Improving rate switching of DirAC rendering
For rate switching conditions, DirAC renderer often resets the whole renderer and memories if there are changes in number of channels (be it transport or output) or operation modes. This is a valid strategy and seems to work relatively well. However, it has been noticed that as many of the final rendering computations are based on the IIR-smoothed values (_prev variable instead of latest value), the rate switching caused reset can cause gain spikes due to small initial smoothed energy levels. For some reasons, these effects also differ between BASOP and float ending up producing larger MLD differences. <!--- Below are labels that will be added but are not shown in description. This is a template to help fill them. Add further information to the first row and remove and add labels as necessary. -->
issue