ivofx I am wondering how many voices I can expect to be possible if I was to write a 16x multi-timbral sampler with a filter for each instance and 3 global effects. Are projects with similar scope I can look at?
these specs are very generic. Probably the only way to know the answer is trying to run the code. If you have a stand-alone app that uses ALSA and whose dependencies are satisfied by using apt-get
on Debian Stretch, then it should be fairly straightforward to build and test it. Note that normally Bela does not run on ALSA, but it CAN run on ALSA - without low-latency guarantees - and this can be useful for quick benchmarking/feasibility study.
ivofx How smoothly is JUCE supported in the first place?
I only built a couple of JUCE projects, maybe two years ago. Here it is: https://github.com/giuliomoro/belaJuce
That was a terminal application only, no GUI. It used to work ok. I remember very long compilation times due to the huge header files, so I would recommend you pick at least one of the following: use pre-compiled headers, use distcc
, or cross-compile.
ivofx Can I control/connect SPI/I2C TFT screens?
In principle yes.
- I2C: there is an I2C bus fully available
- SPI: on BelaMini there is a hardware SPI bus fully available. On the regular Bela there is only soft SPI available, which tends to be limited at 100kHz, and very CPU intensive.
However, I have no idea ho QtEmbedded works. Does it render to memory so that you can send the buffer to an SPI/I2C display? Note that the GPU on Bela is not usable at the moment, though Beagleboard is working towards supporting it (see here).