Multi-Camera Audio Alignment with Mixing

Get answers to your questions about color grading, editing and finishing with DaVinci Resolve.
  • Author
  • Message

Rick Tillery

  • Posts: 30
  • Joined: Sat Nov 14, 2015 4:57 pm
  • Location: TX

Multi-Camera Audio Alignment with Mixing

PostSat Nov 14, 2015 5:30 pm

This is my first attempt to use Davinci Resolve. I wanted to start with the raw footage of a completed project and see how much better DR would be than my previous tool(s) :D

We used multiple cameras to record a choir concert. One of the cameras was wired directly to the sound mixing board, but the microphone setup was poor, and the system didn't pick up good audio when the singers were at certain places on the stage (e.g. standing in front of the curtain line). For those songs, I want to mix in the audio for one of the cameras to which we attached a directional mic. (We don't want to switch completely to the directional mic, so we can get the accompanist from the sound system and to allow tweaking to avoid audience noise picked up by the directional mic.)

We could have mixed live, but we don't really have the capability. Switching between audio tracks sync'd to the frame is usually the most I've wanted to do with my audio. However, in this case, I want to mix two audio tracks. Unfortunately, 1/60 sec alignment is not sufficient to mix the audio together without the audio signals interfering with one another. (Out of phase interference is sometimes a cool effect, but not what we're after here ;) )

So in my previous workflow, I had to extract the audio from both cameras, decode them to PCM, place them into an audio editor, and then manually align them as close as I could to the highest frequency waveform I could identify in both eye. If the tracks had been of opposite polarity, I could have inverted one here as well. (I know...even if I got this aligned to the nearest sample, there would still be up to 1/48000 sec misalignment, but that's close enough for mostly vocal frequencies.) I then padded one of the tracks to embed the alignment offset and re-imported it into the video editor, where I replaced the original audio and re-sync'd it with the video (using impulse sounds present in multiple tracks--I asked the choir directory to clap, but he was pretty busy with 7 choirs and forgot).

Obviously, this was time consuming, and it likely resulted in plenty of eyeball error. And let's not talk about the possibility of drift, which I didn't even want to think about dealing with. :roll:

Oh, did I mention that I had to do all of this overnight, so the choir teacher could use the video in class for a lesson?

Anyway, the question here is can all of this type of sub-frame (approaching sample-accurage) audio alignment be accomplished with Davinci Resolve?


Return to DaVinci Resolve

Who is online

Users browsing this forum: Benoit Breton, Chris Mierzwinski and 23 guests