How can I correct audio drift when syncing 1 hour long video and audio from separate sources?
I recorded 1 hour long church sermon on video (contains video and audio) using
Sony HDR-CX190 HandyCAM
high definition image quality (HD) = [Standard]: 1440 x 1080/60i quality, AVC HD 9M (HQ)
also recorded audio separately using
Olympus Digital Recorder VN-960PC using High Quality Mode Recording Format ADPCM (WAVE)
Adaptive Differential Pulse Code Modulation (ADPCM) compresses files by saving them as WAVE files with a bit depth of four bits rather than the 16 bits typical of other audio formats. It also uses the built-in flash memory more efficiently by comparing a sound wave to a previous sound wave and digitizing the difference.
I am using Pinnacle 16 video editing software.
After I sync the beginning of the video with the external audio, audio drift occurs after a few mins. of playback.
Looking for any help correcting this.
I have been reviewing many items online that refer to frame rate mismatch, etc. but no close matches for my situation that I can tell.
Any suggestions for avoiding issues like this in future also appreciated.
One suggestions so far was that I needed to edit the audio to adjust for the drift which needs to be done outside the editing software.
Can this be done with Audacity? How?
There is a Change Speed effect. You’ll have to calculate how much it’s off, and it will probably take some trial-and-error. And, you may have to do it in 10 or 15 minute segmments.*
Your video editor may have similar capability (usually you can do some audio editing/processing in a video editor).
I recorded 1 hour long church sermon…
Are you going to have to do this every week? If you can find the “secret formula” (correction factor) it might be fairly simple to change the audio speed by the same amount every week. Otherwise, it might be best to get a video camera with high-quality audio inputs so you can plug into the PA system and record the audio & video together together on the camera.
I’m surprised it’s drifting off after a few minutes, but I’m not surprised it’s off after an hour. clocks (timing) in the camera and audio recorder slightly different. They never will match exactly.** Usually these problems crop-up when recording with a computer and consumer soundcard… Usually, video cameras and digital recorders aren’t too bad. Pros use (expensive) equipment with master clock inputs/outputs to keep everything in sync. In the old film-movie days, they used a clapperboard to sync-up the beginning of each scene, and the scenes were short enough that drift wasn’t an issue.
I would guess most of the problem is the audio recorder, but they don’t publish “speed” or “clock” accuracy specs for cameras or recorders, so it’s just hit-or-miss.
\
If you do it in segments, you’ll generally get better results if you break the video into segments also. Then, re-join them with a short cross fade. Or if you edit the audio in segments, you’ll also want to cross-fade the audio for a seemless splice. If you have to stretch-out the audio so there’s a short gap-silence gap between segments, crossfade some background noise (from another part of the service) over the silence so it doesn’t sound like the sound drops-out.
** It’s not the format(s), and the clocks don’t have to be the same. It’s OK if the recorder uses 44.1kHz, and the video used 48kHz. The software can take care of that as long as the cample rates are “exactly” what they say they are.
If you play them one at a time, who’s off? You can’t just go by the timeline. You have to start one playing and come back in an hour. Is it ahead or behind? Play the other one.
You could have the 48000/44100 problem. Video sampling runs at 48000 and sound/CD sampling at 44100, but I would expect the video editing software to know that. That’s a very common cross.
Isn’t there provision in the editing software to change the sound speed? That’s pretty common, too.
Audacity isn’t going to be able to help until you can figure out what the error is and which direction. Import the sound track and Effect > Change Speed. But it’s going to start asking you questions about percent change and you’ll need to know that.
It could be as simple as changing the sample rate in Audacity and export a new sound file. See if it fits.
Any suggestions for avoiding issues like this in future also appreciated.
You are using two aggressively low-end, consumer grade devices and expecting broadcast/Hollywood sync. Maybe only by accident.
You can get sync points the same way Hollywood does by clapping on camera at the beginning and the end
of the show. Use those as sync points in editing.
Pinnacle allows framerate changes and I would assume it could change sound speed, too.
Forgot to mention I have converted the external audio file from wave to mp3 so I can use in Pinnacle.
I put the video on one track and eternal audio file on track below. Then mute the video audio component and sync video to external audio.
After about 2 minutes I can see the video lip movement falling behind the external audio.
That’s about right for the 29.976/30 frames per second video thing in the US. It misses about a frame per minute. After about two minutes some people can see a lip sync error. After three, it’s obvious to a lot more people. By five minutes, it starts to look weird.
But the editor is supposed to know all about these things.
Import the WAV file into Audacity and Effect > Change Speed by 1000/1001. If that’s clearly wrong, go the other way. That’s the conversion in and out of NTSC (US Video) framerates.