Category Archives: NAB 2016

What does Fraunhofer Digital Media Alliance do? A lot!

By Jonathan Abrams

While the vast majority of the companies with exhibit space at NAB are for-profit, there is one non-profit that stands out. With a history of providing ubiquitous technology to the masses since 1949, Fraunhofer focuses on applied research and developments that end up — at some point in the near future — as practical products or ready-for-market technology.

In terms of their revenue, one-third of their funding is for basic research, with the remaining two-thirds applied toward industry projects and coming directly from private companies. Their business model is focused on contract research and licensing of technologies. They have sold first prototypes and work with distributors, though Fraunhofer always keeps the rights to continue development.

What projects were they showcasing at NAB 2106 that have real-world applications in the near future? You may have heard about the Lytro camera. Fraunhofer Digital Media Alliance member Fraunhofer IIS has been taking a camera agnostic approach to their work with light-field technology. Their goal is to make this technology available for many different camera set-ups, and they were proving it with a demo of their multi-cam light-field plug-in for The Foundry’s Nuke. After capturing a light-field, users can perform framing correction and relighting, including changes to angles, depth and the creation of point clouds.

The Nuke plug-in (see our main image) allows the user to create virtual lighting (relighting) and interactive lighting. Light-field data also allows for depth estimation (called depth maps) and is useful for mattes and secondary color correction. Similar to Lytro, focus pulling can be performed with this light-field plug-in. Why Nuke? That is what their users requested. Even though Nuke is an OFX host, the Fraunhofer IIS light field plug-in only works within Nuke. As for using this light-field plug-in outside of Nuke, I was told that “porting to Mac should be an easy task.” Hopefully that is an accurate statement, though we will have to wait to find out.

DCP
Fraunhofer IIS has its hand in other parts of production and post as well. The last two steps of most projects are the creation of deliverables and their delivery. If you need to create and deliver a DCP (Digital Cinema Package), then easyDCP may be for you.easydcp1

This project began in 2008, when creating a DCP was not as familiar as it is today to most users, and a deep expertise of the specifications for correctly making a DCP was very complex. Small- to medium-sized post companies, in particular, profit from the easy-to-use easyDCP suite. The engineers of Fraunhofer IIS were also working on behalf of the DCI specifications for Digital Cinema, therefore they are experienced in integrating all important features in this software for DCPs.

The demo I saw indicated that the JPEG2000 encode was as fast as 108fps! In 2013, Fraunhofer partnered with both Blackmagic and Quantel to make this software available to the users of those respective finishing suites. The demo I saw was using a Final Cut Pro X project file and it was with the Creator+ version since it had support for encryption. Avid Media Composer users will have to export their sequence and import it into Resolve to use easyDCP Creator. Amazingly, this software works as far back as Mac OS X Leopard. IMF creation and playback can also be done with the easyDCP software suite.

VR/360
VR and 360-degree video were prominent at NAB, and the institutes of the Fraunhofer Digital Media Alliance are involved in this as well, having worked on live streaming and surround sound as part of a project with the Berlin Symphony Orchestra.

Fraunhofer had a VR demo pod at the ATSC 3.0 Consumer Experience (in South Hall Upper) — I tried it and the sound did track with my head movement. Speaking of ATSC 3.0, it calls for an immersive audio codec. Each country or geographic region that adopts ATSC 3.0 can choose to implement either Dolby AC-4 or MPEG-H, the latter of which is the result of research and development by Fraunhofer, Technicolor and Qualcomm. South Korea announced earlier this year that they will begin ATSC 3.0 (UHDTV) broadcasting in February 2017 using the MPEG-H audio codec.

From what you see to what you hear, from post to delivery, the Fraunhofer Digital Media Alliance has been involved in the process.

Jonathan S. Abrams is the Chief Technical Engineer at Nutmeg, a creative marketing, production and post resource.

UHD Alliance’s Victor Matsuda: updates from NAB 2016

Victor Matsuda from the UHD Alliance was at NAB 2016. The Alliance was formed about 15 months ago as 4K UHD products began exploding into the market. The goal of the Alliance was to establish certifications for these new products and for content. All of this is to ensure a quality experience for consumers, who will ultimately drive 4K/UHD adoption throughout the market.

Watch our video with Matsuda to find out more.

G-Tech 6-15

NAB 2016 from an EP’s perspective

By Tara Holmes

Almost two weeks ago, I found myself at NAB for the first time. I am the executive producer of color and finishing at Nice Shoes, a post production studio in New York City. I am not an engineer and I am not an artist, so why would an EP go to NAB? I went because one of my main goals for 2016 is to make sure the studio remains at the forefront of technology. While I feel that our engineering team and artists represent us well in that respect, I wanted to make sure that I, along with our producers, were fully educated on these emerging technologies.

One of our first priorities for NAB was to meet with top monitor manufacturers to hopefully land on what UHD HDR monitors we would find to meet our standards for professional client viewing. We came to the conclusion that the industry is not there yet and we have more research to do before we upgrade our studio viewing environments.

Everyone with me was in agreement. They aren’t where they need to be. Most are only outputting around 400-800 nits and are experiencing luminance and contrast issues. None of this should stop the process of coloring for HDR. For the master monitor for the colorist, the Sony BVM-X300 OLED master monitor, which we are currently using, seems to be the ideal choice as you can still work in traditional Rec 709 as well as Rec 2020 for HDR.

After checking out some monitors, we headed to the FilmLight booth to go over the 5.0 upgrades to Baselight. Our colorist Ron Sudul, along with Nice Shoes Creative Studio VFX supervisor Adrian Winter, sat with myself and the FilmLight reps to discuss the upgrades, which included incredible new isolation tracking capabilities.  These upgrades will reinvent what can be achieved in the color suite: from realtime comps to retouch being done in color. The possibilities are exciting.

I also spent time learning about the upgrades to Filmlight’s Flip, which is their on-set color hardware. The Flip can allow you to develop your color look on set, apply it during your edit process (with the Baselight plug-in for Avid) and refine it in final color, all without affecting your RAW files. In addition to the Flip, they developed a software that supports on-set look development and grading called Prelight. I asked if these new technologies could enable us to even do high-end things like sky replacements on set and was told that the hardware within the Flip very well could.

We also visited our friends at DFT, the manufacturers of the Scanity film scanner, to catch up and discuss the business of archiving. With Scanity, Nice Shoes can scan 4K when other scanners only scan up to 2K resolution. This is a vital tool in not only preserving past materials, but in future proofing for emerging formats when archiving scans from film.

VR
On Sunday evening before the exhibits opened, we attended a panel on VR that was hosted by the Foundry. At this event we got to experience a few of the most talked about VR projects including Defrost, one of the first narrative VR films, from the director of Grease, Randal Kleiser, who was on the panel along with moderator Morris May (CEO/founder, Specular Theory), Bryn Mooser (co-founder, RYOT), Tim Dillon (executive producer, MPC) and Jake Black (head of VR, Create Advertising).

The Foundry’s VR panel.

The panel inspired me to delve deeper into the VR world, and on Wednesday I spent most of my last day exploring the Virtual & Augmented Reality Pavilion. In addition to seeing the newest VR camera rig offerings and experiencing a live VR feed, as well as demo-ing the Samsung Gear, I explored viewing options for the color workflow. Some people I spoke to mentioned that multiple Oculus set-ups all attached to a single feed was the way to go for color workflow, but another option that we did a very preliminary exploration of was the “dome” possibility, which offers a focused 180-degree view for everyone involved to comment on the same section of a VR scene. This would enable all involved to be sure they are experiencing and viewing the same thing at the same time.

HDR Workflow
Another panel we attended was about HDR workflows. Nice Shoes has already had the opportunity to work on HDR material and have begun to develop workflows for this emerging medium. Most HDR deliverables are for episodic and long form for such companies as Netflix, Hulu and the like. It may be some time before commercial clients are requesting an HDR deliverable, but the workflows will be much the same so the development being performed now is extremely valuable.

My biggest take away was that there are still no set standards. There’s Dolby Vision vs. HDR 10 vs. PQ vs. others. But it appears that everyone agrees that standards are not needed right now. We need to get tools into the hands of the artists and figure out what works best. Standards will come out of that. The good news is that we appear to be future-proofed for the standard to change. Meaning for the most part, every camera we are shooting on is shooting for HDR and should standards change — say from 1000 nits to 10,000 nits — the footage and process is still there to go back in and color for the new request.

Summing Up
I truly believe my time spent at NAB has prepared me for the myriad of questions that will be put forth throughout the year and will help us develop our workflows to evolve the creative process of post. I’ll be sure to be there again next year in order to prepare myself for the questions of 2017 and beyond.

Our Main Image: The view walking into the South Hall Lower at the LVCC.


Talking storage with LaCie at NAB

By Isaac Spedding

As I power-walked my way through the NAB show floor, carefully avoiding eye contact with hopeful booth minders, my mind was trying to come up with fancy questions to ask the team at LaCie that would cement my knowledge of storage solutions and justify my press badge. After drawing a blank, I decided to just ask what I had always wanted to know about storage companies in general: How reliable are your drives and how do you prove it? Why is there a blue bubble on your enclosures? Why are drives still so damn heavy?

Fortunately, I met with two members of the LaCie team, who kindly answered my tough questions with valuable information and great stories. I should note that just prior to this NAB trip I had submitted an RMA for 10 ADATA USB.3.0 drives, as all the connectors on them had become loose and fallen out or into the single-piece enclosure. So, as you can imagine, at that moment in time, I was not exactly the biggest fan of hard drive companies in general.

“We are never going to tell you (a drive) will never fail,” said Clement Barberis, marketing manager for LaCie. “We tell people to keep multiple copies. It doesn’t matter how, just copies. It’s not about losing your drive it’s about losing your data.”

LaCie offers a three-to five-year warranty on all its products and has several services available, including fast replacement and data recovery. Connectors and drives are the two main points of failure for any portable drive product.

two shot

LaCie’s Clement Barberis and Kristin Macrostie.

Owned by Seagate, LaCie has a very close connection with that team and can select drives based on what the product needs. Design, development and target-user all have an impact on drive and connection selection. Importantly, LaCie decides on the connection options not by what is the newest but by what works best with the internal drive speed.

Their brand new 12-bay enclosure, the LaCie 12big Thunderbolt 3 (our main image), captures the speed of Thunderbolt 3, and with a 96TB capacity (around 100 hours of uncompressed 4K), the system can transfer around 2600 MB/s (yes, not bits). It is targeted at small production houses shooting high-resolution material.

Why So Heavy?
After Barberis showed me the new LaCie 12big, I asked why the form factor and weight had not been redesigned after all these years. I mean, 96TB is great and all but it’s not light — at 17.6kg (38.9 pounds) it’s not easy to take on the plane. Currently, the largest single drive available is 8TB and features six platters inside the traditional form factor. Each additional platter increases the weight of each drive (and its capacity), but the weight increase means that a smaller form factor for a drive array is possible. That’s why drive arrays have been staying the same size and gaining weight and storage capacity. So your sleek drive will be getting heavier.

LaCie produces several ranges of hard drives with different designs. It’s most visually noticeable in LaCie’s Rugged drive series, which features bright orange bumpers. Other products feature a “Porsche-like” design and feature the blue LaCie bubble. If you are like me, you might be curious how this look came about.

rugged

According to Kristin MacRostie, PR manager for LaCie, “The company founder, Philippe Spruch, wasn’t happy with the design of the products LaCie was putting out 25 years ago — in his words, they were ‘geeky and industrial.’ So, Spruch took a hard drive and a sticky note and he wrote, ‘Our hard drives look like shit, please help,’ and messengered it over to (designer) Philippe Starck’s office in Paris. Starck called Spruch right away.”

The sleek design started with Philippe Starck and then Neil Poulton, who was an apprentice to Starck, and who was brought on to design the drives we see today. The drive designs target the intended consumers, with the “Porsche design” aligning itself to Apple users.

Hearing the story behind LaCie’s design choice, the recommendation to keep multiple drives and not rely on just one, and the explanation of why each product is designed, convinced me that LaCie is producing drive solutions that are built for reliability and usability. Although not the cheapest option on the market today, the LaCie solutions justify this with solid design and logic behind the decision of components, connectors and cost. Besides, at the end of the day, your data is the most important thing and you shouldn’t be keeping it on the cheapest possible drive you found at Best Buy.

Isaac Spedding is a New Zealand-based creative technical director, camera operator and editor. You can follow him on Twitter @Isaacspedding.


Fusion launches at NAB 2016 with 4K OLED display

Industry veterans Carl J. Dempsey and Steve Farmer introduced their new company, Fusion, at NAB 2016. They also unveiled their first product — a 55-inch OLED 4K reference display system called the ORD-55.

The ORD-55 features independent-processing quad-mode operation (IPQ), where four individual processors provide independent control of all channels, and uses a single-link 12G input. In quad mode, it provides four independent 27.5-inch FHD displays. The display can also be configured to show one large 4K picture with three smaller preview panes in FHD. The system features a deep “Black Level,” super-wide viewing angle of 178 degrees, a 10 microseconds response time, 100,000:1 contrast ratio, ultra-wide color gamut with 1.07 billion colors, and 12-bit color processing.

Dempsey and Farmer bring a ton of experience to Fusion. Dempsey, a 25-year industry vet, was most recently president and CEO of Wohler Technologies. Farmer brings 22 years in the broadcast industry to the new company, including a stint as director of engineering at Wohler. Starting as a design engineer, he then took on senior management roles in both engineering and product management.

In their new venture, Dempsey will serve as Fusion’s CEO, while Farmer will hold the CTO post.


Digging Deeper: Dolby Vision at NAB 2016

By Jonathan Abrams

Dolby, founded over 50 years ago as an audio company, is elevating the experience of watching movies and TV content through new technologies in audio and video, the latter of which is a relatively new area for their offerings. This is being done with Dolby AC-4 and Dolby Atmos for audio, and Dolby Vision for video. You can read about Dolby AC-4 and Dolby Atmos here. In this post, the focus will be on Dolby Vision.

First, let’s consider quantization. All digital video signals are encoded as bits. When digitizing analog video, the analog-to-digital conversion process uses a quantizer. The quantizer determines which bits are active or on (value = 1) and which bits are inactive or off (value = 0). As the bit depth for representing a finite range increases, the greater the detail for each possible value, which directly reduces the quantization error. The number of possible values is 2^X, where X is the number of bits available. A 10-bit signal has four times the number of possible encoded values than an 8-bit signal. This difference in bit depth does not equate to dynamic range. It is the same range of values with a degree of quantization accuracy that increases as the number of bits used increases.

Now, why is quantization relevant to Dolby Vision? In 2008, Dolby began work on a system specifically for this application that has been standardized as SMPTE ST-2084, which is SMPTE’s standard for an electro-optical transfer function (EOTF) and a perceptual quantizer (PQ). This work is based on work in the early 1990s by Peter G. J. Barten for medical imaging applications. The resulting PQ process allows for video to be encoded and displayed with a 10,000-nit range of brightness using 12 bits instead of 14. This is possible because Dolby Vision exploits a human visual characteristic where our eyes are less sensitive to changes in highlights than they are to changes in shadows.

Previous display systems, referred to as SDR or Standard Dynamic Range, are usually 8 bits. Even at 10 bits, SD and HD video is specified to be displayed at a maximum output of 100 nits using a gamma curve. Dolby Vision has a nit range that is 100 times greater than what we have been typically seeing from a video display.

This brings us to the issue of backwards compatibility. What will be seen by those with SDR displays when they receive a Dolby Vision signal? Dolby is working on a system that will allow broadcasters to derive an SDR signal in their plant prior to transmission. At my NAB demo, there was a Grass Valley camera whose output image was shown on three displays. One display was PQ (Dolby Vision), the second display was SDR, and the third display was software-derived SDR from PQ. There was a perceptible improvement for the software-derived SDR image when compared to the SDR image. As for the HDR, I could definitely see details in the darker regions on their HDR display that were just dark areas on the SDR display. This software for deriving an SDR signal from PQ will eventually also make its way into some set-top boxes (STBs).

This backwards-compatible system works on the concept of layers. The base layer is SDR (based on Rec. 709), and the enhancement layer is HDR (Dolby Vision). This layered approach uses incrementally more bandwidth when compared to a signal that contains only SDR video.  For on-demand services, this dual-layer concept reduces the amount of storage required on cloud servers. Dolby Vision also offers a non-backwards compatible profile using a single-layer approach. In-band signaling over the HDMI connection between a display and the video source will be used to identify whether or not the TV you are using is capable of SDR, HDR10 or Dolby Vision.

Broadcasting live events using Dolby Vision is currently a challenge for reasons beyond HDTV not being able to support the different signal. The challenge is due to some issues with adapting the Dolby Vision process for live broadcasting. Dolby is working on these issues, but Dolby is not proposing a new system for Dolby Vision at live events. Some signal paths will be replaced, though the infrastructure, or physical layer, will remain the same.

At my NAB demo, I saw a Dolby Vision clip of Mad Max: Fury Road on a Vizio R65 series display. The red and orange colors were unlike anything I have seen on an SDR display.

Nearly a decade of R&D at Dolby has been put into Dolby Vision. While Dolby Vision has some competition in the HDR war from Technicolor and Philips (Prime) and BBC and NHK (Hybrid Log Gamma or HLG), it does have an advantage in that there have been several TV models available from both LG and Vizio that are Dolby Vision compatible. If their continued investment in R&D for solving the issues related to live broadcast results in a solution that broadcasters can successfully implement, it may become the de-facto standard for HDR video production.

Jonathan S. Abrams is the Chief Technical Engineer at Nutmeg, a creative marketing, production and post resource.


NAB 2016: VR/AR/MR and light field technology impressed

By Greg Ciaccio

The NAB 2016 schedule included its usual share of evolutionary developments, which are truly exciting (HDR, cloud hosting/rendering, etc.). One, however, was a game changer with reach far beyond media and entertainment.

This year’s NAB floor plan featured a Virtual Reality Pavilion in the North Hall. In addition, the ETC (USC’s Entertainment Technology Center) held a Virtual Reality Summit that featured many great panel discussions and opened quite a few minds. At least that’s what I gathered by the standing room only crowds that filled the suite. The ETC’s Ken Williams and Erik Weaver, among others, should be credited for delivering quite a program. While VR itself is not a new development, the availability of relatively inexpensive viewers (with Google Cardboard the most accessible) will put VR in the hands of practically everyone.

Programs included discussions on where VR/AR (Augmented Reality) and now MR (Mixed Reality) are heading, business cases and, not to be forgotten, audio. Keep in mind that with headset VR experiences, multi-channel directional sound must be perceivable with just our two ears.

The panels included experts in the field, including Dolby, DTS, Nokia, NextVR, Fox and CNN. In fact, Juan Santillian from Vantage.tv mentioned that Coachella is streaming live in VR. Often, concerts and other live events have a fixed audience size, and many can’t attend due to financial or sell-out situations. VR can allow a much more intimate and immersive experience than being almost anywhere but onstage.

One example, from Fox Sports’ Michael Davies, involved two friends in different cities virtually attending a football game in a third city. They sat next to each other and chatted during the game, with their audio correctly mapped to their seats. There are no limits to applications for VR/AR/MR, and, by all accounts, once you experience it, there is no doubt that this tech is here to stay.

I’ve heard many times this year that mobile will be the monetary driver for wide adoption of VR. Halsey Minor with Voxelus estimates that 85 percent of VR usage will be via a mobile device. Given that more photos and videos are shot on our phones (by far) than on dedicated cameras, this is not surprising. Some of the latest crop of mobile phones are not only fast and contain high dynamic range and wide color gamut, they feature high-end audio processing from Dolby and others. Plus, our reliance on our mobiles ensures that you’ll never forget to bring it with you.

Light Field Imaging
On both Sunday and Tuesday of NAB 2016, programs were devoted to light field imaging. I was already familiar with this truly revolutionary tech, and learned about Lytro, Inc. a few years ago from Internet ads for an early consumer camera. I was intrigued with the idea of controlling focus after shooting. I visited www.lytro.com and was impressed, but the resolution was low, so, for me, this was mainly a proof of concept. Fast forward three years, and Lytro now has a cinema camera!

Jon Karafin (pictured right), Lytro’s head of Light Field Imaging, not only unveiled the camera onstage, but debuted their short Life, produced in association with The Virtual Reality Company (VRC). Life takes us through a man’s life and is told with no dialog, letting us take in the moving images without distraction. Jon then took us through all the picture aspects using Nuke plug-ins, and minds started blowing. The short is directed by Academy Award-winner Robert Stromberg, and shot by veteran cinematographer David Stump, who is chief imaging scientist at VRC.

Many of us are familiar with camera raw capture and know that ISO, color temperature and other picture aspects can be changed post-shooting. This has proven to be very valuable. However, things like focus, f-stop, shutter angle and many other parameters can now be changed, thanks to light field technology — think of it as an X-ray compared to an MRI. In the interests of trying to keep a complicated technology relatively simple, sensors in the camera capture light fields in not only in X and Y space, but two more “angular” directions, forming what Lytro calls 4D space. The result is accurate depth mapping which opens up so many options for filmmakers.

Lytro_Cinema_2

Lytro Cinema Camera

For those who may think that this opens up too many options in post, all parameters can be locked so only those who are granted access can make edits. Some of the parameters that can be changed in post include: Focus, F-Stop, Depth of Field, Shutter Speed, Camera Position, Shutter Angle, Shutter Blade Count, Aperture Aspect Ratio and Fine Control of Depth (for mattes/comps).

Yes, this camera generates a lot of data. The good news is that you can make changes anywhere with an Internet connection, thanks to proxy mode in Nuke and processing rendered in the cloud. Jon demoed this, and images were quickly processed using Google’s cloud.

The camera itself is very large, but Lytro knows that they’ll need to reduce the size (from around seven feet long) to a more maneuverable form factor. However, this is a huge step in proving that a light field cinema camera and a powerful, manageable workflow is not only possible, but will no doubt prove valuable to filmmakers wanting the power and control offered by light field cinematography.

Greg Ciaccio is a technologist focused primarily on finding new technology and workflow solutions for Motion Picture and Television clients. Ciaccio served in technical management roles for the respective Creative Services divisions for both Deluxe and Technicolor.