Vegas 11 Quantize to Frames or not

[Graham Bernard]

Q to F always, ALWAYS on. If people haven't experienced an issue, that's luck. When I see the PINK out of whack syncs, I know just where and how I should slip the V or A, and it snaps back.

It's a powerful tool to have available, but only to be used under extreme conditions of Audio matching.

Grazie

[Ron Evans]

I agree with Ray that I can definitely hear the difference if tracks are slightly out of sync. I use Edius for video editing and Vegas for audio exactly because I can edit with quantize to frames off. I export audio from Edius after sync in multicam, edit audio tracks to mix in Vegas and then import back into Edius for final export.

Ron Evans

[Edward Troxel]

Well... there's a difference between "slightly out of sync" and "less than 1/60 of a second out of sync". But, yes, audio can definitely by adjusted at subframe levels. I would NOT adjust the video at sub-frame levels, though.

[Ron Evans]

Less so now most cameras are HD AVCHD etc but with DV the spec allowed the audio to be 1/3 a frame off. Mixing several DV camera audio one can see that when the pictures are in sync the audio can be off by this 1/3 frame, different for each camera . For three cameras the audio could be more than a frame off between them. Just adding these audio tracks with this less than a frame out of sync audio problem is very audible for lip sync dialog and definitely music. Certainly not acceptable to my ears.

I also see this shift if I record one track from my shotgun mic and the other from the facility mixing board. So use the audio shift in lots of situations. My main reason for Vegas since it was an audio editing program.

There is still a noticeable difference though between my NX5U audio and the XR500 or CX700 AVCHD so I still use the work flow. Edius cannot shift audio in these fine increments as easily as Vegas hence my work flow. Video of course is always frame sync. Comments are only for audio.

Ron Evans

[David Wayne Groves]

Have you considered using plural eyes, it should sinc all your media with a click of the mouse....

[Ron Evans]

Plural Eyes sync audio so that if the audio is off by a frame then Plural Eyes will sync to the audio and the video will be off by that frame. You will still have to fine tune the video and then the audio to get fine sync of both. If you have lots of clips it is definitely a time saver. If however you have started the cameras in a multicam show and left running to the end then I find it just as quick to line up manually. If they are consumer cameras without genloc then the audio will still have to be corrected. At least that is the case with my NX5U, XR500, SR11 and CX700 over an hour clip. When I shot with FX1, HC96 and TRV50 in DV there were very big differences.

Ron Evans

[David Wayne Groves]

Strange, I have never had a problem syncing my NX5U, AX2000, Canon HG21 and HG10 on shows that last well over 2 Hours, generally the cams are never shut off till the end of the show...plural eyes just syncs the video as well as the audio perfectly when I pull them into vegas for a multicam edit, I have yet to make any fine tuning.......

[Ron Evans]

Do a little test and you may see what I am talking about. On a resent show 55min Act 1 45 mins Act 2. Cameras were my NX5U, Sony SR11 and Sony XR500 all shooting 1920x1080 60i. All cameras always left running for the show. In video sync at the beginning, if I expand the timeline to a frame then the NX5U and the SR11 audio are in sync the XR500 audio is delayed by 1/2 a frame. With frame sync ON it is not possible to get this any closer, either a 1/2 frame ahead or behind. So in this case it is important to choose which is correct by looking at the video frame. For dance this is important. Hence my comment about fine tuning.

After 55 mins the SR11 audio has drifted to a delay of just under 1/2 frame and the XR500 just a little more than at the beginning. These in reference to the NX5U audio which of course could also be drifting !!!! . Similar time differences for Act 2. I don't view this is a problem for consumer cameras that are not genlocked. This is exactly why pro cameras are genlocked to hold video and audio sync to a reference clock. Vegas and other NLE's will frame sync the multi track video but the audio may thus be out of sync. In my mind this is the real value of Vegas over Edius. The audio on the non reference tracks can be stretched or squeezed ( with audio clock increments) to bring them into sync, effectively correcting for the clock drift in the camera. With AVCHD the drift is small when I used to shoot Hi8 and even DV the drift was significant between cameras. Spec for DV allows audio to be within 1/3 a frame of video so for 3 DV cameras at limits of spec the audio could be over a frame out and drift as well. So I have done this fine tuning for a long time. I don't always mix all the tracks anyway, usually taking a mic at stage level and/or a shot gun as main with a stereo mix to give some body to the overall mix.

Ron Evans

[Chris Hewitt]

I find the trouble with PluralEyes is that it won't allow me to export the audio to SoundForge, it comes up with an 'error' message. I've just tried the demo and it does what it says, great sync. I would have bought it if it allowed me to export. I suppose I could export the audio before running PluralEyes.....

[Ray Turcotte]

Great discussion from everyone. Thanks

For the record I mostly shoot and edit live multicam music events. I've edited a few more projects since and I can turn off QTF and sync muticams exactly provided GPU rendering is Turned Off. The results are great. If GPU rendering is turned On Vegas will crash (VP V511; nvidia V295.73)

As to pluraeyes, I'd understood that it is for syncing video tracks (with audio) together on the time line. After that you can edit the audio tracks anyway you wish.

Ray

[Stephen Crye]

Originally Posted by Edward Troxel

... Who's going to hear a half frame audio difference? Now, a frame or two - you can definitely here - but a half frame (at worse case and should have 1/4 frame or less)? If it averaged 1/4 frame, that's 1/120th of a second!

.

Hi; I know this thread is old, but I thought I would answer your rhetorical question: Classical Musicians. These people have an extraordinary internal clock - they have to otherwise they could not perform.

For example, if the sound is off by +/- 1/2 frame , the strike of the mallet on a kettle drum will be noticeably out of sync with the audio. Kettle drum mallets, particularly during a drum roll, move very fast. The mallet, bouncing off the drum head while the performer keeps downward tension on it for the next strike, can easily hit the drum 5 times in a second. The distance traveled from the apex can be as much as 2 feet. That works out to 20 feet per second. A half-frame time at 30 fps is 1/60th of a second. 1/60th of 20 feet is 4" - in other words, the sound of the drum strike could occur with the mallet 4" above the drum head, easily noticeable.

I do multi-cam recordings of orchestra dress rehearsals, and have had these small sync problems brought to my attention. After I fixed them with quantize off, the musicians were pleased.

Steve

[Juris Lielpeteris]

Here is the key sentence:

Originally Posted by Edward Troxel

audio can definitely by adjusted at subframe levels. I would NOT adjust the video at sub-frame levels, though.

[Ray Turcotte]

A follow up


I've completely changed my approach to question I that posed in this thread. Edward Troxel (and others) are correct ; Always edit and render video on frame boundaries. Vegas is much more stable and error free that way.

The editing process I've adopted since, is:

Import A/V sources and place them on the timeline

Verify quantizie to frames is on.

Choose a particular video frame and match up all the video tracks to the chosen frame. I often use the drummer as a reference as a drum hit or cymbal smash will be pronounced in all tracks, both visually and sonically. Ignore the out of sync audio for the time being. The most the video tracks can be visually out of step is 1/2 of a frame, which is not a problem. Once all the video tracks are matched up, sync lock the video tracks together as a group, or temporarily lock them so the tracks don't accidentally move.

Choose one video sound track as the master sound track. Turn off quantize to frames and then slip the secondary audio tracks until they match the master track. Ignore and mute any audio tracks that are of no use. Once all the good audio tracks are in sync, lock them down, by whatever method you prefer. Then turn quantize to frames back on. (very important)

Continue on editing as normal.

I had asked about quantize to frames as I couldn't fathom what the switch was for. Once I realized the switch is for unlocking audio tracks from the video tracks it became clear to me how to use it. Audio requires a much finer adjustment as compared to video, and so setting quantize to frames off, allows this to happen. In multi cam music shoots, which is the type of shooting I mostly do, it is an important step in the editing process. Now if my cameras all had time code sync, then then everything would match right from the cameras, making most of the above unnecessary.


Happy editing...

[Graham Bernard]

Ray thanks for coming back and updating this thread with your invaluable experience of the power of QtF.

Grazie