View RSS Feed


"Temporal Deblurring", Deconvolution and what MQA Might Be

Rate this Entry

The application of deconvolution to signal processing dates back to work by MIT's Norbert Wiener during the Second World War.

Deconvolution is widely used to deblur signals in both the spatial and temporal domains.

End to end deblurring of an audio signal can be performed by deconvoluting the signal with a measured deconvolution kernel, extending what is the same process as room correction to the entire audio signal chain.

Room correction corrects the spatial domain, and a similar deconvolution kernel can be applied to the temporal domain ... or simply use a multidimensional kernel.


The term "temporal deblurring" has been used as a feature of MQA which promises end to end improvement in sound.

It is certainly possible that the way MQA works is to apply a system wide deconvolution kernel to the music file.

I have no actual knowledge of the details of MQA ... at the time that I am writing this they have not been published, and for all I know never may be. Any relationship to what I am posting and MQA is pure speculation.