One Speaker giving a long presentation
I've been asked by a client to film her 8-10 hour training session delivered over two days using three cameras. The end product will be divided into 20 min. segments and streamed on a password protected video server such as Vimeo. The video is for internal use only to train trainers and not for marketing or selling.
Later we plan to edit the multicamera material using Premiere Pro CC. We'll also cutaway from the speaker to stills at a rate of approximately 30 stills/clips per hour. Because she may respond to questions from the audience we may also need to source referenced visuals after the fact. (The client will provides these)
From everyone's experience how many days editing would you estimate this project would require? Thank you in advance for offering your advice!
If tape-based, that's up to 60 hours, just to ingest and log the tapes. From cards or hard drive based recordings, you can load in much faster, but it doesn't make the logging any faster. It could be useful to have an assistant whose only job is logging, based on time of day time codes. Run all the cameras with free-run time code set to time of day, and start and stop them simultaneously on a cue over the intercom. Then anybody, anywhere in the room, can log in real time, just using a wristwatch. They don't need to be near a camera or even a monitor.
While the multicam in Premiere is good, realize, in the edit, in general practice, you're more or less going to play those streams in real time and try to switch live, on-the-fly. That's three to four days of stopping, starting, re-winding and faking the live switch you could have done in the first place, just one time. There is little time savings to be had there; editing multicam is what you do when your edit time is free or very cheap, and you're gear-poor, or you're unable to switch live at the location for some technical or financial reason, ...or you're just scared to, and want the assurance of being able to undo every decision and make many changes. None of which shortens the post process. So, I'm going to say ten days total for the edit, conservatively. It *could* be done faster, by someone with skills. Maybe as short as six days. They'd be very long, tough days. My rule of thumb is that at a bare-bones, dirt-simple, cuts-only minimum, an edit takes three times as long as the master takes, in real time. it's usually more.
For two-three days' worth of raw material of this kind of thing, with three cameras and a graphics source, you REALLY want to rent a switcher and live-switch the bulk of this at the event, saving the edit time for minor clean-up, color correction, audio sweetening and adding in any stray graphics; it will be a HUGE time saver AND thus, money-saver. I can't emphasize that enough. You can still roll ISO recordings in each camera, certainly just to have insurance, and to make fixing any blown takes easier. Loading in two thirds less footage saves time and drive space, fixing a few blown takes here and there will not cost much time.You can always go back to that, which should make you more self-assured in the live directing.
Seriously, look into renting the switcher and intercoms for this project. For something this big, it would pay to BUY the switcher and have it for all the gigs that come after it. If you have the money for sixty hours of high def storage in hard drives, you have the money for a minimalist HD switcher/recorder setup.
If I can make one more suggestion, or maybe it's more of an observation, based on doing exactly these kinds of jobs, for over two decades. In terms of instructional design, it is often the weakest approach to just record an existing live speech and try to cut it up into manageable chapters or chunks. It's more work, but I think you get a more effective teaching tool if you de-construct the material presented and rewrite/re-format/shorten it to make the most of what video and audio do best. I have found that a lot of hour-long speeches could really be collapsed into fifteen minutes of quality video, if you take out a lot of the redundancies built into speech-giving, and let the viewer decide for themselves if they need to back up and play a segment over. That's just one aspect of how shortening improves the video. I see how long your source is and I cringe, thinking about what viewer will sit thru that, even in a chapterized form they can self-navigate, skipping thru parts they don't need. It almost certainly must be some kind of mandated viewing; nobody binge-watches training materials if they can help it. So, they're being ordered to watch this stuff - they may not be actively seeking it out. That's a tough audience to begin with. How much do you think they'll retain? And is giving it to them in a realtime speaking format really more efficient than letting them read a PDF file with hotlinks? I'm just saying, you can, and perhaps should, divorce the key content from the format it originally came in, to make it more suitable and effective for the medium and method of consumption. That's a battle I'm always fighting, and I don't always win. But when I do, the results are satisfying.
I really appreciate your detailed response. I think your advice to switch on site and record simultaneously in the cameras makes the most sense. Your editing time estimation of 30 hrs of raw footage matches mine which is reassuring. But as you mentioned who wants to store hours of material on a bunch of drives.
We were planning to chunk down this person's talk in post. The gentleman presenting the talk is one of the most extraordinary speakers I've ever encountered. He can talk for ten hours without any notes or supporting material and his audiences are riveted (no exaggeration). I edited an interview with him a couple of years ago and every statement he made was fascinating, detailed and precise with no flab to cut!
The final indexed segments of his long presentation will be used to train other trainers or they can use him in their own seminars to explain concepts no other person on earth can deliver with as much authority and knowledge.
With that said, I will certainly be looking for ways to condense and compress. I agree that it has to be divided into digestible chunks or viewers will bolt.
I have just a quick technical question, I usually record in 1080p when the end product is streamed on the net or projected from a laptop. Most of the switchers I've investigated are 1080i etc with no progressive options with the exception of the BlackMagic ATEM-1 M/E. Should I go progressive or interlaced on this? Again, I'm grateful for you opening this discussion with me.
I don't have a strong opinion on the progressive v. interlace, except where the video will have a lot of graphic cut-aways to stuff with very tiny details, like gargantuan spreadsheets or charts or photos with super-fine details. In those cases pursuing progressive may have a payoff.
Do you shoot everything in 1080? Call me crazy but I like working in 720 over 1080 for my news shooting and for streaming purposes. The sports guys seem to split on their usage depending on how kinetic a sport they cover: golf is pretty static, football very dynamic and motion-intensive. If I am recalling this right, 720 works better in the high-movement coverage and 1080 for detailed frames that are mostly still compositions. Extending that thinking to your lecture, does this "super-speaker" lecturer stand pretty still, or are your camera operators facing a shooting gallery duck target who runs all over the place? Does using 720 make streaming easier?
Thanks Mark. The Lecturer likes to bob and weave when she talks so we may be chasing her with the camera all over the stage. I just prefer the look of progressive film look over the TV look I see in interlaced material. Although, when I think about it Vimeo prefers material in 720p and down converts 1080p to that format so it plays properly. Maybe 720p is the way to go. Although I'll sort it out with my crew this week and decide. Thank you again for your input.