-
RE: 1 frame offset when synching audio to video
I am working on a project that requires synching audio from a sound devices recorder to HD footage. The audio recorder was jam synched with a time code slate and FCP seems to read the time code perfectly. I’m using the merge clip setting to sync the audio and video using in points. I’m using the reference audio track from the camera and the video to mark the time code in point where the slate claps but when I go to the audio track and try to match it with the time code from the slate. The click sound from the slate is consistently one frame beyond the time code from the slate. Does anyone know if this is an FCP issue or could it be something with the Kona LHI Card. I am working with video and audio recorded at 23.976.
Chad Tingle
Producer/Editor -
9 Replies
-
Momentary jam sync without a common video reference will result in potential drift. I am not surprised that a 1 frame offset could occur and I doubt it is a FCP issue.
-
Totally agreed.
-
To get a precise sync of digital audio to video, you need to sync both the timecode AND the word clock (I have also seen this referred to as “frame edge boundary”). Otherwise you’re subject to what is essentially a rounding issue that can cause your audio to be up to one frame out of sync.
-
Sorry, I mixed terms. Let me give it to you straight from my Tascam X-48 digital recorder manual:
Video clock is only updated 60 times per second, where word clock is updated 48,000 times or more per second. Long cable runs can also cause video clock to go out of phase, which may cause problems with digital consoles. Video clock on the X-48 is used to define the frame edge of the SMPTE input.
• Word Clock sets the playback speed. Once the SMPTE input tells the transport where to go, the X-48 plays at the rate set by its Sample Clock source (Settings menu, Source).
• Video Clock gives a precise frame edge to the SMPTE timecode. Not available as a clock reference.
• SMPTE Timecode is used as a location reference, defining where in the timeline to locate to.Now I see why Bob Zelin is such a fan of the AJA GEN10 sync generator.
-
If you want to get really technical, both the leading edge of the frame boundary and a leading edge of the word clock signal should be synchronous. Timecode should not drift in relation to the video or wordclock reference or sync errors will occur. Timecode generators are designed to align bit 0 to that video leading edge. Timecode can also be broken down to 100 bits per frame (NTSC) and 80 bits PAL. My old timecode generator/reader had a display that showed that sync alignment so I could see drift, sub frame, in unreferenced recordings.
20 years ago, I asked my father to design and build a box that created a wordclock square wave signal from a video reference generator so that all my digital audio devices were perfectly synced to all the video devices and to timecode. That box is still the core of my integrated audio and video post production facility. How people survive in a digital video and audio world without such reference amazes me, as does the endless stream of questions on this forum about sync, an issue that I have never been troubled by.
-
In this situation.. the sound devices recorder is generating time code to the slate. I misspoke when I used the term Jam sync.
Chad Tingle
Producer/Editor -
[Michael Gissing] “20 years ago, I asked my father to design and build a box that created a wordclock square wave signal from a video reference generator so that all my digital audio devices were perfectly synced to all the video devices and to timecode.”
You lucky boy.
-
[Jeremy Garchow] “You lucky boy.”
Good luck is 90% good management. That’s why I now own a 40 acre farm with a post facility at the bottom of the world without ever buying a lottery ticket or gambling on a horse or poker machine. All that in spite of working primarily in docos which are charity budgets.
-
Nice one. My dad would think a square wave is some sort of natural oceanic phenomenon.
Reply to this Discussion! Login or Sign Up