Interestingly enough, I’ve never bothered (until today) to test how well HourGlass can reconstruct the original sound when using settings that in theory should do so. Achieving such full neutrality has never been a priority anyway. After all, how fun is it to just have the sound go through all that machinery HourGlass or similar software has, only to process with neutral settings to achieve no change at all in the sound…Not fun at all, really!
Anyway, out of interest I decided to try a null test today. The results are shown in the image. The topmost track has the original sample. The second one has the sound gone through “neutral settings” in HourGlass and the third track is the summed result with the second track’s polarity inverted.
The result certainly isn’t perfect, but it’s much better than I would have expected! I am not probably going to work to make more nulling happen, as there really doesn’t seem to be a sensible use-case to make that work perfectly.