Report
Report
Report
Decoding
Lucas Henry
Above is the image that was decoded from one of the NOAA satellites, specifically
NOAA 18. The process followed for capturing this image aligns fairly well with the process
described in Jacopo’s Lair [1], the blog referenced in the project description. This report
describes and details the process of capturing this image, along with roadblocks and
difficulties along the way. It is designed to allow someone else to reliably replicate the results
shown here.
The basic structure of the process followed was as such: First, the necessary
equipment and software was acquired / built. Then, the relevant satellites were tracked to
identify when they would be passing overhead. Then, the data was captured from the satellite
as it passed. Finally, this data was decoded into an image. Each of these steps will be
described in detail in the following paragraphs.
There were several components needed for the following steps. Firstly, a broadband
digital radio receiver was acquired. Specifically, the RTL-SDR is the best choice here. It has a
large array of software available for it and has a large receiving range. The software that was
used in this process is as follows: SDR++ [2]: a software for interfacing with the RTL-SDR,
GPredict [3]: a satellite tracking software that allows accurate location of the relevant
satellites, and SatDump [4]: a software designed to decode captured satellite signals into the
relevant data (in this case images). All of these pieces of software can be downloaded from
the relevant links. The last and most difficult piece of equipment is the antenna. According to
basic antenna theory, dipole length must be at least 468/fMHz in feet. This gives an antenna
length of 108cm, or 54cm for each leg. For this application, as described in [1], the antenna
legs should have an angle between them of 120 degrees. To connect to the RTL-SDR, an
SMA connector was used. This was done by purchasing a male to male SMA cable, cutting
off and stripping one of the ends, connecting the core and shield to two ends of a terminal
block, and connecting two lengths of 8 gauge aluminium wire to the other ends. See the
images below for the details of this construction.
After having all the software installed and components built, the second step was to
identify a suitable satellite pass. This was done using the GPredict software. The three
satellites suitable for this project are NOAA 15, NOAA 18, and NOAA 19. To track them
using GPredict, a new module was created with the three satellites in it and Victoria BC as
the base station. By viewing the future passes of each satellite their parameters could be
viewed. The ideal pass is during the daytime, on a clear day, with an azimuth as close as
possible to 90 degrees. This means that the satellite will be passing directly overhead with a
clear line of sight, and the image it captures will be bright. Using these parameters, a pass
was found where the equipment could be set up in a place with a clear view of the sky.
The next step was to actually capture the pass from the satellite. This was done by
connecting the antenna to the RTLt-SDR, the RTL-SDR to the computer, and the SDR++
software to the RTL-SDR. the RTL-SDR was tuned to the frequency for the satellite tha
passed overhead (NOAA 18 in this case) (137.62 MHz, 137.9125 MHz, and 137.1 MHz for
NOAA 15, 18, and 19 respectively) and the gain was turned up somewhat. The gain was
adjusted so that the signal of the satellite could be seen, the Mode was set to WFM, and the
Recorder to Audio. The antenna was kept away from while recording, as the experimenter’s
body would harm the reception. Once receiving a signal, the record button was pressed and
the volume was adjusted to avoid clipping. The entire pass was recorded onto a wav file and
the recording was stopped once the signal began to decrease in amplitude.
Finally, this wav file could be decoded into an image. The software SatDump was
used for this. This process consisted of pointing the software to the wav file in the Offline
Processing tab, selecting the ‘NOAA APT’ processing pipeline, setting the input level to
‘audio_wav’, and selecting which satellite the capture was recorded from. No other settings
were changed. The output was shown in the Viewer tab, where settings can be changed to
show a map overlay or weather data. One important note is that the current stable release of
SatDump (1.20) is not actually stable, and would crash during the processing step. This was
solved by using the nightly build of the software instead. The image decoded by the author is
shown at the top of this report.
References:
[1] “NOAA POES (15-18-19) satellites reception | Jacopo’s Lair,” www.a-centauri.com.
https://www.a-centauri.com/articoli/noaa-poes-satellites-reception