Timing video and sound

FoskeyMedia schrieb am 17.04.2017 um 04:36 Uhr

I work a lot with music, dancers, and African drummers. Sometimes (multi-cam) I need to make some very pin point accurate cuts. I know I can expand down the frame level and go by the sound waveform. Does anyone have any other advice? I thought I read somewhere that because light travels faster than sound there's a particular formula for getting them to sync just right.

Kommentare

Ehemaliger User schrieb am 17.04.2017 um 04:44 Uhr

I once worked with a client for a half an hour discussing whether the cut should be on frame "x" or "y". I was previewing the same cut each time, but he saw something different each time. It is all in final perception.

 

 

FoskeyMedia schrieb am 17.04.2017 um 04:50 Uhr

I also find that (maybe because my PC is a bit older..at least 6 yrs old... but I7 with 32gb memory) that the video lags sometimes and is not always accurate. It may sync a bit different in the render.

Ehemaliger User schrieb am 17.04.2017 um 04:58 Uhr

It syncs exactly as you sync it, it just doesn't preview correctly. You always want to test render. I have a media player and a bluray that plays from a USB drive that I always watch to make final tweaks and adjustments. I don't worry so much about exact frames as I do how it looks in realtime trying to keep an eye as if I had not seen it before. Usually my wife gives me good input as well.

Grazie schrieb am 17.04.2017 um 05:42 Uhr

It's great fun! I use Marker generation as my primary Grid mapping for creating cut splits.

OK, 2 tips:

1/- Slow down the Preview playback. This will get you real close.

2/- Use the M Marker key to tap out the beat. Then use the Build RAM Preview often! I have a MACRO and have set this "Control" set to my Control Shuttle Pro.

The free Timeline Tools had added a Marker to Beat tool that I found useful. It has sensitivity and spread options.

You might want to experiment switching Quantize to Frames ON or maybe OFF.

 

Zuletzt geändert von Grazie am 17.04.2017, 05:46, insgesamt 2-mal geändert.

Grazie

PC 10 64-bit 64gb * Intel Core i9 10900X s2066 * EVGA RTX 3080 XC3 Ultra 10GB - Studio Driver 551.23 * 4x16G CorsVengLPX DDR4 2666C16 * Asus TUF X299 MK 2


Cameras: Canon XF300 + PowerShot SX60HS Bridge

Musicvid schrieb am 17.04.2017 um 19:42 Uhr

You can cut at video frame boundaries; or you can ungroup the audio, turn off quantize temporarily, and make more precise audio cuts without disturbing the video. .

You cannot divide a video frame, and doing so on the timeline will give unintended results.

Paul-Fierlinger schrieb am 18.04.2017 um 03:08 Uhr

For whatever this is worth there's an old rule of thumb in animation, established decades ago by the old Disney animators, which is to make lipsynching work the best, make the image track lag behind the sound track by 2 or 3 frames to best achieve the perception of perfect synchronization. Perhaps this is why your client perceived the real sync to not be accurate. Worth the try, though animation may be a lot different than live action.

NickHope schrieb am 18.04.2017 um 06:27 Uhr

You can cut at video frame boundaries; or you can ungroup the audio, turn off quantize temporarily, and make more precise audio cuts without disturbing the video. .

Don't forget there's no real need in the last few versions to turn off quantize-to-frames if this preference is checked.

Still need to ungroup though.