I've been having a similar problem with clips with different start times. Is this caused by the same bug?

I merge the clips normally, (sync them, aline the in and out points and select merge clips) and then drop the entire clip into the timeline and double click to open it in the source monitor.  The sync of the clips is never lost, but something else happens which makes the Merge Clips function quite unusable.

(in the following section, the time between when one of the recorders, be it the camera or the audio recorder, was started and when the other one was will be referred to as "X")

Then one of three things happens:

1) if the files start at exactly the same time (Impossible unless hitting record on one piece of equipment triggers the other) everything works as it should.

2) if the audio recorder started first, the video plays back in the source monitor, but the timecode displayed is not in sync with the video. The timecode that gets displayed is the correct timecode plus X. If I set an In or out point, the audio in point is set at the correct point on the timeline, but the video in point on the timeline is out of sync, and the in point must be dragged back X seconds to match up.

3) if the Camera started first, the Timecode is displayed correctly, but when I set the in point, the audio is X seconds later on the timeline and must be dragged back to match the video inpoint which is correct.

If I open the clip in the source monitor from the project panel, everything works correctly.

JM

I'm afraid you don't quite understand the issue.

I did more testing.

first, the problem also happens on my laptop.

Second, I tested what diference was made by including or excluding the audio recorded by the camera.

The Video was recorded on a RED one and the audio on a Sound Devices 788T

If the audio and video from the RED is merged with the audio from the 788T, the timecode problem goes away, and the video in points edit correctly. The audio from the RED also edits correctly. However, the audio from the 788T ends up being off set by X.

if I don't include the audio from the RED in the murged clip, and only have the audio from the 788T, then the problems I discribed before arise.

if I open the same merged clip from the project panel, set an in point, and then drag it to the timeline, the audio and video in points are not ofset.

Basicaly, here's what I think is happening.

The programers made a mistake with the programing.   When you set an in or out point on a merged clip in the Source Panel which has been opened from the timeline, the inpoints in the verious tracks are all set by the time alapsed from the buginging of the file.    The problem is, if I have a 10 second video clip, and I set then in point at 2 seconds into the clip, the audio tracks are also set at 2 seconds into the clip.  But, if the audio started rolling 1 second early, then the audio ends up starting 1 second before the video.  

however, at no time is sync lost.   The audio Video sync never gets lost in any of this crasyness.

Does this make any more sense?

JM

Message was edited by: epsilonfilms.  fixed a small typo

Yep.  You were right. I checked what you said about markers, and got a few pictures. (And found another unrelated bug in the process)

This is with the camera audio merged with the audio from the T788.   The marker was set by finding the moment the slate shut on the video and seting the marker.  Note that the marker on the last audio track is way to the left, instead of on the audio spike where it should be.  The time between the two markers is X, which for this particular clip is 2 seconds and 11 frames, the time between when I started the audio and the video.

This time, I've removed the audio recorded by the camera.  I did the same thing here.  Open the clip in the source panel from the timeline, and find the slate clap visually.  This time, the marker on the audio track is in the correct place, while the one on the video track is very late.  2 seconds and 11 frames late. 

(if you look at the clip, the maker is at a point where I've removed the slate from the frame already.)

As for the timecode issue:

(the bottom image shows the correct timecode.)

The exact same frame of the same clip, loaded two different ways, showing different timecodes.  The difference between them, in this case, is 2 seconds 11 frames. Also note the difference in the timeline directly under the timecode. 

The bug I found while doing these tests relates to the importing of audio files.  For some reason, when I loaded all the audio from the T788 the timecode of the WAVs wasn't read. (it showed them all as starting at 00:00:00:00000)  Then randomly it appeared, and I was able to sync the files using it.

I still need to figure that out.

JM