"Gone Nuts 3D" Takeaway #3: Move Over Roll Angst, Meet Zoom Disparity

3. Move Over Roll Angst -- Meet Zoom Disparity

The Day Zoom Disparity Became My Top Nemesis

During the introductory film production labs I teach, I have seen the gamut of problems beginning videographers face: forgetting to white balance, neglecting to use neutral density filters, using the wrong neutral density filter, and my favorite, the infamous camera-bag-and-tripod-in-the-shot.  (Granted, you did a great job not leaving your gear unattended!)

Keetin in Booth 3: Home of the S3D editor

Stereoscopic 3D introduces a whole new field for mistakes and disparities that arise between the left and right clips.  In editing the clips together, I performed vertical image translation (VIT) to change the vertical positioning of one clip in order to align the clips.  After VIT, the two images still did not match.  So, I changed the rotation of one the clips.  Still, something was off.  After minutely adjusting the VIT and rotation, I was no closer to adjusting the problem.  Then I recalled what the stereoscopic editor from the first project had mentioned during their project post-mortem: for some shots, he discovered the cameras were zoomed to different distances.  I opened up the Sony XDCam Browser to examine the metadata for the clips and see if this, too, was my problem.  Sure enough, for a handful of clips, the camera zooms were set differently. 

To fix the problem, I scaled one of the images up and proceeded to adjust the VIT and correct the roll.  Voilà!  The disparities were fixed and the clips were for the most part aligned.  

S3D requires an even more watchful eye on the part of the stereographer who now has double the cameras to monitor and watch for small problems and mistakes.  Thankfully, zoom disparity is a detectable problem and easy to fix during production and is fixable during post-production.

 Now whenever I am editing and cannot align the two clips, I instantly suspect a zoom disparity and check the metadata.  The sneaky thing about detecting a small zoom disparity is that the size difference is very, very subtle (especially in grass). So, when in doubt, check for zoom disparity. 

LESSONS LEARNED

1.  When importing your clips, never delete the metadata information.  This may prove to be very useful to you later, especially if your footage log did not include this specific information.

2.  When the two clips do not align, double check to see if there is a problem aside from roll or vertical disparity.

3.  Faithfully stick to the left camera to not only watch for the matte box but more importantly, to check that both cameras are on the same zoom.

4.  Always check with those who have performed your job before.  They hold a rich trove of experiential knowledge.

5.  A 1 mm zoom difference does not correlate to a 1% scale difference.  So, experiment to see which up-scaling percentage minimizes disparities.
6.  Editing angst is bonding.  The editors from the first round of projects enjoyed commiserating and after each project, another editor developed a distaste for zoom disparity.  The first round editors also watched the left camera like a hawk during the second round of projects because we understood the headache slight differences between the two cameras made.
7.  All stereographers should edit S3D sequences. Through the process of editing and aligning the left and right clips, I saw and grasped the different technical challenges in stereoscopic 3D production that I was aware of before because of our reading but did not understand.  I approached our subsequent projects with a changed eye, knowing what had caused me consternation in the editing booth and looked out to catch these problems during production.

"Gone Nuts 3D" Takeaway #4: Roll, Baby, Roll

4. Roll, Baby, Roll

The Day the Alignment Grid Became My New BFF

Grass is fun to roll in, but grass is not fun to roll to properly align the left and right clips of your stereosequence sequences.  Gone Nuts 3D was shot entirely outdoors, affording us a beautiful autumn film shoot.  This also meant that grass was in the background of almost every shot.  

Grass,

and more grass

    

       grass, 

I soon learned that when adjusting the degree of rotation of a video clip to align two homologous points, it helped if the video clips included strong lines and patterns.  These strong lines clearly display when two points are or are not aligned.  In contrast, grass is not a symmetrical and I could not anticipate which direction a given blade of grass should point.  This made it tough to know whether the left side and right sides of both video clips matched when one clip screened the other. 

After this project, our instructor Chris Eller introduced the alignment grid to our filming procedure.  The alignment grid is held within sight of both cameras for a few seconds after the slate.  This alignment grid provides the strong lines to align the homologous points of the left and right clips.  Example A below shows the left and right video clips screened in their stereosequence before any other alignment adjustments were made.  

Example A:
Alignment grid before adjustments

Example B:
Alignment grid after adjustments

The alignment grid introduced an easier way to align the images.  My strategy to fix roll is to adjust the rotation of one of the clips so that the diagonal lines of the alignment grid are parallel to each other.  (See Example B)  The diagonal lines of the two images may line up perfectly depending upon how you adjust the horizontal image translation (HIT).  The alignment grid is useful fixing more than just roll, but the alignment grid significantly speeds the alignment process.  Now whenever I am on a shoot, I monitor to ensure we always begin with the slate and alignment grid.  I am thankful for editing without the alignment grid because I more deeply understand and appreciate its handiness.   

LESSONS LEARNED

1.  Use an alignment chart to speed up manual stereoscopic adjustments like roll, VIT, etc.

2.  Stay current on tips from other stereographers 

3.  Take your instructors' advice.  Groups not consistently using the alignment grid regretted it!

Production Stills "Z-Day 3D"

This has been a busy week for the 3D rig and students with the behind the scenes filming for This Is Tango Now and a class production shoot about zombies, Z-Day 3D.  

Here's a sneak peek from each:

 

     Z-DAY 3D

When shoulder-mounting, closely monitor all zombies

 Keep enough brains on a zombie 3D movie set to keep everyone full

The zombie cast ready to mob

Production Stills "This Is Tango Now"

This Is Tango Now

Musicians setting up

This Is Tango Now show rehearsal

Audio, check!

"Gone Nuts 3D" Takeaway #5: Matte Box Beware

Takeaway #5:  Matte Box Beware

The Day Marking My Compulsion to Obsessively Check the Split Beam and Left Camera

Sitting down in our editing bay, an aptly named Booth 3, I assembled the right and left clips from our Gone Nuts 3D shoot into stereosequences.  I synchronized the clips and quickly checked for any other potential problems when I noticed a giant black shadow in the lower left corner of the left image and the tops of both the left and the right:

Raw Framepacked Clip 1095

I have framepacked the clips and highlighted the problems for ease of comparison.

What are those gigantic black shadows?  Only the Shadow knows...

Examining the field monitors on The Warehouse 3D/Chris Eller

Turns out that those shadows are from none other than our 3D rig's matte box.  In our split beam configuration, the left camera hangs underslung shooting into a mirror.  We filmed on a gorgeous September day which worked well for lighting our set but which worked poorly for seeing anything aside from glares in our field monitors.  This ruled out utilizing the field monitors to compare for side by side continuity. The scene was clear in the right camera, located on top of the rig so we never thought to check that the left camera was clear.  Oops.  

Close-up of the left camera and matte box/Chris Eller

As I sat and stared at the clips, each left clip filmed from this camera position had the notorious lower left shadow.  We had only filmed three takes of one shot from this position.  Thankfully, this meant I only faced the shadow dilemma once.  However, this shot was essential to our story as it was the moment where Arnold confronts and accuses the children of hitting him in the head with the wiffle ball.  

Fixing the top of the images posed no problem at all.  When I fixed the vertical disparity and roll, I scaled the images up in order to align them together.  In doing so, the top parts of the image were cropped out of sight.  Now, just the pesky bottom problem remained.

I tried lightening the shadow with brightness and contrast.  No success.  Someone suggested using Adobe Premiere’s 16-point garbage mask.  I combined my brightening and contrast with the 16 point garbage mask to isolate and lighten the shadow.  This worked decently and would have been an acceptable fix if this had not been in 3D.  However, the left image has such a strong shadow and the right image is so perfectly clear that the discrepancy causes retinal rivalry.  I was left with two solutions, minimize the difference between the two images or mono-ize the shot by removing the left and leaving the right.

In brainstorming solutions, someone suggested scaling the image up even more and cropping as close to the boy in the foreground as possible.  Though the shot is still slightly painful, it is much more comfortable to view than before.  However, this problem would have been avoided with a closer supervision of the left camera.  The consternation this clip caused raised an awareness in me of the necessity to monitor the left camera.  Because the right camera is easily monitored by the stereographer, the 1^st AC is key in confirming that the left camera image is also clear and good to go. 

 

Final Edited Framepack of Clip 1095

LESSONS LEARNED

1.  Recruit a 1^ST AC to monitor the left camera in lieu of field monitor viewing

2. Better to obsessively check the left camera in a split beam configuration than endure the pain of retinal rivalry and failed garbage masks

Sophie's Top 5 Takeaways from "Gone Nuts 3D"

Kate Braun and Frank Buczolich in Gone Nuts 3D (2010)

As university students, this is a prime time in our lives to try something, learn from it, try again, and repeat.  Each project we work on deepens our understanding of what we read in our textbook and weekly readings and what our instructors taught us during class.  As live-action stereoscopic 3D students, each project is an opportunity to learn more about the do's and don'ts of stereoscopic film-making and to adventure with other stereoscopic filmmakers doing the same.

Last semester, after each 3D group finished their project’s film shoots, we held a post-mortem the following class session.  During this post-mortem, the group members shared about what went well during production and shared tips for the rest of the class members to learn from.  This way, we could learn from each other’s successes and mistakes and make new mistakes instead of repeating the same ones. 

So, this week, I will be sharing my Top 5 Takeaways from “Gone Nuts 3D” where I served as the stereoscopic editor.  These takeaways cover what I learned from working as stereoscopic editor for this project and how it has guided my future stereoscopic 3D projects.

Sneak peek - Tomorrow's Takeaway #5 has to do with The Day Marking My Compulsion to Obsessively Check the Split Beam and Left Camera.

newly launched Stereo 3D Jobs website

Whether we are in the production lab waiting for a video to render or downing some coffee at our regular coffee spot, chances are good that the conversation will drift to finding a job after graduation at least once, if not more.

photo by Kenny Louie/Flickr

Instead of wondering if there are jobs available in the s3D field, he newly launched website Stereo 3D Jobs gives us opportunities to talk about.  Stereo 3D Jobs is a site for employers to list prospective s3D jobs.  The site allows the visitor to search using job keywords and location.  It also has a filter to display jobs by Type, Salary, Category, and Industry.

As the jobs continue to be posted, I am curious to read the descriptions of available jobs and the needed qualifications of applicants.  Never hurts to know which skills are desired.  Three cheers to the site's creators.  I hope the s3d HR departments post away. 

Telecom 3D Showcase in the IU Cinema, May 1st

Mark your calendars, the Telecom 3D productions will have their night in the IU Cinema on Sunday, May 1st at 6:30 pm. Dress up and come out to see some great 3D productions from Telecom undergrad and grad students. The event will be added to the IU Cinema calendar soon.


Congratulations to the 3D classes and independent graduate students for showcasing in the state of the art IU Cinema.

Ready for "47 Ronin" in 3D

photo via FirstShowing.net

One of the biggest complaints I hear about 3D movies is that audience viewers find them "gimmicky." (Future posts to come on the usage and meaning of this term.) This complaint is followed up by jabs at the movie's story, many of which are merited.

For this reason, the news Keanu Reeves will be starring in Universal's take on 47 Ronin in 3D brought me joy. The story of the forty-seven ronin (Japanese for "samurai") is rich in its themes of loyalty, honor, and sacrifice as the forty-seven ronin wait to avenge the death of their leader at the cost of their own lives.

From our projects last semester, we found action and fight scenes look sweet in 3D. For me, the depth displays well the spatial relations of the characters. In flat movies, I struggle to track each fighter and often find myself lost, not knowing where each fighter is in relation to the others. For this reason, 3D is helpful to me in following the action of a fight scene. 47 Ronin is a story that affords ample opportunities to visually utilize depth without throwing any ninja stars at the screen. (Though come on, deep down you know you would be sad if you left without one shuriken whizzing at your face.)

Universal gains points in my book for green lighting a 3D movie where the 3D will serve and enhance the story. Hope the film delivers.

For more read: http://www.firstshowing.net/2011/rinko-kikuchi-and-others-cast-in-carl-rinschs-3d-epic-47-ronin/

Welcome!

Welcome to the blog 3D IU Students.  We are excited to share what we are learning about stereoscopic 3D from our classwork, industry news, and our own experiences.


We look forward to connecting with stereoscopic filmmakers worldwide and gain insight from their work in s3D.  Feel free to contact us at our university student group's email:


novationproduction"at"gmail.com


Expect more great s3D videos, movie reviews, and posts to come!