imbr

This is not a issue related to ffmpeg or nvidia-ffmpeg or this ffmpeg port even. Read the docs of motion again. Even if a decoder is implemented by ffmpeg doesn't...

Awesome! Just checked it works perfectly! Thanks! Just remember to make a `sudo make install` so the libraries go to the proper place and for example, if you installed using...

in my case it was missing a `apt install pkg-config`

@danboid I tried on this jetson nano image https://github.com/pythops/jetson-nano-image simple LT4 based. That doesn't work for Focal. I will try now going back to Bionic. I will post results here.

@russ-go I am looking forward for your progress, on the same page as you.

just following your effort on using multiprocessing instead of cython

This really interests me. I really would like to see this done. I can help if needed.

@leouieda my only understand of the subject is practical then I would say Monte Carlo. The second part I really got lost.. hehhe so up to you

Thank you @leouieda Well I might probably be missing something. I modified my code and I am just using a portion of my data for testing. I'm using now this...

Thanks! A little more here. The same line above, results that might help ?  Eq. sources densities ``` python figure(figsize=(10,10)) scatter(layer.y, layer.x, c=solver.estimate_, marker=',', edgecolor='none', cmap=mpl.cm.RdBu_r) colorbar() ``` ...