GH3 Having Trouble Syncing with .Wav file from External Recorder

[Ianscreations]

I am using a backup Digital Recorder (Tascam DR-40).
Just recording using the built-in mics until I can get this issue resolved.

Currently, the .wav file from the Tascam looses sync with the GH3 .mov file after about
20 minutes into the recording. Where it is synced at the beginning, the clips gradually drift
apart such that by the end of an hour, they are nearly a second off.

Curious if anyone else has experienced an issue like this? Could it be due to the settings
I have selected for the Tascam? I am recording in a .wav file as follows:

.WAV (mp3 is another option)
24bit (16 bit is an option)
48 khz (44.1 and 96 are other options)

Are these good settings to match the GH3, or can they be improved?

 

[stoneinapond]

Sounds (pun unitended) like the drift between the internal clocks of the recorder and camera. One second after an hour is not so terrible and wouldn't be an issue unless you absolutely had to record continuously for that length of time. If you do there are several ways to fix it, but a simple solution is to cut up the audio into shorter sections and sync (move) them as required.

And your Tascam settings appear to be OK.

 

[M. Gilden]

Strange, I used a Tascam DR-07, a GH1 and a GH2 to record 1 hour long lectures a 6 months ago. Everything remained in sync for me.

Is this a GH3 problem?

 

[TheDingo]

the .wav file from the Tascam looses sync with the GH3 .mov file after about 20 minutes

Because the internal timing clocks built in to each consumer device is different, it's very common for the sync to drift after a short period of time.

The two simplest ways of correcting this issue are:

1- Cut up your audio into smaller segments ( say a few minutes each ), and manually re-sync each segment. ( or use software like PluralEyes to sync them for you )

2- Adjust the duration of your audio to match the duration of your video. ( either shrinking or stretching the duration of your audio )

One nice feature of the Sony Vegas NLE is that you can click on the beginning or end of your audio track while pressing the CTRL key, and this will allow you to stretch or shrink your audio track to match the timing of your video track. So if you cut your audio to match the start and end of your video, you just sync one end of the audio track and then adjust the duration of the audio track to exactly match the video track. It takes 10 seconds, and now your audio is sync'd for the entire track.

 

[stoneinapond]

...
2- Adjust the duration of your audio to match the duration of your video. ( either shrinking or stretching the duration of your audio )

The only problem with this scenario is that sometimes the in-between syllables get moved to incorrect positions since the software is not aware of the content. It is just mathematically shifting things.

Cutting up the audio file, or using Plural Eyes or something similar is preferable in my opinion.

 

[TheDingo]

The only problem with this scenario is that sometimes the in-between syllables get moved to incorrect positions since the software is not aware of the content.

I find with todays digital gear the sync drift is linear, so when you adjust the duration of the audio it will stay frame-accurate for many hours. ( the longest tracks I've done this with were about 8 hours of conference lectures )

Though I do find that Sony Vegas will sometimes lose track of the duration adjustment if you keep cutting from the adjusted audio track, so I always export the re-sync'd video using an editing CODEC ( i.e. Cineform or DNxHD ) and then edit using the exported footage.

 

[M. Gilden]

The only problem with this scenario is that sometimes the in-between syllables get moved to incorrect positions since the software is not aware of the content. It is just mathematically shifting things.

Cutting up the audio file, or using Plural Eyes or something similar is preferable in my opinion.

Yeah, actually I'm with TheDingo on this one- resizing the content should do the trick because the problem is mathematical by nature, and should not need to be content aware in the slightest.
I did once deal with a recording system that recorded at a faster pace than real time (long lectures, shortened by 5% for quickier review). I, of course, did not know this and wanted to use the hall's audio to sync to instead of using my own.
When I was presented with an audio layer that was at least 1-2 minutes shorter than the video, I thought I was screwed. But some quick stretching of the audio layer to re-time it (also using Sony Vegas- their built in audio support is amazing), and voila! Everything magically fell into place. It really should be that easy unless there is something else awry.