Syncing video events with audio waveform spikes

I just cut a project based on a musical sound track. I turned on audio waveforms and performed hard cuts right on the audio waveform spikes. For my audio track the spikes represented the most dominant drum sounds in the sound track.

Is this the best or the most accurate method of syncing audio and video?

When playing back this project on my system, the video events seen on the monitor seem to be just a hair(hardly noticable) late than the audio events heard through the sound system. Should I depend on the edit placement as being perfectly accurate just because the visual representations in the timeline are matched up perfectly?

Is it best to cut on the 1 and 3 beats(down beats) or the 2 and 4(snare drum upbeats) in typical 4/4 rock music?

What do most of the pros do?

Thanks in advance!

Frank

eMac 1.25ghz/1gb RAM/160gb Internal HD/Superdrive Mac OS X (10.3.9) External Firewire HD's Lacie 160GB & 250GB

You are definately on the right track, that is what I do and I never mess it up. It is almost the same as having a clapper board. The snare will strike on 2 and 4 of the bar and thus the audio waveform will give a spike that is as tight as a nats whisker. phil...

> When playing back this project on my system, the video events seen on the
> monitor seem to be just a hair(hardly noticable) late than the audio events
> heard through the sound system.

You can't monitor sound at the computer while monitoring image on an external playback device. Those two will never sync. Either watch the image on the computer and turn off External Video (APPLE-F12), or connect your speakers to the deck/capture card where the external image signal is going and monitor sound externally.

> Is it best to cut on the 1 and 3 beats(down beats) or the 2 and 4(snare
> drum upbeats) in typical 4/4 rock music?

That's not a very good way to approach cutting music-video style. Every piece of music is different; tempos, the high-hat part, basslines, guitar parts. Within every 4/4 beat there are also 1/8th notes, 1/16 notes, etc. Just try cutting images to Jimi Hendrix's "Spanish Castle Magic" -- there are cut points not only at the 1, 2, 3 and 4, but also in between. Even a sturdy AC/DC beat like "Back in Black" would have notes, chords and accents falling ahead of and behind the beats.

Always cutting dead on the 1, 2, 3 and 4 -- let alone only on the 1 and 3 -- makes for an extremely dull music video where the audience can predict where your next cut will fall. Think in terms of dance, which is fluid and take advantage of the implied polyrhythms within any beat.

>>Think in terms of dance, which is fluid and take advantage of the implied polyrhythms within any beat.<<

Watch this video and notice how he does things in time to the music, but not always on the beat of the music.





I reckon this is a great analogy for cutting to music. Don't be rigid, find the other rhythms and syncopations and movements in the music. And enjoy it! Then you virtually can't go wrong.

A music teacher of mine once told me, "Music is math, but don't ever try to approach music like a mathamatician." Even though there is structure to music, you should try to remain fluid when cutting to it.

I will always make cuts initially by just listening and throwing down markers where I "feel" the cut should happen. Then, later, I will tighten it up to make up for any cuts that don't hold up to multiple passes. But I don't afix everything to the waveform spikes even when cutting to the beat. That would take out the "humanizing" quality of the edit.

Andy

the whole idea of pre-dertermining the cut points is quite alien to me.

in my deep dark history, i DO remember that that is how i approached my first exercises in cutting to music, but that was on single system super-8, and then on a 16mm exercise at my first film school.

very occasionally i will use that technique, but only where i know i want a few short shots cut exactly to the beat.
perhaps surprisingly that's not a regular occurrence.

no, it's the CONTENT of the shot that determines the cut,
then as others have been suggesting, how the content relates to the music,
and where i happen to feel a cut would be good.
it HAS to change, or become boring as Derek explains.

as my favourite producer explains, an audience will lose interest in a film, not necessarily because it's too slow, but because they've gotten ahead of it.
if they get ahead of the film, and know what;s going to happen next, then you are lost.

i edited most of my music videos on tape to tape systems, so this idea wasnt even available.
rather i went with the flow of the music as described so well be others here.
and again content, and performance was king.

more often than not, you'll want to have the action of the shot fall on a beat,
not the cut to the shot.
think of a drummer hitting a cymbal.
if you cut to that on the beat, you've missed it!
same with a dancer doing a turn, a guitarist thrashing away, or a performer singing their lyrics.

record company people used to say "hey, that guy really knows how to cut to the beat"
but i'd be thinking: "i know how to cut WITH the beat!".

---------------------------

by the way, if you ARE putting to your cuts exactly on a beat, it may well seem a bit off in a large space.
mainly i think this is just an issue of human perception, or the way we LIKE to see things,
but partly it's because sound and light travel at different speeds.

when the Disney animators were doing tests for Snow White (first animated feature), they discovered that if they had the dwarves feet fall EXACTLY on a beat, it looked wrong.
they figured that with the size of the cinema, the sound reached them a fraction of a second later than the vision did.
they delayed the dwarves foot-fal by a frame, and their dancing seemed much better.
(again see how it;s what happens IN the shot that is important)

i'm not sure the delay thing was all that was going on there:
it originally looked as if the dwarves were leading the music,
whereas if you watch any dancer, you'll find the music leading them.

cheers,
nick

I would definately suggest you use after effects for a more exact and automatic interpretation of the track.
Trapcodes sound keys does this perfectly and it lets you monitor up to three different beats (Hz)in one track at the same time.


All you have to do is write an expression that links the sounds you want to edit to in the track to the opacity colors that appear on the beat.

Then you render it out, bring it into FCP and use this to decide your edit points.

If you dont understand how to do this I can write the script for you and send it via email in an After Effects file.

Its quicker and easier.



Johan Polhem
Motion Graphics
www.johanpolhem.com