Sync sound and pictures

[Jon Bang]

Hello :-)
I have a frame-by-frame scanner with sound. I have both 8mm and 16mm by the way. But how to sync sound and pictures when i am making filmfile? I am using audiacty today, but i experience that the sound is not 100% synced. When the film is bigger this is a problem. Small films, 50" films are no problem, but when more filmrolls are put in a big roll the problem appears. The scanner gives sync-signal for every picture.

[Mattias Norberg]

Hello :-)
I have a frame-by-frame scanner with sound. I have both 8mm and 16mm by the way. But how to sync sound and pictures when i am making filmfile? I am using audiacty today, but i experience that the sound is not 100% synced. When the film is bigger this is a problem. Small films, 50" films are no problem, but when more filmrolls are put in a big roll the problem appears. The scanner gives sync-signal for every picture.

Hi when i did have a very difficult sync problem in super 8mm film i did use a Avisynth Plugin called "AudioGraph" and i did work very good for me here is a link to the Plugin http://avisynth.nl/index.php/AudioGraph


here is a example

video_clip=DirectShowSource("H:\FILM\Super8mm.avi")
audio_clip=WAVSource("h:\FILM\Film_Audio.WAV")

d=AudioDub(video_clip,audio_clip)

m1=DelayAudio(d,-5.158).trim(0,1749)
m1a=DelayAudio(d,-5.300).trim(1750,2742)
m1b=DelayAudio(d,-5.400).trim(2743,3181)
m2=DelayAudio(d,-5.497).trim(3182,4400)
m3=DelayAudio(d,-5.619).trim(4401,5562)
m4=DelayAudio(d,-5.838).trim(5563,8007)
m5=DelayAudio(d,-5.970).trim(8008,9089)
m6=DelayAudio(d,-6.100).trim(9090,9817)
m7=DelayAudio(d,-6.150).trim(9818,11903)
m8=DelayAudio(d,-6.250).trim(11904,14369)
m9=DelayAudio(d,-6.400).trim(14370,14536)
m10=DelayAudio(d,-6.250).trim(14537,17969)
m11=DelayAudio(d,-6.200).trim(17970,18345)
m12=DelayAudio(d,-6.100).trim(18346,19592)
m13=DelayAudio(d,-6.000).trim(19593,20881)
m14=DelayAudio(d,-5.850).trim(20882,22098)
m15=DelayAudio(d,-5.750).trim(22099,23241)
m16=DelayAudio(d,-5.650).trim(23242,24087)
m17=DelayAudio(d,-5.550).trim(24088,25396)
m18=DelayAudio(d,-5.400).trim(25397,27598)
m19=DelayAudio(d,-5.250).trim(27599,29137)

m=m1++m1a++m1b++m2++m3++m4++m5++m6++m7++m8++m9++m10++m11++m12++m13++m14++m15++m16++m17++m18++m19

return m.AudioGraph(5)

you can see m1 = frame 0 to 1749 have DelayAudio -5.158
and m1a = frame 1750 to 2742 have DelayAudio -5.300 and so on then i join them all like this m1++m1a++m1b++m2++m3++m4++m5++m6++m7++m8++m9++m10+ +m11++m12++m13++m14++m15++m16++m17++m18++m19

[Isidro Costantini]

It looks like AudioGraph will probably generate clicks. You may try using CollEditPro (or Audacity) stretch (preserving pitch, or neither) to match audio length to the one of the video. I think MKVToolNix also has some audio parameters to adjust delay and length without touching original WAV (which I usually save to MP3, or FLAC if audio quality is a must).

[Nantawat Kittiwarakul]

Most video editor software should be capable of audio editing, so that shouldn't be a problem.
The workflow should be...
1. Capture the audio first using the method of your choice. For me I'd use the camcorder & off the wall setup so I have the video as the visual guide when editing. The image quality does not matter at all as long as it's still recognizable.
2. Then capture the film frame-by-frame.
3. Import them into the editor. Having the image track stay fixed, and adjusting the audio track accordingly.

I've found that there's usually no need for dead-accurate sync. Unless it's music video (or the like) the audio could be anywhere from frame accurate to about 0.1 second delay and still goes unnoticed to most people.
But in any circumstance NEVER have the audio AHEAD of the picture. Your brain will immediately catch that up and knowing that something is very "off". This might relate to how we perceive visual and sound in real world - audio will mostly always delay compared to visual, but never ever vice versa.