Time code: how critical?

[Joji Minatogawa]

Hello,

I am new to videography and have been reading past discussions on this forum about time codes.

I'm in the process of 'gearing up' to record solo violin performances. In order to make these videos more interesting, would like to use two cameras. I'm trying to visualize the editing process where there would be only one recording of audio but two sources of video (the two cameras). The result of the editing should be some alternating views of the perfomer but the sound track would be continuous, i.e. the music must coincide with the video.

I'm not sure if I'm asking the right questions, but, I would like to know:

1. How important is time code in this instance.
2. (Will use XL1s for audio because of audio capability); The XL1s manual does not state what time code format this camera uses. Is this a potential problem and are there any limitations as to what kind of tape deck I can use for archiving these recordings and then to tranfering to PC for editing.

If anyone can shed some light as to what I'm getting into, would be much appreciated.

Thank you.

[Rob Lohman]

Timecode is very important to you in such an event. In such
events timecode comes from a timecode generator and is fed
to multiple cameras (and perhaps a audio DAT recorder) to
synchronize it all. This will greatly save time in the editing
process to find footage but especially to be able to cut from
one camera to another.

Unfortunately for you (and other people) the XL1S (as every
other consumer DV camcorder) does not take external
timecode signals, so you cannot synchronize these. The
best thing todo is synchronize on a visual and audible happening
(like they do with a clapper board for example). Then you
can visually and audibly synchronize during editing.

There is also another thread going about using the infrared
record (or LANC controller?) to start multiple cameras at the
same time... check that out.

Hope this has explained some.

[Vic Owen]

As Rob mentions, you can't get absolute time code sync during the recording. You're on the right track to use only the audio from the XL-1, though. Once both video tracks are in the editing program, you can sync them fairly easily. Sometimes a flash will work, a specific audio event, even the position of some object in the frame. I just depends on the clip. With most editors, you can slip it either way to get pefect sync. Short of a master time code, this is your only option with this equipment.

[Joji Minatogawa]

Thanks guys,

Starting make some sense. If you'll bear with me, it looks like there are some alternative ways towards 'synching' audio and video; please correct me where I may go astray:

alt. 1: use a 'time code generator'
alt. 2: use infrared controller to start cameras at same time
alt. 3: record a visual/audio signal on both cameras and sync. during editing

Looks like alt.3 is the most cost efficient way to go, alt. 2 the second cost eff., alt. 3 the least.

But for information only,

1. are time code generators stand alone hardware and are they expensive?
2. do the cameras and DAT have to be hooked up via wiring?
3. some tape decks such as Sony's (DVCAM) supposedly have t-code generators; what for and when are they used? (these appear very expensive!)

On the latter, above, if one is looking at archiving onto tape and need a generator also, wouldn't getting a tape deck make some sense.

Hope I'm not going way off...

Thanks again.

[Vic Owen]

Regarding the time code generator, it gets a lot more expensive, since you would need a different camera -- the XL-1(s) will not accept it.

Using an infared controller would likely not be frame accurate -- you'd still need to tweak the clips in post.

When archiving onto new tape, the DVCR lays its own time code down. Conversely, the VCR will read whatever time code is already on a recorded tape during playback.

Try to keep it simple -- the more "stuff" you introduce, the more chance for problems.

[Alexander Ibrahim]

Look the XL-1 uses DV timecode, and the end result is that you most likely do not have to worry at all.

Using Premiere Final Cut Pro or Vegas Video or whatever you use to edit. You line the videos up at the start, and you are done.

You can line up the video using a flash, or a particular motion on screen, or a clapbaord in the audio...any distinctive thing that is clearly observable on all camera recordings.

That is all there is to the basic answer.

What follows is gravy. You will note that I like just a dash of meat with my gravy;-)

I usually shoot and edit this scenario with two cameras very simply. One camera is tasked with shooting a safety shot, a composition that will always capture the subject acceptably. Once the safety compositon is set DO NOT MOVE THE SAFETY CAMERA AT ALL.

The other "B" camera takes all the close ups etc.

In editing, with Premiere, I place the safety camera footage on the V1 timeline. I then place the B camera on V2 and synch it. (always roll your safety camera a second or two before the other cameras, so you have slack.)

You can add as many cameras as you like on V3 on up, but I'll ignore that for now.

Once they are synched, you just cut away all the problem footage in the B roll, and you are DONE.

The only possible problem is that you may have a defective camera which misrepresents time during the shoot.

To test for this, before the shoot, point all your cameras at an ANALOG CLOCK with hour, minute and second hands. Doesn't matter if it keeps perfect time.

Shoot them starting at some easy to see time, like 12 (noon or midnight matters not.) let them run for an entire tape in LP mode.

Capture all the footage, line it up at the beginning, the move the cursor to the end and look at both clips, they should be in perfect lockstep, if not you have a problem...but is it really the camera ?

I have never seen an operating MiniDV camera that lost time after a single tape even at LP. My experience has been that any problems with this sort of thing has to do with the NLE system, not the cameras.

NLE systems might capture takes at different speeds depending on the CPU load, the codec efficiency etc etc. You can see this on cheap NLE solutions by capturing the same footage from the same deck twice and comparing.

Fixing the NLE systems representation of time can be more problematic by far. Especially since most users build their own NLE's (badly I might add) and are thus their own tech support.

I can say this most generally....OHCI card based editors are most susceptible to these problems. I do not mean cards like RT2000 or DV Storm, but rather simple IEEE cards. This certinly doesn't mean that more advanced cards are impervious to problems...just that they have some built in ability to fix problems.

[Joji Minatogawa]

Ok..., I will try to keep it simple.

Thank you all for all the explanations; I do appreciate the extra gravy, too!.

[Ozzie Alfonso]

All the bases have been covered. There's little for me to add except that I'm facing the same problem - keeping two XL-1(s) in sync. I've done away with any attempt to use the timecode. The clap board is my way of getting a common sync point. In your case - recording in a place where a clapper may be too intrusive and/or the cameras are too far from eachother - I would use a common source of light. The small battery powered laser pointer is a good source to use since it's crisp and can be easily seen by both cameras even at a distance. Just a flash or two is all you need. In editing, sync up the flashes.

The idea of shooting a common clock from time to time is also a good way to double check where in the tape you are while checking the sync.

Tomorrow is my first shoot day on what promises to be a bear of a shoot. I'll let you know how my clapper works.

[Joji Minatogawa]

Thank you Ozzie,

Laser pointer is a great idea!

Good luck on your shoot tomorrow and indeed let me and others know how the clapper works!

Thanks again!

[Don Donatello]

i just finished a 2 camera shoot ( 2 GL's) . this past week end i sync them up ... time code doesn't matter because you are syncing one camera to the other ... i fed a wireless boom mic to both GL's ..on GL A i also fed a boundary mic on 2nd channel .. on GLB i recorded the boom mic to both channels ...

i used Vegas Video to sync them up ..very easy because VV lets you expand the audio waveform HUGE and you can visully see the audio waves ... i sync up to the 1st Voice. (actually the camera's were 1/2 field out of sync - picture and sound)

i recommend that once the camera's are started do NOT turn either off during the performance you are recording ...

once i had them in sync ... i then edited my selects .. then copy/pasted the boundary mic from GLA onto the 2nd audio track of GLB ( had to slide it 1/2 field to match the left channel)
so now both camera's audio tracks are stereo ...

i did find that the audio drifted about plus/minus 1/8 to 1/4 of a frame apporx every 10 min ... i could NOT hear it but with Vegas i could SEE it ( audio waveform graphs)

i then rendered out each camera to a new clip - now BOTH clips have same time code ... i transferred each clip to it's own DV tape (both tapes have same time code ) i'll edited it in the future ..

i also made a WMV8 file of the clips using split screens with burned in TC window that matches the tapes. so i can view both images at the same time ...

[Joji Minatogawa]

Hi,

Thank you for your message.

Was the audio drift due to the cameras during playback or during the capture to VV?

Also, would having two cameras of same model help in synching (time code) as you have done?

Thank you.