Receiving images from space for £20

It’s been ages since my last post – apologies for that, I’ve been busy moving house. Now that that’s out of the way, I’ll be hopefully posting a fair bit in the near future. The laser projector was getting a bit boring so we’ll come back to that soon. In the meantime…

Tim Peake is now on the ISS. Spacex just landed a first stage booster for reuse. I’m in a space mood!

About 18 months ago I became a bit obsessed with the ISEE-3 reboot project. In short, an abandoned space probe launched in the 1970’s was due to return for a flyby of Earth. It was discovered the craft was still functional but no one had the kit to communicate with it again. However, a team did race to put together the necessary kit and software to successfully ‘talk’ to ISEE-3 again during the flyby – amazing.

There is loads more info on that project in the links below. In short, when ISEE-3 was launched, you would need a huge amount of physical equipment to maintain radio communications with a spacecraft (or pretty much anything else for that matter). In the last few years that has changed. A radio receiver can now be replaced by software running on a sufficiently powerful computer. This is called ‘software defined radio’ (SDR), and is exactly what the ISEEE-3 reboot team used to reboot the spacecraft.

The ISEE-3 reboot was the first I heard of SDR, and I had been meaning to have a go at this sort of thing for months. A few weeks ago, I decided I had some free time and decided to have a go. This project can be done in an hour with no prior experience of electronics, radio or programming. There are loads of other tutorials from other people who have done this all over the internet so go ahead and read up on everyone else’s experience with this.

NOAA satellites and APT

Since the 1960’s, the US National Oceanic and Atmospheric Administration has been repeatedly launching polar orbiting earth observation satellites. 

Polar Orbiting Satellites

A ‘polar orbit’ is one which is perpendicular to the direction of Earth’s rotation ie, from pole to pole. The NOAA satellites orbit about once every 90 minutes, so each one passes over most of the Earth’s surface every day.

A total of 19 NOAA satellites have been launched, with new ones replacing those which have become defunct. At present, 3 are operational; NOAA-15, NOAA-18 and NOAA-19. No more will be launched so when these three go belly up there will be no more live APT images (see notes at bottom).

APT refers to the ‘Automatic Picture Transmission’ protocol these satellites use to communicate images back to Earth. These images are used by weather services all over the world to determine cloud cover, temperature etc. As they pass over the surface, they transmit 2 lines of an image back per second. If we can receive this APT signal, we can use a computer to re-construct this image.

What we need to receive the image

I won’t bother drawing a diagram because the setup is so simple. Let’s go through everything in order:


Obviously we need an antenna to pick up radio signals. In this case we need a pretty specific antenna, but it can be constructed from practically junk. Here is what mine is made from:

Some plumbing pipe, a length of TV coaxial cable, masking tape, a broom and a few connectors at the end.

QFH antenna

Seriously, that’s it. It looks like a total joke but it works well. My new neighbors probably think I’m a nutjob when I stick this through the skylight but they’ll get used to this kind of nonsense soon enough. 

Let’s quickly discuss why this works so well. The antenna is tuned to receive signals at 137MHz, which is the frequency used by the NOAA satellites. The transmissions are also Right Handed Circularly Polarised (laymans term – they are twisted around). The antenna type best suited for picking up such a signal is a QFH antenna which is exactly what this is. For anyone interested, there is loads more info on the NOAA transmissions in the links below.

Excellent instructions for building this antenna are given at G4ILO’s Shack (below). His article on this is so good that there is no point in me going over it here – just follow the link.


Radio Reciever

This is the only bit of special hardware need. Long story short, a few years ago, a bunch of folk discovered certain cheap digital TV dongles can be used for software defined radio. The compatible type are designed to recieve European RTL TV. Just plug in your antenna, plug it into the USB and run it through the correct software. I use a NooElec dongle, as recommended by (the go to place for RTL SDR).  It costs about £15 and essentially replaces all the tonnes of kit that used to be necessary for amateur radio.


sdr dongle

My RTL-SDR dongle. Originally designed for digital TV but perfect for SDR.



With the antenna and dongle we can get a signal to the laptop. Now we need some software to do the ‘heavy lifting’ and actually process the signals into something meaningful.

First, we use SDR# (pronounced ‘SDR Sharp’). This takes the signal from the dongle and deconvolutes it into its constituent frequencies. 

Deconvolution is basically the process of turning a raw signal into the individual ‘signals’ (frequencies) which make it up. It is computationally very difficult and is the reason so much equipment as required before SDR came along. More about deconvolution is available at the Wikipedia page. It is pretty complex stuff and no understanding of the process is required for this project.

Next, we need software which takes the deconvoluted signal from SDR# and turns it into an image. There is a free piece of software availible specially for this purpose – WXtoImg. Just go to the website in the links and install it.

We need to send the audio (the signal from SDR# is in audio format) from SDR# to WXtoImg. In Windows 10 this is simple. We go to the audio menu and go to ‘recording options’ we then select the ‘stereo mix’ as the default recording device. Ensure the sample rate is 44MHz. In SDR# and WXtoImg, the recording decive must then be set to correspond with this.

Finally, we need to know when a satellite pass will happen. We need the satellite to be in our line of sight to be able to receive. I go to to predict passes. Orbitron is a great alternative for satellite tracking. It does require you to periodically update the trajectories but the positions can be monitored in real time.


Receiving an image

Here is the process I used:

  • Find an upcoming pass – I go for those with a maximum satellite elevation above 50 degrees.
  • Connect the antenna to the dongle and plug into the USB.
  • Go outside or stick your antenna through a skylight. 

QFH recieve

  • Open SDR# and start playing audio. Select the appropriate frequency (specified by WXtoImg but always around 137MHz for NOAA satellites). The bandwidth is usually around 30KHz.
  • 10 minutes before the pass, open WXtoImg and start auto recording. This will begin recording the audio from SDR# and decoding it into an image as soon as the satellite comes into range.
  • Monitor SDR# -there should be a peak at the satellite frequency as the pass begins. Make sure the selected frequency for recording corresponds to this peak and ensure the bandwidth selected covers the whole peak. As the satellite elevation increases, a distinct chirping is heard. This is the APT signal. I think it sounds like something from a ’60’s sci fi film (it is 1960’s technology after all).
  • Keep an eye on the volume level in WXtoImg. If it falls below green, turn up the gain in SDR#.
  • Keep watching SDR# and adjust the selected frequency as necessary. Some drift in the signal will be observed due to the Doppler Effect as the satellite passes over. The effect is pretty small but it is important to account for Doppler shift to recieve a coherent image.
  • When the pass is complete, you should be left with 2 raw images. One is visible light and the other is infra red. There are all sorts of enhancements available in WXtoImg, such as false colour, temperature etc. Have a play around.

The first few times I tried this, I saw the signal in SDR# but it wasn’t strong enough to make an image from. This is because I tried just sticking the antenna out of a window. As soon as I tried the skylight, I started getting better images. Here is the first one I got (on my fourth attempt):

1212170 NOAA15 2

false colour visible light image of UK – received on 12/12/15.


12121700 NOAA 15

Raw image as taken on 12/12/15


A few other points

  • For any location on Earth, there are usually several passes per day of NOAA polar satellites. So, don’t worry if it doesn’t wok on the first attempt. It didn’t for me!
  • The lines of interference you can see on the images are due to pager signals. Pagers operate around 137MHz in the UK. Why anyone still uses this antiquated piece of technology beats me (I sometimes have to carry one for my work – I don’t understand why they don’t just phone me).
  • The images uploaded here are far from perfect. I’ll upload any better ones in the future if I can improve the quality.
  • There is a series of Russian satellites, called ‘Meteor’ which transmit images at 4 times the resolution of the NOAA orbiters. These use a more up to date digital image transmission protocol, called LRPT which supersedes the APT protocol. I may do a post on receiving from these satellites in the future. Again – many people have already posted about doing this.

Further Reading

Click to access Users_Guide-Building_Receive_Stations_March_2009.pdf  The ISEE-3 reboot project. Not directly relevant to my project but this was my intro to SDR.  Info on the NOAA satellites. There is a PDF available which gives the full details of the APT protocol and requirements for a receiving station.  Great guide to building a QFH antenna for receiving satellite transmissions.  The deconvolution process. Probably only of interest to the mathematically minded!  The go to place for anyone starting RTL-SDR.

As always, just leave any questions or corrections in the comments.


3 thoughts on “Receiving images from space for £20

  1. Pingback: How to build a supercomputer – part 1 | brain -> blueprint -> build

  2. Pingback: Searching for extraterrestrial intelligence with a Raspberry Pi and a neural network – part 1 | brain -> blueprint -> build

  3. Pingback: The Great Disappointment | brain -> blueprint -> build

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s