Feature Request: BladeRF 2.0 micro Oversampling and 8 bit mode support for 122.8Mhz instantaneous bandwidth

[nanostructure]

Feature description
The BladeRF 2.0 micro has supported 122.88Mhz sampling rate and 8bit sample mode. ref: https://www.nuand.com/2023-02-release-122-88mhz-bandwidth/
This feature is to support Oversample mode and user selectable sampling bitmode for BladeRF 2.0 micro

• add support for oversampling
• add sampling bitmode selection menu
  -- SC16_Q11_PACKED can support up to 110.0Mhz sampling rate with stability(with oversampling enabled)
  -- SC8_Q7 can support up to 122.88 MHz sampling rate with stability(with oversampling enabled)
• updated sampling rate dropdown list to be more user friendly

Feature Screenshot

suggested code update
source_modules/bladerf_source/src/main.cpp
[removed huge code block spam]

[AlexandreRouma]

The feature has been in the todo list for a while but it has to be implemented properly.

First, having a drop down for the sample type is useless. The only situation in which it is necessary to use 8bit samples is when operating > 61.44MS/s. Using 8bits under that limit is degrading the signal for no reason, and using 16bit over that limit is impossible. It only makes sense to automatically set it with the samplerate.

Second, the reason this hasn't been done is when it was first added, there wasn't a reliable way to detect if the hardware supported it. I'll have to check if that is still the case.

[nanostructure]

Regarding the sample rate:

• Depending on the cable and USB type (2.0/ 3.0) used , it could be possible that the achievable wire speed may not be at the optimum. So it still make sense to use 8 bit mode on lower sample rates to have wider spectrum
• I feel it makes sense to let the user be aware that he is on 8bit mode instead of silently degrade for the user
• With the introduction of BLADERF_FORMAT_SC16_Q11_PACKED format in the 2025.10 release. It is actually possible to run up to 110Mhz with no degradation of signal quality (Tested on my own Mac mini). But again it's up to the wire speed. So it's not necessarily that we should automatically downgrade to 8 bitmode with sample rate > 61.44MS/s

[AlexandreRouma]

Depending on the cable and USB type (2.0/ 3.0) used

The blade is never meant to run in USB2, and any compliant cable will handle the full data rate just fine.

I feel it makes sense to let the user be aware that he is on 8bit mode instead of silently degrade for the user

It makes no sense whatsoever to let the user downgrade performance when it's not necessary or let them choose a setting that cannot possibly work.

110Mhz with no degradation of signal quality

That mode can simply be used automatically when in a certain sample rate range and when a supported libbladerf version is used.

[nanostructure]

Thanks for the advise. I have updated the code

1. hide oversample and sample_format control from end user. User only needs to choose sample_rate
2. indicate on the sample rate dropdown that it's in 8 bit mode
3. add version checking for both hardware and software support checking. (according to nuand releases)

attached source_modules/bladerf_source/src/main.cpp

sample rate dropdown list

running in 122.88Mhz 8 bit mode

main.cpp

[AlexandreRouma]

Cool, but as stated everywhere I don't take code contributions, so I'll implement it myself when I get the time.

Also, I guess you're new to open source but you don't just send a code file like that. The proper way to shares changes is called diffs, look it up.

Also the way this is written makes me believe that shit is LLM slop.