[Bela Gem Stereo] [Pure Data] - adc~ analog input issues and workflow problems

Lor93

Hi everyone,

I'm a new Bela user working with the Bela Gem Stereo, and I've run into what seems like a significant workflow and documentation issue regarding analog input reading in Pure Data.

The official Bela documentation suggests using [adc~] to read analog inputs in Pure Data, stating that [adc~ 3] corresponds to A0. However on the Gem Stereo [adc~] does not work correctly for analog inputs; the values returned are nonsensical and do not respond to changes in the physical sensor (in my case, a simple 10kΩ potentiometer with verified correct wiring: 0V and 3.3V confirmed at the extremes with a multimeter).

After some trial and error, I discovered that [bela/analog-in~] works correctly and returns the expected values between 0 and 1. Note that with this object the pin numbering is different: A0 corresponds to inlet 0 (not 3 as with [adc~]).

Using [bela/analog-in~] introduces a serious problem for users who develop patches on a PC (standard Pure Data) before uploading to Bela. The object is not recognized in desktop Pure Data, so no outlets are created, making it impossible to wire up the patch correctly before uploading.

The only workaround I found is replacing a placeholder object with [bela/analog-in~] while keeping the connection structure intact. This is functional but quite cumbersome for something as basic as reading an analog sensor.

Is this a known issue with the Gem Stereo specifically? Is there a recommended workflow for developing Pure Data patches on PC that use Bela-specific objects? Any clarification would be greatly appreciated.

Thanks!

giuliomoro

As the language-overview example says, you can install the abstractions via deken (PureData's built in extension manager) or simply downloading them from here: https://github.com/BelaPlatform/bela-pd-abstractions.

stating that [adc~ 3] corresponds to A0. However on the Gem Stereo [adc~] does not work correctly for analog inputs.
I discovered that [bela/analog-in~] works correctly and returns the expected values between 0 and 1.

That's weird. [bela/analog-in~ 0] on the Stereo simply creates and [adc~ 3] for you, taking care of calculating channel numbers ...

Lor93

Thank you for pointing me to the abstractions library, I had missed it while following the Pure Data course on the Bela website.

Just to add: I was actually monitoring both [adc~ 3] and [bela/analog-in~ 0] simultaneously with a [print] in the same patch, and only [bela/analog-in~ 0] was returning correct values.

giuliomoro

And this is a Stereo, not a Multi?

Lor93

giuliomoro Yes, exactly

giuliomoro

Can you show your test patch?

Then try this patch:

Then open the scope, you should see:

channel 0 (red) and 1 (blue) are exactly the same (they are so much the same that you'll have to disable channel 1 in order to see channel 0 underneath it)
channel 2 (green) is fixed at 0 (being the difference between two identical signals
channel 3 (orange) will also be at 0; disable it to be able to see channel 2

Lor93

OK, I've tried your test patch, and everything's working fine! I don't know why I was having problems in my previous test; my patch was very similar, except that I was printing the results to the console.
Thank you very much for clarifying this and for your prompt reply.