ADR Syncing

I have acquired video files for a TV film that I’m restoring to its original cut.
Thread here: http://originaltrilogy.com/topic/AD-Anno-Domini-1985/id/54709

I haven’t done a restoration project of this magnitude before, and I’m having a lot of difficulty with sensitivity to the A/V sync. The audio is definitely leading the video (not by much) but it feels inconsistent and it’s driving me insane.

I’m really hoping that I don’t have to go back and correct the audio scene-by-scene for an almost 10 hour miniseries. I’ve tried lagging the audio by a frame or two, which sometimes helps, and sometimes still feels off. And I’ve tried lagging the audio by as many as 10 frames, where it is clearly behind the video. Nothing seems to work.

Does anyone have experiencing syncing A/V with only one audio source?

Can you describe the source of the video and the source of the audio that you are sync’ing to that video?

Thanks for your reply Alex. I have .VOB files of the movie. Just the same single audio and video source. The sync looks good in the .VOB files but the video is distorted and needs some cropping and other work, plus I want to do a fan edit as well. When I import into my editing software, Premiere Pro, the sync is inconsistent…it varies without rhyme or reason from watchable to distracting. When I slow down and inspect frame by frame, usually the audio leads the video by a few frames. When I export, the opposite is actually happening. The audio lags the video even more noticeably.

I was hoping someone with experience in Premiere Pro would respond to your query by now, as I am not familiar with that software. However, with some sync’ing projects that I’ve done, I’ll attempt to answer.

From my reading, both the video and audio are from one source, the .VOB files, right? If that is the case: from your description of your sync’ing issue, when you imported the .VOB files to Premiere Pro, the frame rate of the video—the amount of frames displayed per second (fps)—may have changed, but the audio that was sync’d to the .VOB files remained the same. You may have to convert your restored video back to the fps it was in the .VOB files so that your finished video can be sync’d to the original audio.

So, first, compare the fps of the .VOB files, your imported video, and/or your finished video and see if there is a difference.

Thanks again Alex. I was told on the Adobe forums that I might need to convert to AVI DV NTSC before importing into Premiere Pro. I haven’t really nailed down that convo yet though. Here are the specs of the source file and the exported file:

.VOB source files:
VIDEO_TS\VTS_02_1.VOB
VIDEO_TS\VTS_02_2.VOB
VIDEO_TS\VTS_02_3.VOB
VIDEO_TS\VTS_02_4.VOB
VIDEO_TS\VTS_02_5.VOB
Format: MPEG-PS (MPEG-2 Program Stream)
Duration: 2h 19m 42s 144ms
Size: 4.12 GB
Bitrate: 4219 kbps

Data:
Codec: DVD_NAV

Video:
Codec: MPEG-2 video
Duration: 2h 19m 41s 906ms (251,205 frames)
Bitrate: 4000 kbps
First frame: 0.00s
Frame interval: 33366 (0.033s)
Framerate: 29.97 frames/sec
Resolution: 720x480
Aspect: 4:3
Pixel format: YUV420P
Interlace: TFF
Color range: MPEG
Color space: BT409BG
Color primaries: BT470M
Color transfer: BT470M
Profile: Main @ L8.0
B-frames mode: 1
Maximum B-frames: 2 frames
B-delay: 1 frame
ReFrames: 1 frame
Packets per frame: 1 packet

Audio:
Codec: ATSC A/52A (AC-3)
Duration: 2h 19m 42s 144ms
Bitrate: 256 kbps
First frame: 0.00s
Frame interval: 32000 (0.032s)
Framesize: 1536 samples
Channels: Stereo
Layout: L R
Samplerate: 48000hz
Bit depth: 32 bit

FPS = 29.97

Looks like the duration length of the audio and video are off, with different frame intervals.

When I exported based on parameters another user suggested:

Container:
Format: QuickTime / MOV
Duration: 2h 17m 16s 138ms
Size: 9.89 gb
Bitrate: 10315 kbps

Video:
Codec: H.264 / AVC / MPEG-4 AVC / MPEG-4 part 10
FOURCC: avc1
Duration: 2h 17m 16s 94ms (246,835 frames)
Bitrate: 9988 kbps
First frame: 0.03s
Frame interval: 33366 (0.033s)
Framerate: 29.97 frames/sec
Resolution: 720x480
Aspect: 1.363
Pixel format: YUV420P
Interlace: Progressive
Color range: MPEG
Color space: SMPTE170M
Color primaries: SMPTE170M
Color transfer: SMPTE170M
Bit depth: 24 bit
Profile: High @ L3.1
B-frames mode: 1
Maximum B-frames: 2 frames
B-delay: 1 frame
ReFrames: 3 frames
Packets per frame: 1 packet
Language: English (ENG)
Disposition: Default
Title: Alias Data Handler

Audio:
Codec: AAC (Advanced Audio Coding)
FOURCC: mp4a
Duration: 2h 17m 16s 138ms
Bitrate: 317 kbps
First frame: 0.00s
Delay: -33ms
Frame interval: 21333 (0.021s)
Framesize: 1024 samples
Channels: Stereo
Layout: L R
Samplerate: 48000hz
Bit depth: 16 bit
Profile: LC
Language: English (ENG)
Disposition: Default
Title: Alias Data Handler

Framerate is also 29.97 fps and the video duration length is again off from the audio/container length. Different here though are things like audio and video lag, Reframes, and frame intervals.

In that case, one suggestion would be to fix the audio to fit the length of your final rendered video. But, there is a way you can do that without any trimming. You just have to adjust the duration of the audio to fit your finished video.

A free software download, like eac3to, can do that for you. Use it to demux the audio from the video, giving a .WAV file. Then, just adjust the duration–making it slow or fast.

Finally, mux your adjusted audio to your finished video. Check the beginning of the video and check the end of your video. If both ends are in sync, you’re done. If not, go back to eac3to again and adjust the audio to another duration time. You may have to play with the duration to get the right fit.

I, personally, have not used it, but other members here have.