Manual Synchronization Trick

[Jon Fairhurst]

Today, I called into a meeting on my mobile phone that was also available online. I plugged in my ear buds, and after completing a short errand, I logged onto the computer and enabled the headset. To avoid a gap, I put the headset over the ear buds.

The time offset was nearly 1/2 second, yet it was hard to determine which source led and which lagged. After a minute of deciding, "phone, no, online, no, phone...", I pulled the left ear bud and lifted the right headphone and the early source was obvious. I could easily hear the sound arrive left (online) and then right (phone).

Clearly, this trick can be used when matching dual system sound with in-camera sound. Put one on the left and the other on the right and adjust the time offset until there is no left-right echo effect.

I've always done this with the sources mixed. I rely on the visual aids of steep plosives on the timeline. It works, but it can be a challenge as we go from clips to minutes to seconds to milliseconds.

I'll try the split L/R approach during my next sessions. I expect that this will be much less confusing to the ear I dial in my offsets.

[Jay Massengill]

+1 for temporarily panning sources to left and right to better determine timing.

This is especially true with very low sync tracks recorded with on-board mics that don't contain much visual information on the waveform.

That's one of the pitfalls of the "it's only for sync" shortcuts. You have to be able to hear it, and better yet see it to easily drive sources into sync. Otherwise it's more work and resorting to tricks like panning to sort it all out.

[Jon Fairhurst]

Frankly, I was shocked at how confusing it was to distinguish which source was early when the two were mixed. The tone of the two sources was very different, and I could clearly hear how much delay there was, but my brain couldn't sort out "who's on first" at all.

I was equally surprised at how obvious it was when panned. No golden ears required.

I guess this makes sense in the real world. When we hear two sources, they almost always come from different locations. We didn't evolve hearing two things from exactly the same point in space.

[Paul R Johnson]

Always good to remember when people extol the virtues of mixing on phones. Without the 'real world' headphones have some problems, don't they.

[Don Palomaki]

Interesting.
How well did it hold? Often not unusual for long haul comm channels to have varying latency, especially internet channels.

[Roberto Diaz]

I have used this technique for years to sync audio to video. put the first, i do a coarse alignment using the waveforms. then to get it dead on, i put the videocamera's audio track full left and the "real" audio track full right. then slide one against the other. with headphones of course!

it's amazing to hear the sound when they get lined up. one way i've described it to others is that when they're out of sync you hear a sort of echo, but when they're in sync, the sound seems to come from the center of your skull -- literally right between your ears, inside your head.

more recently, i've been using plural eye (with sony vegas). much quicker, of course, but the poor man's version is definitely effective.

[Jon Fairhurst]

Paul, you're so right about mixing on headphones being problematic. Then again, there are times when that's the only practical solution (late at night, in an office cube...) It's fine to do an initial mix on headphones to select the right tracks, set initial levels, and apply initial EQ. This gets the grunt work out of the way. But for the final, tweak everything on monitors. That said, I like to double check the mix on cans (and other systems) to make sure I didn't overcook anything.

Don, the connection was good and I didn't listen long enough for GoToMeeting to slip the delays, so it held steady. That said, I've heard some woooords aaaartificially streeeeetched during meetings as latencies are adjusted. The worst problem I have is that the AGC is overly aggressive. I have to mute when I stop talking and start typing - even with a headset mic an inch from my mouth.

Roberto, yes Plural Eyes is easier but there's something nice about just dialing it in. :)

[Seth Bloombaum]

Originally Posted by Roberto Diaz

...more recently, i've been using plural eye (with sony vegas). much quicker, of course, but the poor man's version is definitely effective.

Speaking of Sony Vegas, if you have it you should be aware of a fine-adjust method for sync that only Vegas supports AFAIK.

Rough in the sync
Turn off Quantize to Frames (older versions, new versions don't conform audio track slips to whole frames)
Select the audio event on the timeline that you want to slip for sync
Zoom in as far as you want; you can go right down to *extremely* zoomed in (amount of zoom affects amount of slip for each key press)
Loop-play over some sync-sensitive area of content, like lip-sync or drums
Listen to original bad reference audio (why not pan it left, per Jon's method!), and 2nd system good audio, (pan it right!)
Now, use the Num-pad keys 4 and 6 to slip the new audio left an right to reduce and eliminate echo. When there's none you're as synched as you can get it.
Don't forget to turn on Quantize to Frames if you turned it off!

This, and Jon's method, both work because we're very sensitive to slight delays/echos. We use them subconsciously all the time to help localize audio sources in the real world. Even better as Jon suggests, with the two sources hard-panned to L & R!

One of these days I'll run into a project where I'll have to purchase Plural Eyes, but for long-take projects the manual methods work great, I've synced hours and hours of footage this way.

[Rick Reineke]

As a Pro Vegas user, (and beta tester for SCS = SONY CREATIVE SOFTWARE), you can edit the audio to the sample level (typically 48Kkb/s) If so selected in the Options /Preferences menu. However the A/V files must un-groupted first.. (select, U)


though... do not blame me for they're p_s poor tech support (or lack thereof) and other short comings