Rhea

A whole bunch of years later I'm experiencing the same problem. No matter how I export the video, spoken word is clipped and disgusting, barely understandable. Mind you this is audio track linked to the video, recorded with GoPro h265. Original audio quality, and the preview playback are both perfectly fine, the issues is just with exported video and this started to happen with the update to version 7.32 and "Automatically normalize audio volume" in the Settings/Media is disabled. I tried exporting it with default settings and also with the highest bitrate on both h264 and mpeg4, nothing changed. (Note that exporting to h265 doesn't work at all.)

So I have extremely simple project with just chopped up clips, single track. Exporting uses only two cores. As a programmer I don't understand why because you can very easily alternate frame numbers on different cores, etc... Here is an idea: * Use (numCores - audioTracks) cores to deal with video. * Use (audioTracks) cores to deal with audio (one core per track) Average computer today has at least 4 cores, not two, and an average computer of a content creator has at least 8 cores, if not 12, not two. Please allow full multithreading on export so that our new AMD chips are not bored =) Here is an idea - open source it and I'll implement it 😇