Bela
Real-time, ultra-low-latency audio and sensor processing system for BeagleBone Black
|
This sketch demonstrates how to read from and write to the audio and analog input and output buffers.
In render()
you'll see a nested for loop structure. You'll see this in all Bela projects. The first for loop cycles through audioFrames
, the second through audioInChannels
(in this case left 0 and right 1).
You can access any information about current audio and sensor settings like this: context->name_of_item
. For example context->audioInChannels
returns current number of input channels, context->audioFrames
returns the current number of audio frames, context->audioSampleRate
returns the audio sample rate.
You can look at all the information you can access in BelaContext.
The simplest way to read samples from the audio input buffer is with audioRead()
which we pass three arguments: context, current audio frame and current channel. In this example we have audioRead(context, n, ch)
where both n
and ch
are provided by the nested for loop structure.
We can write samples to the audio output buffer in a similar way using audioWrite()
. This has a fourth argument which is the value of to output. For example audioWrite(context, n, ch, value_to_output)
.
The same is true for analogRead()
and analogWriteOnce()
.
Note that for the analog channels we write to and read from the buffers in a separate set of nested for loops. This is because they are sampled at half audio rate by default. The first of these for loops cycles through analogFrames
, the second through analogInChannels
.
By setting audioWrite(context, n, ch, audioRead(context, n, ch))
and analogWriteOnce(context, n, ch, analogRead(context, n, ch))
we have a simple passthrough of audio input to output and analog input to output.
It is also possible to address the buffers directly, for example: context->audioOut[n * context->audioOutChannels + ch]
.