The State of 3D

I wanted to know who's doing 3D, how, and for what types of productions. Dave Helmly, North American tech sales manager for pro video/audio at Adobe; CineForm CEO David Taylor; and Sean Kilbride, Nvidia's technical marketing manager, workstation products provided the answers in email interviews.

What's your sense of how many videographers are now producing or trying to produce in 3D?

This is common question I get all the time and not an easy one to answer from the NLE side. We are getting a ton of questions about 3D workflows from the broadcast market, the independent events market and there is a lot of interest from our Premiere and After Effects user base about 3D which shows a huge interest. This is the main reason I made the 60 minute video.

Related Links

3D on the Cheap, Part 1 In this article, I'll discuss how to shoot, edit, and upload 3D videos to YouTube, and what you'll need to view them from that site... 3D on the Cheap, Part 2 Last time out, I detailed a consumer oriented workflow for 3D production, culminating with uploading a 3D file to YouTube. Since the YouTube video has only been viewed about 50 times so far, I'm getting the message that consumer isn't exactly the level that you care about...

What's their motivation?

Many of these customers say the same thing- the time to make money with 3D is now while it's still new and fresh. With new TV channels coming on line, the need for 3D content is growing. This includes local and national commercials. As you already know, there are a lot of desktop editors that make their living from national and local TV commercials and having the ability to do 3D stereo output gives them a leg up on their competition. We saw the same excitement with the HD. For some high-end wedding markets, 3D could also be an advantage. Many of these editors have more competition (lots of Pro video graduates with a sense of storyline, color, and proper use of effects) in the HD space and they need a new way to separate them from other HD editors.

What's the demographic? Are you seeing anything in the business market, or is it all indy film?

This has been split. I've had customers ranging from federal and state markets wanting a 3D kiosk in their lobby showing a promotional type piece to event video. 3D is just a way to get them to view their content and message. With the indy film market, the workflows get more costly and more complicated as each 3D scene needs to be carefully adjusted for the best 3D effect. Programs like CineForm's First Light is perfect for this as you can now have adjustable keyframes for convergence changes.

What camera/rig are most videographers using to shoot the video?

Most people appear to use a side-by-side rig. The rigs are basically the same with the main difference being the type of camera. These can range from two SDI cameras to AVCHD, to two Canon DSLRs. There are more and more rigs popping up with more cameras which can help when needing to adjust convergence without needing to auto zoom in the software. These side-by-side cameras all have the same issue of trying adjusting zoom, iris, and gain, which is really not possible without stopping the shot and making adjustments. Once you understand your limitations and adjust your shooting style, you can end up with some pretty amazing stereo video by using simple side-by-side rigs. The quality for the money is amazing. The cameras I mainly use are a pair of Canon HF10's which cost around $490 each.

Clearly Panasonic's AG-3DA1 camera has the advantage here as you can adjust any of these while you are shooting, including ocular spacing (convergence). At this point, this AVCCAM camera comes with a $21K price tag for that advantage. It's the closest thing I've seen to a point and shoot solution with all of the normal controls you would expect as a shooter. The broadcast customer seems to be the main market for this unit and CS5 and Cineform workflow fits this need perfect.

What are the most common target outputs? YouTube 3D? 3D DVD? Film?

This really depends on how you plan to show the video. Most 3D TVs will take Side By Side or Over Under stream in various formats like H264 or MPEG2. The TV can detect this mode and switch to 3D mode.You can also manually switch the 3D TV to 3D on the remote. For YouTube HD and Vimeo HD, anaglyph is the only answer. The key to anaglyph is to give your eyes about 10 seconds to adjust to the color shift and stand back at least 6 feet to help with the ghosting effects.

3D DVD and 3D Bluray require special encoders to encode 3D Stereo for these 3D hardware players to detect the 3D stream. There are not a lot of options for 3D authoring out there yet. You can create a 3D video exported to an AVCHD stream and burn it to BD disc and set it for auto play. The 3D TV will see the Side By Side stream but it requires you to put the 3D TV into 3D viewing by pressing the 3D button on the remote. This solution will work for most if you just put a note at the beginning of your video that tells the viewer to press the 3D button on their remote. Again, the real solution should be for the player to detect the 3D stream and switch automatically and this requires additional encoding on the authoring side.

The film market at this point is using 10 bit Cineon Log DPX files. The main size is still 1920x1080 but these DPX files can really be any size 2K or 4K . These files usually have separate Left and Right files. (Side By Side workflow is not a Film workflow as it's only half res)

What 3D viewing systems are editors using to edit at all ranges of the market, from James Cameron down to the prosumer?

For the pro/film market they use mainly projection systems like the Christie Dual HD SDI RealD units. There are a number of other projection systems around which use spinning discs and other methods of creating the 3D effect. I've seen a number of these in use in Hollywood and it's pretty impressive. Anaglyph in some parts of the film workflow has been used for years as well

For broadcast markets, they are mainly using side-by-side viewing either dual SDI or 3D TV. prosumers are just now getting started and many are using anaglyph but could benefit by buying a 3D TV like I mention in the video.Additionally, the new Nvidia 3D Vision display solution is well suited for all 3 markets for a clean editing and preview solution for small groups. The 3D Vision glasses allow you work at normal distances to your computer screen without eye strain. This is quickly becoming the preferred NLE solution. The Alienware 2310 is the current monitor of choice for editing and viewing. This is an amazing solution by Nvidia.

What's unique about CS5's 3D workflow?

The Adobe CS5 Premiere Pro 3D workflow with CineForm is unique because we can work with stereo files on a single timeline, add effects like 3D titles, dissolves, 32-bit color correction and so on in a single pass as well as preview in real time in a variety of ways. We support Nvidia 3D Vision Active Shutter Glasses (also know as Page Flip); passive monitors like the Hyundai; commercial 3D TVs with Active glasses like the ones from Samsung, Panasonic, and Sony. All from a standard DVI port using Open GL, AJA Kona 3 dual SDI, Nvidia dual SDI, and various anaglyph formats.

Previewing and exporting your timeline with effects is a huge advantage when working with 3D and needing to make constant adjustments on a quality display system. Most 3D solutions from other NLE require you to use anaglyph (need to stand back 6 or more feet and the color is off) or passive monitors (prices start at $1,450 and using $2,000 Dual Link SDI boards – most expensive solution) Adobe Premiere Pro allows you to use a $350-to-$450 120Hz computer display with a $200 Nvidia 3D Vision Kit for the highest possible quality for editing and viewing. With the addition of using consumer based 3D TVs (prices start at around $1,699 for a 40in. Samsung UN40C7000 (the main one I use) and $100 glasses makes this an affordable large LED solution.

With Adobe CS5 Premiere Pro, you can basically view 3D Stereo any way you need to including Dual SDI to a Christie projector.

Any movies or high-end productions built using the Premiere Pro/CineForm workflow?

There are several projects going on. One example is with second unit director and visual effects supervisor Rob Legato. He is currently evaluating CineForm Neo 3D with CS5 Premiere Pro as a VFX editorial tool for 3D review on Martin Scorsese's The Invention of Hugo Cabret.

What's your sense of how many videographers are now producing or trying to produce in 3D?

Good question. I don't have a good sense about the number of videographers yet, but I know its growing rapidly, and I also know that those experimenting with 3D span a wide range including film, broadcast (news, sports, game shows), independent film, documentaries, advertising, education, training, museums, terrain mapping, science (from NASA to under the ocean), and even home users. 3D weddings seems to be getting some traction also. (In fact one of our guys got married last summer and had his wedding shot in 3D). All these categories represent our existing customers.

What's their motivation?

Honestly I think people are experimenting right now. The experiment is, "Does 3D add to me storytelling experience?" I think from some demographics the answer will be "Yes," and for some it might not. If you can convey a story in a more compelling manner, whether its props, dialog, a pretty girl, or 3D, you will try different things.

What's the demographic? Are you seeing anything in the business market, or is it all indy film?

Ah, I guess I answered this above. Experiments with 3D training are underway.

What camera/rig are most videographers using to shoot the video?

Everything under the sun! From SI-3D (records directly to CineForm 3D files), Red, F35, Arri, Sony 150s, to Canon 5D/7D, and of course the Panasonic AG-3DA1 3D camcorder which is about to ship, and for which we are a launch partner. Rigs range from professional P&S Technik, Technika, 3D Film Factory, or custom rigs made by 3D production facilities like Pace or 3ality, to home-made rigs. The industry is ripe with innovation right now, and there is no "right" answer.

What are the most common target outputs? YouTube 3D? 3D DVD?

I haven't surveyed this, but I think YouTube is a big target. Also burning side-by-side onto conventional Blu-ray will be big because you don't need a BR-3D player and all new consumer TVs support side-by-side. Also, broadcast is using side-by-side because they can broadcast through their existing infrastructure. Blu-ray 3D will become more important, but we need Blu-Ray 3D players and HDMI 1.4 3D TVs to be available first.

What 3D viewing systems are editors using to edit at all ranges of the market, from James Cameron down to the prosumer?

For the most part people are viewing side-by-side as the most common format because it is the most widely support input format for 3D monitors. 3D gaming monitors can also used, and they are affordable, but it is less common to get from an editing environment out to a sequential frame monitor (although CineForm can do this because of collaboration we have with Nvidia).

How does the CineForm codec help?

CineForm is two things: At the lowest level we are a codec, but as you know that is the entry point for CineForm and is the most mature of our technologies. From a compression standpoint CineForm offers visual quality that is at parity or exceeds that of the respected HDCAM SR format used throughout post production. But SR is not an editing format; it is a tape recording format. CineForm becomes a file-based format used throughout post production for acquisition, editorial, mezzanine storage, archive, effects, and backend distribution. CineForm compression is compatible on Windows, Mac, and Linux, and with most all applications, such as from Adobe, Apple, Avid, Sony, and many more.

How are most editors using First Light?

The second part of our technology is about workflow, not compression. Yes, we leverage the extremely high quality of our compression and the speed. (We can encode or decode 200+ HD frames frames per second. Why is speed important? Well, the faster your compression works, the more CPU you have left for other post processing.)

This leads into what First Light is about. First Light introduces a new concept we call "Active Metadata." Active Metadata allows you to manipulate images non-destructively (think Lightroom for video, but with more advantages). First Light allows you to adjust normal image development parameters like exposure, contrast, saturation, lift, gamma, gain, all without flattening the changes into the underlying images. All these adjustments are applied after decoding the base image and before handing the image to the calling application (PPro, FCP, Media Composer, Vegas, etc). In addition to basic color processing First Light adjusts 3D parameters like convergence between images (including keyframing those adjustments), keystoning (toe-in/out of your images), disparity zoom (matching focal lengths on mismatched lenses), and many more professional features. (Here's a list of First Light Features.) You can also add 64x64x64 cubic film or artistic LUTs to material. So First Light is a non-destructive image development tool for your images).

I said "better" than Lightroom, right? What I mean by this is that metadata adjustments you do with First Light always remain with your CineForm images— you don't have to flatten the files when they are used in another application. So you can take CineForm files and Active Metadata through an entire workflow from acquisition, into editorial, and then into finishing without ever flattening the Active Metadata. This means that at any point in the workflow you can tweak the parameters using First Light. BTW, we have only scratched the surface on where First Light is going!

What's your sense of how many videographers are now producing or trying to produce in 3D?

Difficult to determine a number on this one, I guess suffice to say almost everyone we have talked to lately has plans to at least start exploring stereo 3D production. It's hard to estimate how much of that is real world use and how much is just curiosity.

What's their motivation?

More and more every day and the pro, prosumer, and hobbyist levels. Stereo 3D (S3D) is a tool for creative expression, another 'level' to be used to tell your story.

At a high level an artist uses S3D along with pacing, color or lighting to generate the correct emotional response in their audience. For example if you want the audience to feel warm and intimate with the scene you'd light in a warm golden colors, put neutral colors on set and create a very shallow stereo depth, in essence to draw you in. Now if you want to set the emotion of dark scary tunnel, darken the scene, create flickering point lights on rough grimy textures and set a very deep stereo depth.

Coraline 3D is one of the best movies I've seen take advantage of this. While it might not be out in 3D BluRay yet, there is a great article in Cinefex talking about this.

What's the demographic? Are you seeing anything in the business market, or is it all indy film?

All the major studios are embracing S3D as a motivator to get the public back in to the cinemas as well as counter the piracy issues. With S3D Digital Cinema Packages being fairly secure and encrypted, the pirates can`t just steal a movie by pointing a camera at a screen any more.
The Independent film makers have jumped at S3D because it gives their project a real value proposition and longevity to distribution which is essential to getting a film released and any hope of recouping costs. With the S3D filming techniques exploiting digital camera technology the cost of entry into production is lower than that for shooting 35mm film.

What camera/rig are most videographers using to shoot the video?

It varies on what they are trying to do and the budgets they have. There is no one "correct" stereo rig different shots, shooting styles and budgets call for different stereo rigs.

Technic, Arri and others are currently manufacturing rigs for filming but no single rig fits every shooting situation due to their size and lengthy calibration. Setup on an S3D rig can take up to 30 minutes or more so if the rig is going to be moved better make sure its factored in to the schedule.

What are the most common target outputs? YouTube 3D? 3D DVD? Film?

True film is the one piece that probably doesn't belong here. One of the key pieces that's making stereo movies comfortable to audieneces is removing film and going all digital. A properly shot piece can be re-purposed easily, but artistic choices do need to be made on the stereo cinematography depending on the size of the screen. A stereo shot that looks good on the big screen may not look right on a 50in. LCD or 24in. LCD so some amount of stereo touch up should be expected as you change output media.

Where does 3D Vision fit into the 3D workflow from a price/performance standpoint?

3D Vision fits on the production side as well as the home consumption side.

Production: 3D Vision is the only stereo solution that delivers a true 1080p stereo experience to desktop LCDs though the 3D Vision glasses and 3D Vision Ready LCDs. Other desktop alternatives rely on interlacing which removes texture detail. 3D Vision can be used on set as we showed at NAB with Silicon Imaging and during editing with Premiere Pro and CineForm. Many effects tools, like the Foundry's Nuke also support 3D Vision.

Home Consumption: The same features that make 3D Vision beneficial for the production side of things are also allowing it to grow the home 3D PC market as well. Every Nvidia 3D Vision PC out there enables their customers to not only enjoy games in a new way, but also view all the 3D content just as the creator had intended whether it comes from BluRay 3D or streaming over YouTube 3D or through fixed deployments like the NASCAR RaceBuddy 3D.