Zero latency convolution with convoengine
I'm trying to implement zero latency convolution with the objects in convoengine.
My strategy is to use...
WDL_ConvolutionEngine headEng;
WDL_ConvolutionEngine_Div tailEng;
...with the 'head' engine performing brute convolution on the first 64 samples then the 'tail' engine performing partitioned fft convolution on the rest of the impulse.
I'm finding the output of the tail engine very unpredictable in terms of timing, so it doesn't seem I can just choose 64 samples as the split point as the 'tail' engine mightn't have samples ready in time.
Not to mention the housekeeping with this strategy is getting messy!
Is there an easier way?!?!
Many thanks in advance,
Suarez
|