Slightly OT: Multi-Camera Shoot – 384 Hours

[Kevin Monahan]

WHY must you capture so much footage? Couldn’t this process be shown in an hour or a day? How about time lapse options so you wouldn’t generate so much footage?

[Steven Ludlow]

Believe me…I’ve already pursued that option with the end customer. I tried to talk them into just recording every OTHER second — they don’t miss anything, and the footage drops to 190 hours. But…no.

I’m not editing it in the end…so it’s neither here nor there. I’ve just got to figure out how to efficiently capture it.

[Zak Mussig]

Steven,

I don’t have a lot of personal experience with setups like this, but no one had replied yet, so I thought I’d give it a shot.

That’s a massive amount of footage you’re dealing with. I would personally be very wary of trying to do that whole project without tape backups. Even if you can spring for that much storage, with tapeless formats you need to have your footage protected. I would suggest renting 4 HDV camera that support timecode in. Just change tapes in a waterfall pattern over the course of a couple minutes so someone is always rolling. The odds of you absolutely needing the shot from any of the camera in that minute or so are pretty slim when you’re dealing with that much footage. Even you want to do some sort of time-lapse, I really doubt you’d notice those gaps.

This give you tapes to pull from rather that trying to keep an insane amount of data around. Unless some part of the process happens at super high speed, HDV should suit your needs just fine.

Of course all of this is from knowing how long and with how many camera you want to shoot. Knowing what you want to do with the footage and the goal of the final product might impact some of these decisions.

Zak

“You can’t fix coverage in post.”

[Jeremy Garchow]

96 hours of of DVCPro HD 720p30N and two channels of audio is 2.59 TBs.

24pN is right around 2TBs.

You’d need to rent four DVCProHD capable cameras (such as the HVX200), four firestores and set them all up correctly.

I bet you could do that for around 5K.

[Bouncing Account needs new email address]

Just to stir the pot…

I have written down what I think you will finally end up doing on a 3×5 file card and sealed it in a mayonnaise jar.

My only hint is that you will end up NOT shooting and editing in the way that you said you are intending to do.

Once you get closer to the project, if you will post back with your revised plans (and I know you will have them,) I’ll unseal the jar and read you my prediction.

😉

[David Bogie]

I’ll second this observation.
If the shoot is so (freakin’) important to the client, it’s important enough to have a backup of at least half the cameras. That’s a RAID or tape.

We wish you good luck, we’re glad it’s you and not us, and I hope you’ll come back and tell us how you pulled this off.

bogiesan

This is my standard sigfile so do not take it personally: “For crying out loud, read the freakin’ manual.”

[Ben Scott]

for large amounts of hd footage on a budget cant agree more HDV is the way to do this project, tapeless cant be backed up and will be more hassle than its worth

also might be an idea to work offlineRT if the plan is to really cut the footage down, plus I wouldnt want to see multicam editing with HDV native footage and Prores or dvcpro might get more expensive storage wise than can be afforded especially if tapeless and therefore needing good redundant storage

learn offlinert workflow and using media manager and all the tricks to not mess up with multicam sounds like best bet

good to know if people have HDV working natively with 4 angles simultaneous multicam, if so capture in that and then finish in prores

make sure you use time of day timecode and then use the multiclip sequence feature, this is designed workflow for what you are trying to do. this should be easy to setup in camera if cameras time of day are in synch to begin with and wouldnt need timecode feed to each camera therefore reduce cost probably.

my bit finished!

[David M.]

I think I have a solution for you: you need a spreadsheet program like Excel. That way you can educate your client on the real cost of producing a shoot like this in column A – and in columns B, C, and D you can give him/her alternative solutions to get ‘er done that may fit into their meager budget. And based on your description it does seem meager for such a big undertaking. One caveat – you may lose the job – BUT you will retain your sanity and you will more than likely give your client a needed reality check in the world of video production/ editing – a world that is ALIEN and archaic to most normal people.

Take hold of the reigns brother. Steer the ship. Educate your client!!!

Just wanted to say that this year alone I have successfully media managed over 5 large projects with amazing success thanks to these fine folks here at the cow!!! Now that is saying something.

-DM

[Arnie Schlissel]

I’d rent 4 digital still SLR cameras, with intervalometers, and do this as a timelapse. If you experiment for a day or 2 with different intervals, I’m sure you can get this to work at a speed that will better serve your client’s needs than creating a 96 hour video will.

Arnie
Post production is not an afterthought!
https://www.arniepix.com/

[Steven Ludlow]

Thanks to all for your input. We’ve been hashing this out over the past few days.

I could have done a better job of describing the requirements of this project at the outset, but to clarify a few things:

1) Time lapse, generally speaking, is not an option. We are capturing the entirety of a massive manufacturing operating that takes 12 days to complete. The purpose is to track the movements of the various teams involved and study the order and manner in which different teams/individuals accomplish tasks. The goal is to ultimately lean out the process and improve the quality of the product. Teams of people will be poring over this footage like it’s the Zapruder film. (Now…I say time lapse is GENERALLY not an option, because I still think recording every other second cuts our footage in half, but won’t miss any critical movements either. And…we may do that.)

2) We’re in-house corporate…saying “no” isn’t an option. We do educate our clients, but sometimes the objective can’t be compromised, and it’s easier to throw more money at the project than it is to simplify it.

3) HD is not cricitcal. We wanted HD because, if we’re buying equipment, we don’t want to invest in SD technology. What we buy for this process study could easily be put to use in a variety of other more glossy Marketing capacities. And…even though we deliver in SD almost exclusively now, shooting a wide shot in HD still gives us the option of going tighter in post if they want more detail. (Which is likely…because they anticipate more uses for this footage than just process mapping.) This, however, was obviously the first criterion to be sacrificed.

So…

Our original plan was to mount 4 HDD cameras (more than likely, the Sony HDR SR-11), capture in SD for THIS project, and offload all 4 cameras at the end of each day’s shift.

But…logistically, this proved too time-intensive — hours each day just offloading, then figuring out how to deliver that volume of footage to the client without many MORE hours of encoding. (Not to mention, the requirements eventually increased to 10-hour days and 6 cameras).

We eventually threw a little more money at it and are purchasing a system typically used for security purposes. It’s a DVR that will record at 720×480 and utilizes MPEG-4 compression. It can handle 4 cameras at 30fps, or–in our case–6 cameras at 20fps. It’s fully programmable, and the proprietary viewer will make navigating the footage in the end very easy. We can also convert to QT for import into FCP, should we need to edit further. Once set up, all we have to do is go dump footage to external drives every couple of days. The cameras are 1/4″ single chip, but the quality will be sufficient for this project.

Given the intensive use of video in Lean Manufacturing (and other like process-improvement methods), I post all of this in some detail hoping it may help someone else in a similar situation. Or…maybe someone will see a better way.

If there’s a way to cross-reference this in the Corporate Video forum, it might be better served there.

Thanks to all for input, and don’t hesitate to provide feedback.

Steven