Editing another project that has the same issue as before. Taking clips of video + audio from a GoPro camera and adding background music. Same issue happened with other types of cameras.
Camera audio format is WAV and the music is MP3.
The problem is that the Music track volume fluctuates between and within various clips like Audio Ducking. If I remove the WAV audio track from a clip, the MP3 track gets much louder.
This happens somewhat in preview but it is much more prevalent after export.
What is the content of the WAV audio? Is there road noise or heavy wind in it? The problem being described sounds like destructive interference between the tracks, where the frequencies in one track will cancel the other track due to being out of phase. All Shotcut does is add the values together.
I tried something different for the audio, to see what the results would be and I got no sound at all. Changed to 384k and AC3. Now now audio. Really wondering what is going on.
For compressed audio, that would be Opus or AC-3. But AAC should be very decent under most circumstances too. This problem sounds more like an issue at the processing stage than the codec stage.
Can you share a screenshot of your project? Maybe someone will notice something. For example, maybe you have a compressor our normalize filter on your output track.
You have a very wide dynamic range on the WAV track that the camera has recorded. I recommend turning on the compressor audio filter for this track and adjusting the compression ratio that is comfortable for you. This will allow quiet sounds to sound louder and loud sounds to be muted and quieter, meaning there will be less dynamic range.
That is loooooooud for audio of wind and road noise. It’s like a white noise generator that will destroy clarity in any other signal.
If the music is professionally produced, then it will not be the problem and will not need additional normalization or treatment. The problem is the GoPro audio overpowering the music by way too much.
Options are to compress the GoPro audio like @dimadjdocent suggested, then drop the final gain to a reasonable level. Or just go straight to the Gain filter and make it match the music level better. If you want the GoPro audio to be more prominent than the music, then drop the Gain on both, so that when they’re added together, they aren’t so close to the clipping point.
I’ll have to review and make sure my statement about the video clip audio is correct. I’ll try and remember to post a screen shot of the levels.
The low frequency of the wind and road noise might be a culprit but I have the same thing happening with other cameras where the levels aren’t so high. This is why I’m confused about it.
I applied a high-pass filter at 150hz which seemed to help some but didn’t elevate it all.
As requested, I have a screen shot of the volume max for the Music Audio Track 1st and then the Video Audio Track 2nd at the same frame.
This is for the final song of the video and one of the worst culprits. There was not a lot of wind or bike riding noise in this clip but the level is high.
I used a high-pass filter on it at 150hz which seemed to help. I appreciate the suggestions on the Compression and Audio Levels and will give them a try on the next one as I’ve already published this project.
OK. Nothing else comes to mind. Screenshots can still be helpful. Sometimes when users share screenshots of their UI another person will notice something that the user did not notice.
