Convolution Implementation for ADAU1466

Hello,

I am trying to implement a process that involves convolution in one step, but I am having difficulty. I have made a simple testbed project in SigmaStudio that routes a stereo input through block processing, then out. If I place an FFT algorithm then an IFFT algorithm in line in the block schematic, I get unmodified audio out (as I expect.) When I place an additional input and FFT algorithm, then use the conjugate multiply block to multiply the two FFTed signals together before the IFFT (thereby convolving them in time domain), I get an attenuated, noisy, and badly distorted mess. I get the same results using the conjugate multiplication block. Turning off the "Normalize Input" setting for the FFT algorithms results in an overdriven, noisy, and badly distorted mess. Changing the FFT algorithm type to "Memory Optimized" doesn't improve things either.

What do I need to do in order to convolve two signals accurately?

Audio Signal Processors

SigmaDSP Audio Processors

adau1466