Make sure you've got scratch track sound on every camera PLUS your "good mix". Trying to get things in sync w/o audio will be brutal.

On multicamera shoots, I use cheapo digital clocks (like $10 each) all preset to the same exact time. Then anytime someone starts a camera he quickly shoots the clock FIRST. Then he doesn't cut until the end of the roll.

Once you've laid the first camera into the time track, set the track code to sync with the clock. From there on in, it's a piece of cake to sync ANY of the cameras to that main clock.

Stagger the start times of each camera and especially the STOP times. You don't want to have two cameras changing rolls at the same time; not when you've got four running.

Start the first camera (locked off) at least ten full minutes before the show even looks like it's going to start. That way you won't miss the start. (They promise to tell you when it's starting.... THEY LIE ! )

Don't worry TOO much about color balance - it's a r & r show - if the cameras are horribly off, cut to black and white - then to the next camera.

Hope this helps,
v

Also, don't try to mix 24p footage with 30fps unless you are going for an effect.
Uh...why not? Do you mean don't put 24p on a 30p timeline? 60i timeline? 60i on a 24p timeline? 30p on a 50i timeline? 30p on a 24p timeline? It's done every day...by hundreds of editors. In both directions. In fact, Avid is making a large campaign on just exactly how well their products do it. Vegas has been good at this for 3 revs now, and terrific at it with Version 6.
Since David specifically mentions his VX, it doesn't do 24p anyway, nor does the PD 170. Shooting VX and PD at 60i, and shooting the DVX at 24p will be just fine, if he wants that smooth look, especially on pans. That said, I'd also somewhat assume he's shooting the DVX in 60i anyway.

Were it me, David, I'd lock down the VX's, and use the others as handhelds or panned cams. Shoot the VX's fairly tight since they're not as pretty as the PD's are when it comes to wide.

Anyway David,....Victor's time method works great. Kinda like a visual jam sync. When he told me about it I thought he was nuts. Then we tried it. He's not nuts.
If you can store two white balance settings at the same time, do it. That'll *help* Be sure to slightly underexpose rather than over expose, which is easy to do in a rock show. Lotsa changing lights and proximity can make it tough to keep exposure happy. Be sure to calibrate/adjust your LCD so it's at least close to what your eye sees.
Avoid Autofocus like the plague. Not just cuz it's difficult, but fog *really* messes with it.
Make sure your "good" recording is at 48k. Bitrate doesn't matter much, so long as it's 16 or higher. (which suggests 24 if you've got the space, the extra dynamic range is nice)
If you can get one cam to do dutch tilts or off angles, those are always great for cutaways. It's usually worth it to shoot an extra cam as a crap C roll cam with a 12 year old running it, as long as he actually points at the subject, simply because sometimes you'll get something cool in there. Even a junker cam looks good for closeups, and you can always muck it up with a filter.

Sounds like a fun shoot. If it comes out anything as good as the project you shared in Dallas....I'd like to see it.

Stage lighting can be brutal, and to get faces right, you often have to lose the background. Err on the side of underexposing.

I agree that the VX's should not be left on automatic exposure. Try looking at the exposure data being recorded on your trial shoot to see what looked good and set your locked down cameras to that.

The Sony's have a very obvious step in brightness when using the exposure wheel. Try the custom preset function instead where the exposure is essentially continuous.

Forget auto, white balance and check monitors,,, being you are using diff cams you will be in for some cc in post, but maybe not too bad. Check the monitor and set each cam as close to each as possible , and all will be fine,,, I agree with gary, shoot all the same 24,30,60 whatever.

Thanks for the tips everyone. The digital clock idea sounds great - we have lots of leeway with the artist (I was his bass player for 3 yrs) and he's prepared to take a 2 min break on our cue for tape changes, if we wanted to change all at once. It's a shoestring deal for sure, but should be fun.

And thanks Spot! I'll keep you posted...

I shot a Christmas play with three cameras last year, and was amazed at how much flexibility I had when editing the three together. At least one camera had something interesting.

I remember watching the Yanni "Live at the Acropolis" video about a decade ago on PBS and being impressed with the large number of video cameras used. I would guess about a dozen, when including the ones on stage, down in the audience, flying overhead, mounted to the drummer's stage, etc.

OK, the shoot has come and gone and it seemed to go well. I haven't reviewed any video yet; the first thing I fire up is the audio mix, and the isolation is surprisingly good...levels are good (bass is low but can be normalized)...but what the....

ALL of the audio is about 25% slower and lower pitched than it should be. I've searched the forums and it looks like it may have something to do with an incorrect bit depth setting, or having it locked in when it shouldn't be (or not locked in when it should be). I'm not in front of the gear right now, but let me describe the recording vs. playback environments to see if this is workable...

Recording - New-ish Layla 24 bit unit locked at 48K as master, original 20-bit Layla as slave connected via BNC video cable as specified. The newer 24-bit unit used it's internal clock, and the older 20-bit unit used (I think) Word clock from the master. All recording seemed to go fine.

Playback - The 24-bit Layla was borrowed, so all playback, mixing, etc is being done on my older 20-bit unit. I haven't changed any settings on any control panels or within Vegas, just fired up the project and played a couple of sections. If I adjust the playback rate slider to about 1.13 or so, it's pretty close. But I really, REALLY don't want to do this. Can I resample in Forge somehow, or can I use the 24-bit Layla for playback and mixing? Anything? It needs to be sync'd to video so the proper pitch and speed need to be intact. And, to start with, the band wants to review just the audio in rough mixes to see if anything needs to be fixed.

Perplexed.....

You can resample in Forge, or you can use the 24 bit box, either way should get you set up. Sounds like you've got sample rate differences more than anything else. Forge is great for converting this

Thanks Douglas, I'll try that out tonight. As long as the answer isn't 'scrap the whole thing, you screwed up" I should be OK.

I run 4 cameras in all my wedding and concert shoots. I have a created a very automated approach to sync'ing the multicam clips:

Step 1... Instead of the 'shooting the clock' idea, I simply sync all my camera clocks before the shoot. Go to "time set" on all of them, set to the next minute, then hit start-clock on all 4 as quickly as I can. Maybe a second or two off from each other, but no more.

Step 2... For capture, I use SceneAlyzer (www.scenealyzer.com) (~$25 shareware download). it puts the camera clock time as part of the avi clip file name. (Sony PLEASE make clip date/time available via script in a future release).

Step 3... I wrote 2 Vegas scripts. One that loads all media from a directory onto the timeline. I use a convention where I start all clip names with 1, 2, 3, or 4 for the camera number. The script loads all of the clips for each camera on a separate track. The second script places the clips at the precise time on the timeline when they were shot... e.g. clip name (named by SceneAlyzer): "2 Sternfeld Reception A'20050924 20.14.57.avi" will be put in a .veg file named 20050924.veg and placed on the time line at 20:14:57.0.

So in a couple of clicks, your entire multicam shoot is loaded into a project and completely sync'd up (minus some minor tweaking) and ready to begin edit.

If anybody is interested in the scripts (as-is):

http://www.2ndGenerationProductions.com/vegasscripts/LoadAllMedia.js

http://www.2ndGenerationProductions.com/vegasscripts/multiCamSync.js

Jerry

Jerry,

Very timely. I've downloaded the scripts and will try your technique this weekend.

I am about to shoot what I think will be my 40th Nutcracker (I used to do three per year, but I only shoot one now). I'll only be using two cameras -- one for the principle dancers and the other for entire stage. Works pretty well, and is easy for me to handle by myself.

If this works, I'll use it on a dance concert I'm scheduled to do in January. I'm toying with the idea of using three cameras for that one.

John, I tweaked the scripts just a little before uploading to make them a little more generic and a little less 'custom-for-me'. If by chance I messed something up and you have problems, please email me using the contact email address on my web site. Feedback is definitely welcome. I said 'as-is' just as a disclaimer. I'm very interested in making these scripts usable for others.

Good luck with the Nutcracker...

Jerry

Jerry,

No problem. I'm reasonably "script literate," so I'll have no trouble tweaking them, if necessary.

John, I have some other mulitcam-related scripts that might be useful for you as well. If you think you can use them, I'll clean them up and upload as well. (I love editing video... I detest mundane repetitive time-consuming and error-prone manual tasks... So I write scripts to automate everything I can think of...)

1) Std multicam edit -- like others that exist... use markers 1, 2, 3, 4, etc to cut to that camera. Unlike some other similar scripts, it doesn't move all clips to a new track (hard to do track-level effects, etc when clips from many tracks get merged into one). It edits each camera track. Customizable fade time. Audio tracks are untouched. Cannot remove/compress 'time' from the project with this script.

2) Similar to #1... but a marker with '0' cuts out the block of time and compresses. I use this one for multicam reception dance segments where I may want camera 1, then camera 2, then remove a couple of minutes completely, etc.

3) Multicam block remove -- This allows you to take a fully edited real-time full-length segment ("Dance of the Sugar Plum Fairies") and cut it down to a 20-second highlights segment perhaps for a lobby trailer. Cuts out everything in the time block and compresses the freed space. Keep only blocks between a '1' marker and a '0' marker. Discard and compress everything else. Keeps all audio/video in sync inside the preserved blocks of time.

Let me know if you'd like me to upload these as well. They may require a little bit of 'genericizing'. But I could get them uploaded fairly quickly.

Jerry

DSE wrote:
You can resample in Forge, or you can use the 24 bit box, either way should get you set up. Sounds like you've got sample rate differences more than anything else. Forge is great for converting this


A new twist, and we can take this offline if need be. I went into Forge and didn't see anything between 24 or 16 bits. This box is 20, and I'd like to go no lower than 20 (honestly though, not sure why, it just seems like the thing to do. Or not do.)

So without doing anything in Forge, I opened a new project in Vegas, dragged the wav files in, and it is playing fine in terms of tempo and pitch. But, about 20 - 30 minutes in, some of the tracks are a little out of synch with the others. That must be the 20 vs 24 bit difference. But, how to "downsample", and what setting to use? 16 bit? Should I down sample all wav files in the project?

notes - It was all recorded originally at 48K; the driver model for recording was Windows Classic Wave Driver (this is the only setting that would allow me to record 16 tracks any longer than 4 minutes at a time) and now on playback it is ASIO Echo WDM.

Jerry,
I would be very interested in the multicam scripts. I often have to slice out sections of the program and often have problems with sync. The block remove is intriguing also.

Regards,
Jerry

Yes, I'd definitely be interested. I am still using a work-in-progress script written by someone who then decided to make the script commercial, removed the script from free distribution, but then never finished it. The script you have for doing the highlights for the lobby trailer, as you call it, would be extremely useful to replace this older, not-fully-functional, script.

(bump)
Curious if Spot or anyone else had an answer to my last post in this thread. I guess I'll experiment with 16-bit resamples in Forge.

In Forge, set file properties to desired bit rate.
In Forge under "Process" find resample.
Set Antialias filter on, set accuracy to high.
Since your pitch is off, you just need to set the sample rate only, or should just need to do that.
Bit rate isn't related to your pitch issue, but your sample rate is.

Thanks for the tips Spot. But pitch isn't an issue now, it's timing of some of the individual tracks - as I said, I'll take it offline if need be (side note - also sent via email a few days ago)

I created a new project in Vegas, dragged the wav's in, and after about 20 mins or so the drift gets worse - for example, between kick and snare. I'm assuming one was plugged into the Layla 24 and one was plugged into the Layla 20. Sorry to sound like a broken record - I tried all the steps above, even changing down to 16 bit, and the timing is still off. It starts fine, and starts to drift.