Well it worked. I’ve decoded NOAA-19 as of the 30th of June from my QTH. I think that’s an achievement. There are still a lot of bugs to work out, most notably rtl_fm seems to be incapable of performing anywhere near the level SDR# or GQRX do at this sort of thing. I’m not sure why, but this image was a total fluke.
I’ve also built a new antenna out of the old dipole I was using, the benefit of using a folded dipole is the material required to build it is a full wavelength, so simply cutting it in half yields two half-wave dipoles for use in phasing. So now I have a fancy looking RHCP crossed dipole.
So hopefully this works out, I have the pi automatically upload all renders to va7eex.ca/wx-up, so hop on over to that directory for my many failures. I’ve also updated the main site to show satellite predictions local to me, because why not.
I’ve been experimenting heavily with my code, its grown significantly since I first began, I briefly experimented with trying to compensate for doppler shift with multiple, shorter, recordings made in rtl_fm then merged together to be rendered, but that didn’t produce any better images of white noise than the normal method so its back to the simple way: One recording of the satellite pass, no accounting for doppler shift.
I’m also planning on maybe making a more permanent antenna out of not-tape measures. Trying to find some light weight metal rods to use with I can bolt to something has been surprisingly difficult. I could just waltz over to a hardware store, but their metal rods are typically quite large and expensive.