FreeDVNeo verification and a small correction to the procedure. - #78
peterbmarks wants to merge 2 commits into
Conversation
| | Date | 2026-07-28 | | ||
| | radae repo commit hash | 25d58fe30e3765444d8e2a70133f7f39095c50ca | | ||
| | rade_c repo commit hash | Via freedv-backend 5a4210b7548426571b2dd7635e1585dfb7fad098 | | ||
|
|
There was a problem hiding this comment.
@peterbmarks - pls quote the rade_c Git #, freedv-backend is part of the "device under test"
| cd ~/radae | ||
| lpcnet_demo -features wav/all.wav features_in.f32 | ||
| ./build/src/lpcnet_demo -features wav/all.wav features_in.f32 | ||
| # a python environment is needed with torch numpy matplotlib |
|
@peterbmarks - looks like your very close. Give me a call if you want to work on it together. |
|
Thanks David, yes very close. I'm not set up for it today but any tips about optimising RF over cable would be good to share. What did you think of the idea of a live error display from a known test signal of features? |
|
Hi @peterbmarks - rade_c for Linux had some recent bug fixes and diagnostic improvements merged last week - what's your time line for closing out this verification report? If you have RF set up issues, happy to offer phone support. |
|
I'll try to get it OTC done in the next few days. |
|
Hi @peterbmarks - re your question via text yesterday, the latest verification procedure and form can be found in #80 |
2 participants
Currently failing OTC by 0.0009. Will try harder next time.