Super large FLDIGI modes test

Daniel Miranda

Hi, everyone. I have just finished testing more than 7000 combinations of FLDIGI's digital modes, Linsim's profiles and white noise. It took about four days for my notebook to process what probably amounts to several weeks worth of audio.

The results are in (no need to download anything or run code, just scroll down to see the results). If you find mistakes, please file issues on GitHub.

To make this project possible, I implemented new functionality in FLDIGI (new macros, see my changes at and Linsim (command line arguments, bug fixes, see my changes at and I found a few bugs in the process. If the developers want to merge the changes upstream, please fell free to do so, and let me know if I can help in the process.

I already explained the project I am working in some time ago. These tests are helping me make more informed choices of digital modes before my schedule allows me to travel to implement it.


P.S.: example plots attached. See github page for the full data, including the raw datapoints.

Poul K

Nice work, appreciate it.
Do you consider a commented report also? That could be beneficial when preparing presentations to "spread the word about digital modes".
73 de SA7CND



Thank you for this comprehensive research.

My Shortwave Radiogram project, and its predecessor VOA Radiogram (2012-2017), have also compared modes, although in a less controlled environment. We send  digital modes on shortwave broadcast transmitters. The transmitters are in AM modes, but reception can be in AM, USB, or LSB. Reception is even possible on cheap portable shortwave radios with no SSB capability -- as long as the audio reaches a PC with Fldigi or an Android device with TIVAR or AndFlmsg.

In a shortwave broadcast environment. we have found that MFSK32 and MFSK64 offer the best combination of speed and performance, and they can be used for images as well as text. 

Shortwave Radiogram this weekend will include Olivia 64-2000, Olivia 32-2000, and Olivia 16-2000, as well as our usual MFSK32 and MFSK64. We will also transmit an interesting waterfall graphic. (Reception of at least one of the broadcasts should be possible in Brazil.)

Details here:

See results from listeners at @SWRadiogram or