Straggling video from yesterday’s post finally live … stereographer Sean White on the mission to rig a 3D camera system portable enough to shoot the prep and implosion of a condemned sports stadium in Salvador, Brazil for the explosive demolition series Blowdown:
The great lens showdown is over: Meuser Optiks it is. After an intense push to After an intense push to choose lenses for our A cam system so we can shoot our first 3D documentary, these German lenses – 3.4 mm, designed for a 1/3-inch CCD sensor, and HD capable – were the ones that made the cut.
They work with our Iconix 1/3-inch sensors, and the interaxial distance can be set close enough to allow us to fil 1 ½ -2 metres away from our subject.
Another to follow ASAP on why the system should work to shoot the prep and implosion of a condemned sports stadium in Salvador, Brazil for the explosive demolition series Blowdown – YouTube’s fighting me.
This wouldn’t be a huge problem if the stock mirror that came with the rig were up to QC par.
But it isn’t – which is why we were using the backup in the first place.
Here’s the issue: one side of the stock mirror is coated and one side isn’t – the coated side is reflected to the camera that’s mounted on the top of the rig.
This disparity seems to cause a difference in light between the two cameras, to the order of one stop (ie. the upper camera gets about half the amount of light as the horizontal camera).
Our stereographer, Sean White, also found that there’s a slight colour difference between the two cameras, and a mild fogginess to the stock mirror.
We plan to use this rig to shoot the implosion of a condemned sports stadium in HD. And we deal with high-profile broadcasters – we have to deliver high-quality footage (for them and, more importantly, for a worldwide audience).
So Sean grabbed the spare – which we bought in case we break the original one while filming on an industrial site in Salvador, Brazil for the explosive demolition series Blowdown – to see if it gave us more consistent picture.
It did – there’s only 1/3 to 2/3 of a stop difference between the two cameras, the colour is almost matching, and it’s fog-free.
Though none of us are sure why this is the case, we moved to film with the superior mirror (ie. the backup became the starter).
But while Sean was re-screwing it into the rig, it cracked.
The placement process is quite a balancing act – screw the mirror in too loosely (or mess up the gaffer tape), and it threatens to slide out the bottom when you position it at 45 degree angle.
Sean thought he had found the sweet spot for tightening, but he was used to adjusting the stock mirror … the backup shattered under a similar amount of tension (which makes us think the first mirror may contain more plastic, but that’s just a theory).
Needless to say the setback is disappointing.
Sean got in touch with 3D Film Factory. They won’t exchange the stock mirror for one that’s like the backup. But they will sell us another backup at a deep discount.
Granted, this rig is cheaper than most of the other ones out there, so we can’t expect the moon.
But considering how integral a working mirror is to the system, hopefully quality discrepancies and the finicky screw-in process will be rectified in the future.
Vision alert. So we’re thinking of designing our own mini beam splitter rig to shoot close ups for our first 3D documentary . Here’s why: as I’ve mentioned,we’re trying to sort out our B cam system using Film Factory’s 3D Indie BS Rig – this should hopefully work for the majority of our B cam shots.
But it isn’t the most portable system – something that will surely prove challenging on an industrial site, where the explosive demolition series Blowdown takes place.
For the odd close up of the crew loading explosives etc. in a condemned sports stadium in Salvador, Brazil, it’s painful to think about moving the whole rig and setting up just to get a shot or two.
And we’d like to avoid using a side-by-side rig because it’s a mission (if not a mission impossible) to get the cameras close enough together to film these types of shots.
If there’s a way to build a mini beam splitter unit, one person could just move in and get this footage with way less trouble (ie. instead of two people to set up the full-sized rig, etc.).
As an added bonus, we could also use this system in a temporary pinch if the main B cam system happened to go down.
We bought an extra mirror anyway, in case the one that came with the Film Factory Indie rig system malfunctions mid-shoot … why not try and put it to use?
Our stereographer, Sean White, knows a machinist who may be able to help us build this mini rig we’ve envisioned.
We’d probably use Canon Vixias – smaller than the Sony EX1/EX3 combination we’re trying to make work with the Film Factory Indie rig – and rig it as an underslung system.
One camera would be positioned below (instead of above), and the spare mirror would be angled the opposite way as the one on the full-fledged beam splitter rig.
This rig would theoretically snap onto a tripod, and would achieve a small enough interaxial distance for us to get the close-ups we need without a mega hassle.
If any stereo, etc. issues arose, we’d correct them in post – theoretically a relatively small workload considering we don’t use tons of these types of shots in our shows.
At the moment, it’s just an idea – but seems fittingly par for this unchartered course.
As I’ve mentioned, the 3D files that Cineform creates only have two audio tracks.
To capture ambient noise as well as a conversation between two subjects for the explosive demolition show Blowdown, we have to capture at least three channels (a boom mic and two lavs), sometimes four (camera mic).
Our editor, Brian Mann, has been in conversation with Cineform developers to see if we could find a way to edit with more than two channels.
They’ve been very prompt in replying and helpful.
But unfortunately it looks like there’s no way to edit more than two channels of audio using the current version.
There’s no particular reason why the program’s this way – it’s just a design factor that isn’t optimal for our specific post production needs.
As far as we’re concerned it’s the best high-end game in town, and otherwise it’s working great.
Cineform’s lead Mac engineer plans to add it to the list of things to add to their future release, First Light.
In the interim, we’ll have to figure out how to adjust our workflow.
It seems we’re not alone in the quest to find a beam splitter rig/camera system that will work as our B cam setup for shooting our first 3D documentary.
My previous blog outlining the issue has been posted on the Yahoo! Group thread P+S Technik 3D Stereo Rig + 2 Sony EX3, where there are entries from several people looking for information on how to make this – and similar systems – work for them.
Here’s our particular issue: we’ve been testing a Sony EX1 and a Sony EX3 mounted on a Film Factory 3D Indie BS Rig, but are having issues getting the cameras optimally positioned.
One alternative mentioned in the Yahoo! thread is a P+S Technik rig, but from what I gather it’s more expensive than the Film Factory Indie unit we’ve purchased.
And it’s almost time for the ultimate showdown: pitting three different models against each other, head to head … to head.
As I’ve explained, the winning candidate will ideally work with our Iconix 1/3-inch sensors, capture in HD, and allow us to film anywhere from 1 ½ -2 metres away from our subject to as far out as we want to go.
Easier said than done.
The Schneider Cinegon 5.3 mm lenses we ordered from New York are meant for a 1/3-inch sensor, but they’re not designed to shoot in HD, so I suspect the quality will be too low.
Our HD 3D monitor has arrived!
Moving to polarized is a relief – no colour loss, no red and cyan ghosting … and no headaches.
Purchasing the monitor ($6,600 later) was obviously a must – you can’t produce a high quality 3D show, like the first 3D documentary we’re going to shoot, editing in anaglyph.
We managed to fit it into an edit suite and set it up. Brian, Jakub and I check out some VFX footage on our newest 3D toy:
When it comes time to cut this Blowdown, Brian will use the monitor to see what he’s editing in Final Cut Pro (which, as I’ve mentioned, can’t edit in 3D without third party program help – we’re trying Cineform Neo3D out).
This view is key – cutting 3D shots means there’s a lot more to consider – parallax, convergence, wide and close, how much positive depth/negative depth exists in each shot.
If you cut shots with huge discrepancies in depth it’s really uncomfortable to watch, so you can’t just chop shots together – even with a flashy transition.
The only rub with the monitor is that it doesn’t do dual stream, which means the footage is technically at half-resolution (ie. don’t get both eyes full res).
So even though we’re editing in dual stream (to deliver the highest quality possible), we can’t view it that way on the JVC screen.
We looked into dual stream monitors – Panasonic’s due to release a 25-inch unit in the fall – but it’s prohibitively expensive (approximately $10,000).
And, more importantly, it’s too small for us to view our footage in a size that’s representative of our final product (how many 25-inch televisions have you seen lately?) – a shortfall that could lead to convergence that isn’t optimal for our audience.
Oh, and the JVC came with two free pairs of polarized glasses.
Looks like we’ll do just fine:
Editor Brian Mann works with the JVC HD 3D LCD monitor
Shooting our first 3D documentary’s not only going to change the way we capture content in the field, it’s also going to drastically change the way we store it.
Not only will this be the first show that my crew shoots tapeless, they’ll also have to contend with twice the amount of footage.
And if they lose any of it, it could be a huge disaster (think: Blowdown without the implosion. Yikes).
So we’re looking into a system that will allow us to move footage from camera/nano3D compact flash cards to a storage unit during the day, the transfer it into a mega conduit each evening.
Our data journey would start with ShotPut Pro. This copy utility application automatically copies and verifies all transfers off of flash cards. It can also copy multiple cards to multiple hard disks at the same time.
We’d use ShotPut Pro to transfer our footage to a G RAID mini. Using its RAID 1 setting, the crew would put two copies of everything onto the mini’s two SATA drives. This redundancy means that if for some reason we lose one drive, we won’t lose the farm.
Each G RAID mini stores up to 1 TB of data, so we should be able to carry our footage (up to 500 GB, copied twice) on it until the end of the day (if the crew’s shooting more than that amount, they’re shooting too much!).
Each evening, we’d then transfer the footage from the G RAID mini to a G SAFE. Each of these storage units takes up to 2 TB of data, and only stores RAID 1 (mirrored), which means two copies stored no matter what.
The data journey would end when both 7200 RPM SATA II drives are removed and shipped back to the production house separately, in case one gets lost in transit.
Approximate cost: $100 for ShotPut Pro, $300 for the G RAID mini, $700 for the G SAFE with two drives (ie. the first 2 TB of storage).
After that, we’ll be buying drives just like we bought tapes – I’m interested to see how costs compare out the other end.
And another first – to keep track of audio tracks, locations and dates (in lieu of the tape, sticker and marker technique) we’ll be using an electronic slate, courtesy of the iPad.
Good news on the post front: we’ve upgraded from Adobe CS4 to CS5.
And the stereo scripts our compositor will use to create VFX for our first 3D documentary work with the newer version.
We’ve been gearing to get CS5 running since we moved from Leopard to Snow Leopard to take advantage of the 64-bit architecture and improve workflow.
But our VFX artist, Jakub Kuczynski, was concerned stereo 3D scripts he found online that have given him a much more efficient pipeline for stereoscopic workflow in After Effects wouldn’t transfer over smoothly to CS5.
He contacted the scripts’ developer, Christoph Keller, to ask if they’d be compatible, but he didn’t know.
Now we do. And it’s very good news – work that would take Jakub a day to do manually takes him about an hour, thanks to the scripts.
As for the CS5/Leopard upgrades, we haven’t noticed a marked increase in speed, but even a little more juice over the long run means more efficient post production overall.