András Retzler

Frank, Thank you for your valuable thoughts on this topic. I've just drafted this very quickly, without much research. As far as I can see, I need to look into...

Thanks! Teensy-ConvolutionSDR also looks great!

Edit `PARAMS_NEON` in the `Makefile`.

Currently different compile flags are needed for different architectures. This is not nice, and should be fixed (when I'll have time, but not very soon).

> To give some numbers on its speed, it takes 20% of one core on an Intel Q9400 processing a complex signal at 30 Msps. Wow! Looks great!

Yes, some documentation would help me to better understand the code. There are 3 places where I usually put documentation: * at the beginning of `csdr.c`, where the output of...

Anyway, what you do sounds really great, thanks for working on that!

Hi Michele, I've just added `convert_f_s24` and `convert_s24_f` to deal with signed 24 bit integers. 32 bit floats are already supported (`f` means that data type). Best regards, Andras

You're welcome. Some notes: I might have exchanged the LE and BE formats. If you experience problems, you may try the `--bigendian` switch: ``` csdr convert_f_s24 --bigendian csdr convert_s24_f --bigendian...

This is still undocumented in the README, I should fix that.