth. Use apps to find the general direction of the satellite, move your antenna around until you see the signal. Once you see the signal, move slow and with purpose to have the signal be as strong as possible.
Doppler tracking when it is not needed → Some older guides feature doppler tracking - shifting the frequency throughout the pass to keep the signal centered. All signals described in this guide do NOT require any such tracking to be performed, doing so can only hurt your reception.

Doing the above is especially hurtful with LRPT, as your machine knows where the signal is by using a costas loop, changing the frequency makes it momentarily lose its lock causing data loss.

Blindly maxing the gain setting → Upping the gain only makes signals stronger up to a certain point, after which it starts amplifying the noise floor much more than actual signals. This can drown them out due to overloading, as well as damage the signal’s quality due to your SDR saturating. Turn it up only until you see, that the signal isn’t getting any stronger. Only up your gain until

SDR specific information

These are things you should know about when using some specific SDRs:

RTL-SDR

These apply to all SDRs using RTL chipsets (RTLSDR blog, Nooelec SMART…)

This band does not require anything more than 2.048 Msps, using a higher setting can lead to sample drops due to instability.

HackRF One

DO NOT USE THE Amp (preamplifier) OPTION!!! The preamplifier is flawed, can get killed very easily. An exception to this is when you’re using Clifford Heath’s redesign, which adds several protections making the preamplifier safe to use.
The HackRF has a fairly outdated design which suffers from heavy phase noise. This means, that you have to receive signals stronger than other SDRs to achieve the same quality.

Preferred software

The best program for decoding data from satellites is arguably SatDump. It is an open source project that has everything you will need, and then some.

Always download the latest nightly build of SatDump, make sure to update it frequently. It’s in active development, has new additions on a daily basis.

To track satellites and figure out their future passes (Most are orbiting the Earth after all), you can use these:

Cross-Platform:

SatDump - Windows, Linux, macOS, Android - SatDump has an inbuilt module you can use for tracking found in the Recording tab. It can only track one satellite at a time, doesn’t do long term predictions.
N2YO - Web - Does the job, however lacks the polish of other apps

PC:

Gpredict - Windows, Linux, macOS - A relatively young tool, arguably the best choice for tracking on your computer
Orbitron - Windows, Linux (wine) - Quite dated but functionally sound

Mobile:

Look4Sat - Android - Provides everything essential in a simple UI.

There are a few apps for IOS, but they have severe limitations. Using any of the above instead is heavily encouraged.

Make sure you update your TLEs every few days/weeks, not doing so can make the satellite locations be outdated or just outright incorrect.

I personally use Orbitron for long term and Look4Sat for short term predictions, the SatDump tracking module while live decoding in case I lose track of the satellite.

Using SatDump

This guide will focus on using SatDump for all recording and decoding purposes. Before we get to actual satellite reception, we have to familiarize ourselves with the SatDump UI. If you already have experience with it, feel free to skip this heading.

Preparation for the pass

Configuration

Before anything, we need to configure a few things in the Settings tab. Make sure to do the following:

General SatDump

Set your location (QTH Longitude, QTH latitude)
Hit the Update button on the Update TLEs line
(optional) Disable Automatically process products

Output directories

Set your live processing directory
(optional) Set your baseband recorder directory

Once these are set, hit the Save button on the bottom of your screen, restart SatDump afterwards to make sure all settings are applied. You shouldn’t have to reconfigure these again.

Offline processing vs Recorder

When you launch SatDump, you initially face the Offline processing tab. This is where you can decode recordings in addition to semi-processed data (raw frames/binary etc.). We won’t be using this tab very often, since we’ll be using the Live processing feature instead, which decodes signals straight from your SDR eliminating the need to make huge recordings prior to getting data. We can find it in the Recorder tab.

This tab might seem overwhelming at first, but don’t worry; we’ll cover it one step at a time.

Selecting source

Expand the Devices dropdown, this is where you will manage your source - where you are getting signals from. SatDump supports several inputs such as baseband recordings and various servers, we will be using your SDR. You can select it as such:

Reload the source list
Click on the dropdown
Select your SDR

If your SDR isn’t showing up, make sure that it is being recognized by your computer and that the drivers are installed properly. If it’s visible by other programs but now SatDump, ensure all SatDump dependencies are installed.

Once you have selected your SDR, you can select a sampling rate appropriate for the signal you are about to decode; this depends on the symbol rate of the target signal. Your sampling rate should be at least roughly twice the symbol rate of the signal. 2.048 Msps is perfectly adequate for VHF signals described here, while 2.4 Msps is enough for most L band signals. Please note that higher sampling rates don’t hurt the signal as long as you don’t begin to drop samples.

I.e. if a signal has a symbol rate of 1 Msym/s, a sampling rate of at least 2 Msps is recommended

You can now press Start, you should see a blue waterfall appear on the right-hand side of your screen (see image below). You can configure the gain slider(s) once you are ready to receive a signal.

UI elements

Now that we have something to look at, we can break the UI down into a few different sections:

Screenshot of the SatDump UI

FFT spectrum - This shows the slice of the radio spectrum being sampled by your SDR. The higher a point is, the stronger that part of the spectrum is.
FFT Waterfall - This shows how the FFT spectrum has changed overtime using a color gradient. By default, blue is the weakest and red is the strongest.
Frequency range - This gives you a scale of what frequencies you’re looking at.
Status bar - Shows what SatDump is doing right now, clicking it gives you a full log.

Adjusting FFT

If we don’t adjust the FFT settings, we won’t see much happening on it; the default FFT range is set up to two extremes: -150 dB and 150 dB. To make sure we can actually see signals on it, connect your SDR to your receiving setup (antenna) and do the following:

Open the FFT dropdown
Select a reasonable FFT size (resolution), 16384 should be more than enough for most use cases
Move the FFT Min slider until the Noise floor (bottom of the sampled spectrum) is just above the bottom of the window
Move the FFT Max slider until you have enough of a range from the Noise floor, that you will be able to see the signal at its maximum strength with a decent margin on top. You can also simply adjust it when a satellite is in view, since it has no effect on reception. Position it, so it’s aesthetically pleasing to look at.

I.e. if the noise floor is at -65 dB, you expect the signal to reach 20 dB, make the maximum -65 [Noise floor] + 20 [Maximum signal strength] + 5 [Margin] = -40 dB
Set the Avg num to anywhere between 10-50, it smoothens the FFT to make it more visually pleasing.

Your FFT should be configured properly now, here’s an example of the same signal at 1691 MHz as seen above, but with a properly adjusted FFT:

Screenshot of a properly adjusted FFT Everything tinted yellow here is a signal. We can now see the shape and strength of it.

Setting up tracking (optional)

If you set your location properly and your TLEs are up-to-date, you can use SatDump’s Tracking dropdown to see the next satellite pass. It allows you to have the satellite’s location in a convenient spot, eliminating the need for a second program running next to SatDump. To use it, choose the Satellites option and select the target satellite from the dropdown. You will now see a summary of the next pass.

The ‘radar’ is like a compass, where every circle is equivalent to 30° of elevation. The closer to the center a satellite is, the higher in the sky it is in the real life, with the center point being 90° - straight up.

A screenshot of SatDump's tracking tab
SatDump showing the next NOAA 15 pass

Using pipelines

Now that we can see what is going on in the spectrum, we can prepare and start a pipeline! Once a pass is incoming, open the Processing dropdown and select the entry (pipeline) corresponding to your satellite pass, in this example I will choose METEOR M2-x LRPT 72k for Meteor M2-3 LRPT.

Once you configure it and hit start, a lot of things will pop up which will be different based on the satellite you are receiving. Let’s break it all down:

Screenshot of SatDump with the LRPT pipeline

On the bottom of your screen you can see the pipeline appear, it separates into two primary sections:

The left side has information about the signal itself - How strong it is, where it is, its constellation
The right side has information about the decoder - If we are decoding a data, if error correction is working etc.

Now let’s look at everything individually:

Constellation - Without going into much detail about radio theory, the constellation is a visualization of the signal you are receiving. It is helpful when determining issues with your reception, is able to tell much more than just by looking at the signal on the spectrum. It should match the modulation found right above it, in this case OQPSK = 4 dots in each corner. The smaller the dots are, the stronger the signal is.
SNR - How strong the signal is expressed in dB. This is just a calculated value, can be wrong! The constellation and BER are the ultimate judges of signal strength.
Viterbi - Without going into much radio theory again, the viterbi algorithm shows whether SatDump knows where the error correction bits are. It is useful when figuring out how strong a signal is (Lower BER = Cleaner signal), or if you are dropping samples.
Deframer - The deframer shows, whether SatDump knows where individual frames of data start. If it doesn’t know (NOSYNC), SatDump isn’t decoding any data, because it has no idea what it is looking at.
Reed-Solomon - This describes the forward error correction bits, where: Green = “Bit is OK”, Orange = “Bit had to be recovered”, Red = “Bit was lost”.

Other things might show up here depending on the satellite you are receiving, such as an image or frame counter. Please note that numbers 3,4, and 5 only show up with signals that have FEC. In this band, that’s only LRPT. APT will have a live image preview of what SatDump is decoding.

This should cover everything you might encounter while decoding these satellites.

VHF APT/LRPT reception guide (137MHz)

Now that we have gone over the terminology behind these satellite and know how to use the necessary software, we can finally move on to the actual reception!

Receiving VHF broadcasts is incredibly easy → all you need is some wire, an SDR, and some patience
As of writing this article, there are 5 weather satellites currently broadcasting imagery in this band
While easy to receive, they have a relatively low quality (4 km/px on APT and 1 km/px with JPEG compression on LRPT) and transmit only 2-3 channels while broadcasts in higher frequencies usually transmit 5+ raw, 1 km/px channels

Sample processed APT and LRPT images

Note: The raw images don’t have maps on them, they were added in post-processing.

A processed APT image NOAA 18 APT received on 02/01/2024 using a 5 element yagi-uda antenna. Processed using SatDump with the HVC RGB composite. Equalized, median blur applied. 65% quality lossy JPEG compression with 0.05 gaussian blur applied, click here for the full resolution image.

A processed LRPT image Meteor M2-3 LRPT received on 02/01/2024 using a 5 element yagi-uda antenna. Processed using SatDump with the 221 RGB composite. Equalized. 65% quality lossy JPEG compression with 0.05 gaussian blur applied, click here for the full resolution image.

Detailed satellite information

NOAA

There are 3 NOAA satellites broadcasting imagery in this band, NOAA 15, 18, and 19. These are the only remaining members of the POES (Polar Orbiting Environmental Satellites) constellation. They were launched in 1998, 2005, and 2009 respectively.
Have an analogue APT (Automatic picture transmission) broadcast that transmits two channels at a 4 km/px quality. Its analogue nature means, that crackling (noise) during the recording translates to grain in the output images.
These satellites also broadcast an auxiliary DSB (Direct Sounder Broadcast) signal which contains the HIRS and SEM instruments as well as telemetry. I won’t be covering it in this guide, as it’s primarily focused on visible imagery.

During day these transmit one visible and one infrared channel, during night they switch to two infrared channels.

New satellite launches are a part of the JPSS constellation, which only include an incomparably harder to receive X-band signal. No future satellite launches from NOAA are planned to include a VHF antenna.

NOAA APT screenshot from SatDump NOAA 19 APT, Credit: SigIdWiki

METEOR-M

There are 2 satellites currently broadcasting imagery in this band, Meteor-M N°2-3 and Meteor-M N°2-4 (colloquially referred to as Meteor M2-x or just M2-x for short), both a part of the Meteor-M constellation. They were launched very recently - in June 2023 and February 2024 respectively.
These satellites have a digital LRPT (Low rate picture transmission) broadcast that includes 3 channels at a JPEG-compressed 1 km/px quality. It includes FEC to make sure the picture doesn’t come out grainy as well as allowing you to decode the signal properly even if it’s fairly weak.

Reception note: M2-3 LRPT fades significantly when facing away from you at lower elevations.

These transmit any 3 channels the operators choose, as of the latest commit:

Satellite	Channel numbers	Channel types
M2-3	1, 2, 3	3×Visible
M2-4	1, 2, 3	3×Visible

Meteor LRPT screenshot from SatDump Meteor M2-4 LRPT. Spike on the left isn’t a part of the signal.

This satellite series has been plagued with accidents, faults, and delays: Meteor M1 and M2 lost altitude control, M2-1 exploded on launch, M2-2 got hit by a micrometeor making it unable to broadcast LRPT, M2-3’s LRPT antenna didn’t fully extend, leaving it in a tilted angle making the signal improperly polarized, experience random drops as well as making it generally weaker than it is supposed to be.
As a long awaited change of luck, M2-4 has successfully launched on leap day in 2024, is broadcasting LRPT at a full strength. Its JPEG compressor has had a few hiccups presenting as a lot of artefacting in its imagery, but it has always recovered after the satellite restarted it a few orbits later.

Broadcast issues

NOAA 18 has had a configuration error present ever since its management was transferred to Parsons tech, making it broadcast a visible channel on APT during the night instead of an infrared one. This presents itself as half of the APT image being black during nighttime.
Meteor-M N°2-3 has an incorrectly deployed VHF antenna, making the LRPT signal weaker than intended and unexpectedly experience drops at certain elevations.

NOAA 15 has had several major hiccups with its scan motor current spiking due to it grinding through debris. These spikes caused a loss of synchronization between the scan motor and the processor presenting as major glitches (NOAA 15 received on 13/10/2023 using a V-dipole) appearing in place of actual imagery. A large enough spike could lead to a complete motor stall, from which recovery would be highly unlikely. The satellite has since recovered and is currently broadcasting fine. Its status can be checked on the STAR ICVS monitoring page by selecting Telemetry and Scan motor current.

Hardware needed to receive these satellites

You will need an SDR and an antenna, no other special equipment is required for VHF reception.

NOTE: You also need the appropriate cables and adapters to connect antenna to your SDR, make sure to order these in advance. An example is a coaxial cable and an F female to SMA male adapter.

SDR

Most SDRs should be able to natively sample this band without any issues, but you should ideally look for reputable brands to avoid poor results or unexpected issues. A great starter stick is the $30 RTLSDR Blog v3/4.

Antenna

You have the choice between:

Omnidirectional antennas - Don’t need tracking to be performed (remain stationary throughout the pass)
Directional antennas - These need tracking to be performed during the pass, can only receive one satellite at a time, generally have a higher gain leading to better signal strength

Popular antenna types for receiving signals in this band include:

1. V dipole antenna

Omnidirectional
Very easy to make, very portable
Fair results, has some nulls due to its inconsistent radiation pattern
Arguably the best for beginners
When using, point directly north/south & hold it horizontally about 50 cm from the ground (Personal tip: Play around with it and figure out when the signal is the strongest, stick with what works!)

To build it:

Get a chock block (electrical terminal), a coaxial cable and two preferably copper, unshielded wires that are ~54.5 cm long
Stick the shielding of the coaxial cable in one hole and the copper core into the other (keep the core run as short as possible)
Screw the two wires into the holes and spread them 120° apart making a V shape. That’s it. Really.

Note: This antenna can be used for permanent fixtures, but should have a proper ground plane. A good example is a V shaped reflector about 1/4 wl below the dipole, a bit larger than the V dipole itself.

A visual guide on how the antenna should look
This image suggests a wire length of 53.4 cm, while that would work the actual length should be approximately 54.5 cm. This is because the v dipole elements be a fourth of the target wavelength, which in this case is 299,792,458 ms⁻¹/137,500,000 hz = ~2.18 m; 2,18/4 = ~0.545 m = ~54.5 cm. Why? It’s how a v dipole works. This image also suggests using aluminum rods, while that’d work, copper is about twice as conductive and will therefore lead to better results.

2. Yagi-Uda antenna

Directional
Fairly easy to make, but is rather bulky
Very good results thanks to its high gain
Requires manual tracking

An image describing the parts of a yagi antenna
Source

To make it:

Get something long and non-conductive, this will be the base for the yagi and is called a “boom”. Afterwards, figure out how many elements fit on it based on its length.
More elements make the antenna:
- More directional (Harder to track)
- Have a higher gain (Better signal strength)
- Have a longer boom length (Bulkier)
Shove appropriate numbers into this calculator (For frequency choose 137.5 MHz)
Cut copper wires to length, place them onto the boom according to the values from the calculator
For the driven element (dipole), cut it in half and put one side in a terminal with the shielding of a coaxial wire and the other with the core of the same wire
Make sure the cables don’t touch or are short together in any way, this will make the antenna not work
Coil the coaxial cable up a few times right after the feed point in order to convert the unbalanced signal to a balanced one (In essence creating an HF choke)

The choked balun and dipole feed I use on my yagi

3. Quadrifilar helical antenna (QFH)

Omnidirectional
Fairly difficult to build
Very good results thanks to its circular polarization and consistent radiation pattern
Best choice for permanent fixtures

Sample QFH antenna
Source

A good guide I have followed to make this antenna can be found here, while being in Spanish it offers great information about the dimensions and the installation. Google translated page

WARNING! This antenna has a circular polarization, meaning you HAVE to match it with the satellites’, that being RHCP!!! (For APT & LRPT) Correct QFH polarization Bottom to top, RHCP is twisted clockwise, while LHCP is twisted counter-clockwise. CC for image: muellermilch.de on Discord

Using the incorrect circular polarization presents itself as the signal being weak, fading significantly throughout the pass.

Other choices

This list isn’t exhaustive, different antenna types such as (including but not limited to) turnstiles or quads can also be used. Images of a turnstile and a quad antenna Turnstile on the left, source. Quad on the right, source

Frequency reference

Please note that you don’t need to know these by heart, SatDump offers a frequency drop-down with every pipeline. These are listed in case you use a different program that requires you to tune to their frequency manually. As of the latest commit, the frequencies the satellites transmit at are as follows:

Satellite	Frequency
NOAA 15	137.62 MHz
NOAA 18	137.9125 MHz
NOAA 19	137.1 MHz
Meteor M2-3	137.9 MHz
Meteor M2-4	137.9 MHz

Actually receiving the satellites!

Get to a place with a good view of the sky - The more you can see, the longer you can receive the satellite for and the longer the resulting image will be. VHF signals should easily pass through dead trees and some foliage.
Open SatDump and navigate to the Recorder tab, select your SDR, set the sampling rate to an appropriate value (~ 1-2 Msps for this band) and hit Start
- Rise the gain until your noise floor starts to rise more than the target signal, or until your SDR overloads, whichever comes first. Try not to move it around much during the pass, it will lead to the image having sections with a different brightness with APT. Doing so has no effect on LRPT.
In the side panel, open the Processing menu and do the following:

FOR NOAA APT
- Select the NOAA APT pipeline
- Enable DC Blocking and SDR++ Noise Reduction
- Select the proper NOAA satellite
- Open the frequency menu, select the correct satellite
Example configuration for an NOAA 19 pass

FOR METEOR-M LRPT
- Select the appropriate pipeline:
  - METEOR M2-x LRPT 72k for Meteor M2-3 and Meteor M2-4
  - METEOR M2-x LRPT 80k for <x>
- Enable DC Blocking
- (Optional, preferred) Enable Fill missing data
- Select the appropriate frequency:
  - Primary for 137.9 MHz
  - Backup for 137.1 MHz
Example configuration for an 80k bitrate Meteor-M pass. Note that the satellites usually transmit 72k instead.
When the satellite comes into view, press Start on the processing window

When using a directional antenna, move slowly and try keeping the signal as strong as possible using the SNR with LRPT and the FFT along with the sound (trying to avoid crackling) with APT. You might not get the hang of it on your first try, tracking is a skill you have to learn!

If everything is right, you are now receiving a beeping APT signal or see four dots on the demodulator in case of an LRPT signal!
SatDump mid LRPT decode. See the SYNCED viterbi and deframer.
Once you see the signal has completely disappeared and/or the satellite has passed under the horizon, press Stop on the pipeline
SatDump will now begin automatically processing the results, you can see the progress on the status bar and the Offline processing tab. (You can disable the auto-processing in the settings if you want to tinker with the images yourself and don’t want SatDump creating automatic images)
Once it finishes processing, head to the Viewer tab and select the pass you decoded on the top left

You are done! Feel free to play around with the image settings and enhancements, you can figure it out :)

Make sure to check the common issues header if you experience any trouble with the decode/product processing.

Alternative LRPT setup (SDR++)

As an alternative to directly recording in SatDump, SDR++ can be used to record an LRPT pass as well. This is especially useful when you aren’t fully certain of the satellite’s bitrate. To use SDR++ to record the LRPT pass, do as follows:

Open SDR++ and start the radio
Open the Module manager and configure as follows:
1. Type any name you want into the left window and select meteor_demodulator in the right window
2. Click the ‘+’ icon
3. Tick OQPSK in the newly added module
4. Hit record when a satellite is in sight
Move the demodulator over to the correct frequency
Receive the LRPT signal as usual, once the pass is over hit Stop
Open SatDump, move to the Offline processing tab and configure appropriately:

Make sure to select soft as the input level!
Hit Start

Your LRPT pass should now decode as usual.

Common issues

Generic

There is a map but no clouds are present

You likely selected the MCIR/MSA RGB composites, which overlay a predefined map over your image. If your original image was just static (no actual signal was decoded) or solid black (visible LRPT channels at night), all you are left with is the predefined map. You can confirm this by looking at the folder with the products, in the case of APT look at the raw_sync.png file, it’ll likely be full of static. In case of LRPT look into the MSU-MR folder, you are likely going to find 3 solid black images.

The signal is fading a lot near the horizon

TODO: image

This is normal, happens due to atmospheric phenomena.

APT specific

There is grain all over the resulting image

APT is an analogue transmission format, so what you hear is exactly what you get. If you heard any crackling during the decode, it will have shown up as grain on the resulting image. The most common reason for this is that the signal was simply too weak. This can be remedied by using the Median blur option.

There is a thin line in the middle of the image, even though the signal sounded clear

A crop of an APT image with a thin line in the middle

This line is caused by APT passing through the center of your FFT spectrum, which has a thin DC spike created by your SDR merging together the two sampled radio branches. APT’s analogue nature makes this thin spike be interpreted as part of the APT signal, which can create erroneous lines in the image. You can easily avoid this by using a Frequency offset, then tuning the value you typed into the field away from the target signal. Use enough of an offset, so that the APT signal doesn’t travel through the middle of your spectrum, 100 kHz is usually enough.

I.e. with a frequency offset of 100 kHz, tune 100 kHz above the APT frequency. For NOAA 15 broadcasting on 137.62 MHz, that would be 137.72 MHz.

There were occasional bursts of crackles, the image has several small lines of static in it

A crop of an AOT image with several small static lines in it Small lines are most visible on channel A (left)

This is caused by interference from VDL, a data transmission format used by planes in the neighboring aviation band. Because of its strength and proximity to this band there is sadly no consumer-level remedy for this other than moving away from areas where planes frequent in.

LRPT specific

The image is solid black

If the image is from LRPT on an evening pass, you need to select an infrared channel (4/5/6) to see imagery - Channels 1/2/3 are all visible, so during night they are solid black. If the satellite was in 123 mode, you won’t be able to see any imagery from this decode. Sorry!

I saw an LRPT signal but got no SNR and/or NOSYNC on the viterbi

SatDump has recently had a period where the demodulator got modified in a way that has caused issues with locking onto the LRPT signal. This was fixed in a later update, should be fixed by simply updating SatDump.

I have noticed this issue happening in the recent past as well, will investigate when able

The imagery I received is blocky and/or corrupted

An image of the 221 LRPT composite with extreme damage
Meteor-M N°2-4 received by t0nito on 31/07/2024
A crop of Meteor-M N°2-3 received by KentuckyFriedData on 05/02/2025

This happens when the JPEG compressor on the satellite gets stuck or glitches. The satellite has a watchdog that will restart it after a while, until then there’s nothing you can do.

Colors of received imagery

After receiving these satellites and checking the resulting images, you will notice, that they are all black and white even though the images you see online are colorful. The colored images are created by processing the raw channels into RGB composites - applying different channels to different colors. You will quickly find the term False Color being thrown around, but how can color be false? To understand why this is the case, we have to look at some color theory and instrument descriptions:

How is color perceived?

Our eyes sense different colors by sensing different wavelengths that objects reflect using three types of cone cells:

Short → 380 - 540 nm, corresponds to Blue
Medium → 440 - 670 nm, corresponds to Green
Long → 500 - 690 nm, corresponds to Red

Color spectrum showing the S, M and L wavelengths

Satellites work similar to our eyes, but instead of having different cones for different wavelengths they have different channels.

But why are the images in black and white when we see color? Because we are looking at individual channels (Equivalent of only one type of cone at a time) instead of a mix of channels as perceived by our brains.

Why is it called false color?

Put simply, the term ‘false color’ refers to an image that does not represent what you would see with a naked eye. We can’t see electromagnetic waves that have a lower or higher wavelength than RGB (400-700 nm). Due to technological constraints and a focus on scientific value, earlier satellites’ instruments usually didn’t bother sampling all color bands, instead receiving other wavelengths that are more important for research such as different infrareds.

As an example, these are the channels present on the AVHRR/3 instrument flown on MetOp and NOAA POES satellites:

Channel number	Wavelength	Description
1	580-680 nm	Visible
2	725-1000 nm	Near infrared
3A	1580-1640 nm	Near infrared
3B	3550-3930 nm	Middle infrared
4	10300-11300 nm	Thermal infrared
5	11500-12500 nm	Thermal infrared

As you might see, we only have two channels covering the visible spectrum (Ch2 partly covers it), meaning we can’t get the actual colors the sensor could have seen from the data it collects - True color isn’t possible on AVHRR/3.

But how do we get colored composites if we can’t see the actual RGB wavelengths? We can assign arbitrary wavelengths to the R, G, and B channels, this results in a False color composite. These are useful in cases, such as highlighting things that might not be visible in the true color spectrum. They’re also prettier than looking at raw black and white images.

Let’s take the 221 RGB composite as an example, it assigns channel 2 (centered at 630 nm) to the red and green output channels, and channel 1 (centered at 862 nm) to the blue output channel. This does not represent the actual RGB color wavelengths, hence it isn’t true color.

Example of the 221 false color composite A crop of NOAA 18 received on 02/03/2024 using a 125 cm dish and a SawBird GOES+. Processed using SatDump with the 221 RGB composite Median blur applied, equalized. 65% quality lossy JPEG compression with 0.05 gaussian blur applied.

Of course there are much more complex composites such as the NOAA Natrual Color Composite which applies channels using the following formulas:

Color	Formula
Red	`ch1 < 0.065 ? (ch4 - 0.7) * 2.66 : ch2 * 2.2 - 0.15`
Green	`ch1 < 0.065 ? (ch4 - 0.7) * 2.66 : ch2 * 2.2 - 0.15`
Blue	`ch1 < 0.065 ? (ch4 - 0.7) * 2.66 : ch1 * 2.2 - 0.15`

You can play with the different composites and see what looks the best for your use case.

Do any satellites broadcast true color?

Yes! New satellites often have channels that individually sample R, G, and B wavelengths. For example, here are the channels of the VIRR instrument flown on FengYun 3 [A/B/C] satellites:

Channel number	Wavelength	Description
1	580-680 nm	Visible
2	840-890 nm	Near infrared
3	3550-3930 nm	Middle infrared
4	10300-11300 nm	Thermal infrared
5	11500-12500 nm	Thermal infrared
6	1550-1640 nm	Short wave infrared
7	430-480 nm	Visible
8	480-530 nm	Visible
9	530-580 nm	Visible
10	1325-1395 nm	Near infrared

Channels 1, 7, and 9 sample R, B, and G wavelengths respectively; this makes them viable for a true color composite! In this case, the composite is 197 - 1 to red, 9 to green, and 7 to blue.

A true color image from FengYun 3C FengYun 3C received on 29/03/2024 using a 125 cm dish and a SawBird GOES+. Processed using SatDump with the 197 RGB composite. Median blur applied, equalized. 65% quality lossy JPEG compression with 0.05 gaussian blur applied.

Miscellaneous stuff

Pass rating scale

This scale is not official or perfect, I just created it to be able to gauge how good each xRPT satellite pass was. Which metric to use is debatable, but I personally prefer how many kilometres of the ground track you received from the satellite. Given Meteor-M's spatial resolution of approximately 1 kilometre, you can use the vertical length of raw MSU-MR images from your LRPT pass to get the amount of kilometres you tracked

For example: /path/to/product/MSU-MR/msu-mr_1.png, has a 2,048 × 4,750 px resolution. This means, that the resulting image has approximately 4,750 km of the satellite’s ground track.

Kilometres tracked	Pass quality
<250	Very poor
250-1000	Poor
1000-2000	Okay
2000-3000	Good
3000-4000	Very good
4000-5000	Excellent
5000-5500	Almost perfect
>5500	Perfect

The theoretical limit for this is about 6000 km, which has been received by some by using atmospheric phenomena to receive the satellite down to just below the horizon.

Why does the horizon appear as a straight line during skew tests?

If you received Meteor M2-4 in the few months following its launch, you might have captured one of its skew tests! While performing these, the imager turns a perfect 90° to its side.

Meteor M2-4 skew test picture Meteor M2-4 received on 21/04/2024 using a V-dipole. Processed using SatDump with the 221 RGB composite. Equalized. 65% quality lossy JPEG compression with 0.05 gaussian blur applied.

You might be puzzled to see that the horizon appears as a perfectly straight line. Does this suggest that the Earth is flat? Not at all!

The key to understanding this phenomenon lies in how weather satellites transmit this data. All signals described in this guide are direct broadcasts, meaning they send the image line-by-line rather than capturing and transmitting an entire, instantaneous picture. This process is akin to how a scanner works, scanning each line individually.

Thanks to the satellite itself always pointing straight down and the nature of the line by line broadcast, no curvature is present on the resulting image. Sorry flat earthers!

Epilogue

This guide has covered everything you need to know about VHF reception of weather satellites, if you want to learn how to use the data you gathered for scientific purposes, you can look at Lego11’s guide on weather prediction from the data you receive.

If you crave for something more interesting while remaining relatively easy and “5+ raw 1 km/px channels and full disk Earth imagery” piques your interest, you can check my L-band reception guide out! Thanks to Nooelec’s SawBird GOES and people commonly giving dishes away after switching to terrestrial TV, this band usually costs no more than 50 bucks to set up.

Contacts

If you have any questions or remarks regarding this page, feel free to message me either on Twitter (@Meti172172), or on Discord (@cpt_dingus).

I hope you had fun reading this article, now go get some pretty pictures!

Table of contents