In prepareToPlay(), as the doc says that estimatedSamplesPerBlock is just a hint and isn’t any kind of guarantee,
I always set the size of my buffer as twice the estimatedSamplesPerBlock :
For now I never had more samples than 2 * estimatedSamplesPerBlock in the processBlock, but I guess the behaviour may change depending on the audio drivers and hosts (?)
Is that good practice ? is it always enough or should I double-check everything is fine in processblock?
How are you dealing with this ?
thanks for your opinion!