June 8th, 2012, 09:53 AM
I edit HDV and over the years I have used camera flash, quick movements, and sounds to provide markers to match clips. They work but it is a slow process. I have tried Singular Softwares Plural Eyes which created more problems than it solved. Any other suggestions to line up clips would be greatly appreciated.
June 8th, 2012, 02:56 PM
I'm am not sure I know exactly what the difficulties are, but maybe the following will help sort out some solutions for you. At least, it may help better focus the discussion.
First, the only way I know to get video easily synched without any work with flashes etc is to lock and synch cameras' time codes. There are not many HDV cams that offer this capability and it is not a practical solution for many of us. Also, as you mentioned Plural Eyes, I am guessing that you might also be looking at ways of better synching multi-cam audio tracks among themselves. Unfortunately, time-code locking will not fix synching between audio tracks when the tracks were recorded at different distances from the audio source.
Second, because of the potential for differing distances from audio sources, using linked audio tracks may not provide a good way to match video or match up with each. For that to work, each camera (or its audio feed) either needs to have been the same distance from the audio source or else each camera has to have been fed a common audio signal. The reason it can be frustrating to use on-camera audio for synching video (either manually or via Plural Eyes) is that, depending on the distance, there may be a 2 or 3 or more frames of difference between the various audio tracks even when their linked video tracks are perfectly synched from flashes. Sound travels slowly enough that differences in distance between from the sound source can result in noticeable differences in audio timing between cameras.
Trying to synch video by using both audio and camera flashes might well tie you up in knots.
On the other hand, if your video is from two cameras shooting interviews and the cameras (or mics) are within ten or twenty feet of the subjects, then using audio synch may give you excellent video synch.
Third, editing workflow can ease the burden of synching. For event shoots, I've always found it much easier to first sync event video tracks with a camera flash, then find the audio source closest to the people --- which would be mics on wedding participants for ceremonies and the camera closest to the onstage performers for dance and theater --- and leave that linked to its video track.
On the PPro timeline, I then unlink the other audio tracks and slide them around to match them up to my reference track. I have done this manually by matching wave forms or by using Plural Eyes when it was harder to spot clear start and stop points in the waveforms. (I have been doing this since PPro CS2, and for years before that with the old Cineform NLE.) This is fairly easy when there are clear breaks in the audio -- say, the start of music for a dance or when a subject starts speaking after a quiet moment. It may be harder for events like weddings.
Fourth, you mentioned HDV. Are you recording to tape or to tapeless media? If you are using tape, there is always the risk of dropouts. Finding them is time consuming. I know of no non-time-consuming way to synch tracks from HDV tape where you might have some drop outs. This problem is what pushed me to tapeless recording and it has made my editing life considerably easier. It may or may not be a problem for you.
Fifth, while I use camera flashes, there can be a couple of quirks when working with interlaced footage from cameras with rolling shutters, and there are also some quirks with some flashes. These quirks can add a bit to the time it takes to check the video synch, but I have never found checking particularly time consuming.or onerous. YMMV.
The quirk with some camera flashes, particularly those from some professional flash units and some digital still cameras, is that the flash goes off twice, almost a stutter step a frame or two apart. You have to be sure that you are using the same flash, which usually is not time consuming unless you have one of those moments where all the cameras might not pick up all of the flashes. The related quirk from interlaced footage and rolling shutters is that the flash may go in one field and not the next on one camera and may show up in two fields on the next camera. So which field do you use? My solution to this dilemma is to wait about 20 seconds or so and shoot a second flash. I also look for a flash at the end of the program (or maybe shoot off another of my own), so I have a another reference that may guard against cameras' internal clocks drifting relative to each other. I generally do that with the longer multi-cam shoots.
Finally, "time consuming" is a relative and rather subjective term. For me, something that takes a few minutes may be acceptable. For you, that same amount of time might seem inordinately burdensome. Or vice versa. It depends on what your frame of reference is.
So, here is my frame of reference. I shot a 3½ hour dance recital last weekend. I had six cameras, a mix of AVCHD and tapeles HDV It took me about three or four minutes to synch the three AVHCD tracks and two of the HDV tracks. Most of the time was in locating the initial flash on each track The final track (a locked down HDV cam) had not been started before the beginning flashes (ooopps). I had to look for other flashes and work backwards to get that track synched with the other five. (Having the flash at the end helped a lot.) I am using only some of the audio tracks. It took a few minutes to identify match points in the waveforms (sharp starts to the music) and slide the tracks into alignment with the audio from my stage cams which were the reference point. (It is a dance recital, so I want the audio to synch up with what the dancers were hearing.) Most of the synch time was in just listening to the playback to check if (or where) adjustments might be needed.
This seemed reasonable to me but YMMV.
June 9th, 2012, 03:47 AM
Devil's advocate here:
If it's difficult to tell where a clip belongs, then it doesn't matter where it goes! The audience won't notice if it's not in chronological order. Mwahahahaha.
June 11th, 2012, 06:54 AM
I just do two camera shoots with an audio digital recorder for a third audio track. When I was shooting tape with an XL2, I synced by both camera strobe and audio. I find it easier finding the sync points by looking at the audio waveform than scrubbing and finding the frame with the flash. I just have to wait for the PEK fiels to be built.
The camera flash was just for backup and because of the short duration of the flash, I usually pop two because sometimes it's right in between frames.
Now I shoot with two Canon XF100's, and I'm able to sync with the timecode of the video files. I still use the clapper (actually, it's a spring loaded eye glass case that makes a nice loud clap when snapped closed) to sync the digital audio recorder with one of the cameras. I don't use the camera strobe anymore.
It takes me less than a minute to get everything synced in PPr.