ISDCF Chairman’s Meeting Notes - April 2, 2013

Here are my cryptic meeting summaries. These are NOT the same as the once-in-a-while Chairmans notes - these are just my summary of the past meeting delivered soon after the meeting. Full meeting notes will follow. THIS IS A DRAFT AND MAY NOT BE ACCURATE. Notes are from memory, not from the transcription.

Next meetings schedule: Wednesday May 7 (HDR in afternoon), Wednesday Jun 25, Wednesday July 30, October 6/7 Plugfest, October 8 Main meeting (subject to change)

1. 1)SMPTE-DCP roadmap and status
   

ISDCF again discussed the SMPTE-DCP roadmap after new information came following the last plugfest. The refinement was with the audio configuration of MCA - we found that all systems can read MCA and will probably work if the channel layout was in interop order. The group decided that this was NOT the way to proceed - rather create test packages of true MCA - meaning the audio in a random order. So Mike Radford and Pierre Lemieux will build three SMPTE-DCP packages - A, B, and C. We anticipate that all deployed systems will be able to properly play back A & B and that C can be used for plugfest testing at this time. If this turns out to be correct we will use SMPTE-B-DCP for wide field testing.  There will be a 5.1 and a 7.1 version for testing.

The plan is to prepare the test material (Mike Radford/Pierre Lemieux), the test material will be initially tested in a few local theaters (Jerry Pierce/Dean Bullock) to confirm descriptions and suitability for unassisted in-theater testing. The packages will have a wide release via the trailer package vehicles and supported by NATO messaging. Deluxe has volunteered to help gather and confirm functionality.

Then release of trailers followed by full features.

It still remains to be seen if we should ask for SMPTE configuration 4 WTF to be redefined to be interop and define the channel layout to be consistent with ISDCF audio track layout configuration. That will be discussed at the next ISDCF meeting.

Additional SMPTE-DCP test material will be built for testing 3D subtitles. A draft diagram was shown and there seemed to be general acceptance of the usefulness of the proposed test material.

2. 2)Alternate Frame Rates / Subtitles
   

We reviewed the alternative frame rate chart that was generated at the last ISDCF meeting. This “new” chart is NOT a result of the plugfest, but our collective understanding of what faces distributors and those running festivals.

While we would like to have more “green” on the chart, it was generally acknowledged that there are going to be problems with some systems / some versions that will have difficulty playing the yellow or orange marked segments. The group wanted to error on the side of caution to let film makers / distributors know that if one of these frame rates is chosen that there may be problems in the field - distributor and festivals beware.

In particular there have been problems observed with subtitle performance at these alternative frame rates

3. 3)KDM delivery ecosystem
   

The following diagram was shared with the group:

The concept was to see if there was any agreement on the big-picture for gathering device lists, distributing KDMs, and interaction between various participants. Many interesting comments came up … Such as: My exhibition circuit does our own device list management - therefore you are also a DL resource.

Not sure what to do with this, but it may help formulate some solutions, especially in the FLMx design.

4) System Updates

There was a number of comments that field systems are not being updated to the latest version of software/firmware. This is causing problems for new movie releases, especially ones that use the later features. The current device list system is not up to date in keeping track of all the systems and their current state of version numbers.

A suggestion is that the FLMx include polling of the version number of the equipment and provide access to that information back to the equipment manufacturer. This could help understand the scope of the problem for getting version updates.

5) High Dynamic Range

We held a special meeting of a new sub-group of ISDCF on HDR. This is co-chaired by Jim Smith, Paramount and Jerry Pierce.

At the Inter-Society meeting we were given an assignment to help find a place to have the discussion. There is a desire to have as few new distribution masters in the future.

While there are other organizations doing some testing on HDR there doesn’t seem to be anywhere (at this time) that is approaching HDR from a high level. So for now ISDCF will keep holding meetings to discuss issues surrounding HDR.

For definition HDR means many things to many people. It may include brighter whites, blacker blacks, wider color gamut, it MIGHT include alternative frame rates, it also might include a defined brightness for 3D. It MAY be laser light related and MAY include traditional xenon light sources.

We may be looking for a new reference projector. But it may be too early to specify a reference projector yet.

At this stage there needs to be more demonstrations and discussions. One place will be ISDCF to have and coordinate the demonstrations/discussions.

Jerry

Chairman ISDCF

ISDCF Chairman’s Meeting Notes - February 26, 2013

At the meeting I was encouraged to create these cryptic meeting notes. These are NOT the same as the once-in-a-while Chairmans notes - these are just my summary of the past meeting delivered soon after the meeting. Full meeting notes will follow. THIS IS A DRAFT AND MAY NOT BE ACCURATE. Notes are from memory, not from the transcription.

ISDCF held a plugfest Monday/Tuesday February 24/25 at CineCert. Many issues were being tested and there were some revealing results. These notes cover more of the results than the actually blow-by-blow of the two days of testing.

It should be noted that the systems being tested may be using beta software and the test packages may not be properly created. The goal of our plug-fests are to allow equipment manufactures and content creators to test their systems in a “safe” way - allowing for mistakes here and not in the field.

SMPTE-DCP roadmap and status

SMPTE-DCP has had limited use in the wild - to date only HFR (Hobbit 2 in the US) and Dolby Atmos releases. While we believe that almost all deployed digital cinema systems can play a SMPTE-DCP there is some concern that all features will properly work - in particular audio routing and SMPTE subtitles. We thought we had a solution for general release of SMPTE-DCP with the use of MCA (Multi-channel audio) and constraints for authoring subtitles. During the test this appeared to not work, however a week later it was discovered there was a problem in the test file and the problematic player worked fine!

The good news is it gave the group an opportunity to create a roadmap for SMPTE-DCP deployment and authoring. Figure below.

We have called these SMPTE-A to SMPTE-D. We anticipate all deployed systems and all authoring systems can play and make SMPTE-A TODAY!! Further, we anticipate all deployed systems will be able to INGEST SMPTE-B today without problem - however many (most) will not respond to the additional features added by CPL Metadata (naming), TKR (theater key retrieval), and Markers (beginning of credits for in-theater automation).

We believe it is time to create a new test package of SMPTE-B in 5.1 and SMPTE-B in 7.1 for testing in-theater. This should give us a high degree of confidence for deployment of trailers and features in SMPTE-B DCP, although realistically most first releases will be in SMPTE-A since authoring tools are not common that add these new features.

The plan is to prepare the test material (Mike Radford/Pierre Lemieux), the test material will be initially tested in a few local theaters (Jerry Pierce/Dean Bullock) to confirm descriptions and suitability for unassisted in-theater testing. The packages will have a wide release via the trailer package vehicles and supported by NATO messaging. Deluxe has volunteered to help gather and confirm functionality.

Then release of trailers followed by full features. It would be hopeful for first release in October 2014, with full deployment goal in April 2015 - see it will be April! (We never knew what year.)

SMPTE-B adds new features that current production authoring systems. These important features should be a high priority for mastering systems for SMTPE-DCP, especially TKR.

Of special note is AUX data. Currently this is being used for Atmos / MDA and in the future may include motion seat data. Today these systems are very specific and are know to work with AUX data.

At the plugfest it appeared that getting to SMPTE-C would take significant time since one system rejected MCA (SMPTE Multichannel Audio), but that was found to be wrong. The test content had an authoring error. (The error was missing duration date either in Timecode component of Header metadata or Container duration of Essence descriptor.) While this is great news, the software tested was beta software and would take about a year to be deployed. We might be able to perform our field-test using SMPTE-C!

Both SMPTE-A and SMPTE-B use the Interop-DCP audio format as defined in ISDCF audio documents. It is SMTPE audio configuration 4 - Wild Track Format. A suggestion was made to submit to SMPTE a revision to the standard to re-name configuration 4 to be “Interop Audio” format and specifically call out the channel assignments defined in ISDCF. This suggestion should not delay the use of Configuration 4.

Why not use SMPTE Configuration 1 (5.1 audio) and SMPTE Configuration 5 (7.1 audio)? The original intent was to move to these well defined audio configurations. The first challenge was to get hardware platforms to route audio (mission accomplished!). Also, these two configurations would not work for Atmos and D-Box (and possibly others). The second challenge was to configure each screen with proper settings. The second challenge and the testing of the configurations was very problematic, so we turn to a more pragmatic approach.

Given how close we are to do real use of SMPTE–DCP the general feeling at the meeting was to stop "fixing" Interop–DCP and put all forward efforts in supporting and extending SMPTE–DCP.

One concern I had was the availability of offering tools that supported SMPTE–DCP, but but our test of six different tools had very positive results for creation of both Interop- and SMPTE–DCP packages.

Alternate Frame Rates / Subtitles / 3D subtitles

There has been renewed interest in using 3-D subtitles as well as some filmmaker interest in alternate film rates. We used the same content as last years plugfast (Thank you to Mike Radford, Fox and Steve Llamb, Deluxe). It is clear sufficient progress has been made, although software may take a year to be fully deployed.

3-D Subtitles

Note 1: Tests have only been conducted at 24 frames per second for 3-D rendered subtitles– this is not for HFR 3-D.

Note 2: 3-D subtitle thing will not work on TI series 1 projectors unless the server renders subtitles outside the projector.

All other cases(Series 2 dual link SDI, Series 2 IMB, Sony, server rendered) can work, Although software has not been fully deployed.

Bluntly, I was not expecting all systems to be working and I did not reported on this fact at the ISDCF meeting. I only recognized this after someone (Mike) pointed it out. We may be able to move faster on the use of 3-D subtitles in field.

We did not test on a 3-D screen at the plug fest, so polarity needs to be confirmed.

It may be time to create a test package for fields testing of 3-D sometimes - ideally at the same time as SMPTE–DCP testing.

Additional Frame Rate

We repeated the test with the same material provided by Mike Radford at Fox. This material tests frame rates from 24 to 60 frames per second in both 2-D and 3-D. Up to 3D60 or 120 frames per second. There are some mistakes in the authoring– 3D30 is at 500 per second (?) which is too high. It did point out that the specification may need to be clarification and fixing the max frame size in the JPEG2000 constraint.

There were 14 test clips for the full set. We expect no-playback with some frame rates that utilize dual link SDI connection between the server and the projector. The chart below shows the tests and the results from one vendor with series 1 (no one tested series 2 with dual link SDI) and the Sony test. (We could not confirm the polarity of the 3D subtitles.)

This was not a test of all systems and the usual caveat about the software has not been fully deployed, but the results were very encouraging.

If a studio/filmmaker wants to distribute any nonstandard frame rate, more testing will be needed.

Authoring Tests

We have heard about issues at film festivals that use many of the lower cost or free DCP authoring tools that result in playback issues on some servers. We wanted to do a simple test– Not definitive, but educational.

Chris Witham, Disney and Bill Hogan wrangled a short piece of content (reel 1 of Sintel) with 5.1 audio, closed captions and subtitles. These were submitted to 9 different authoring tools -

Wailua, Doremi CineAsset, Dolby, OpenCinema Tools, Easy DCP, Clipster, Doremi Rapid, Qube

Each tool created both and Interop-DCP and a SMPTE-DCP. For all the returned test material the total playback time for all the tests was about 45 minutes, so we didn’t do a complete playback test on all servers.

Bottom line: Most everything played! There were problems with color space (probably as assumption of 709 color vs RGB color in the package.) All tools built valid SMPTE-DCP and Interop-DCP packages!

This was a pleasant surprise. We really expected more problems given the “pain” that occurs at festivals… so why? Operator skill? Source material restrictions? Clear instructions?

So we know that most authoring systems are capable of making valid SMPTE & Interop-DCPs. We assume they allow the creation of bad DCPs. May high end authoring tools try not to allow an invalid DCP to be created. In this case the source material was well defined and in festivals cases the source material may be unconstrained.

In our case the “operator” was very knowledgable on the creating of DCPs and tested each one with high end tools - perhaps modifying the encode parameters to avoid problems.

In general, failure is a better educator than success and this time we had a great success.

Typography - JDCF

Our sister organization - Japan Digital Cinema Forum (JDCF) - contacted ISDCF on the issue of rendered subtitles. JDCF is similar to ISDCF operating on the “pain” approach to problems with pragmatic approach to solutions. The first pain point is the inconsistency of rendered Japanese subtitles.

To say the least this is a tricky problem. Yoshi presented a few slides (in the meeting folder) that described some of the issues. The figure below shows the complexity of rotating a few characters from horizontal to vertical.

In an effort to address this topic, JDCF’s members created 25 minutes of subtitles vs burned in on-screen text. We played these on a number of systems (although not all of them).

Bottom Line: The rendering engines were not all the same, but they were close to each other, but not matching the desired results.

The two sets of images on the next page show the two types of rendering engines not matching the desired results.

One member of the audience had an interesting observation, the company Adobe has worked on typography for years to try and perfect text placement. The rendering engine in digital cinema was not designed with the attention to detail needed for good typography.

One suggestion is to move to the PNG style of subtitles for these type of critical presentation of on-screen subtitles.

It’s not clear what the next step is for this work. Some needed to go and look at the details of the MXF files.

Closed Caption naming and performance

We’ve discovered that on at least one closed caption system, different data in the closed caption language element in the CPL causes the system to route that output to a different channel. In other words, if 5 trailers in a show playlist contain “en” in the language element and the feature contains “English”, the audience will not receive closed captions on their displays for the feature. It is strongly recommended that these elements shall be populated only with language tags specified by RFC 5646." In fact CPL metadata says "All instances of type xs:language in a Composition Playlist instance shall be language tags conforming to IETF RFC 5646."

Other

Selective Forensic

We also did testing of the selective forensic marking. This time we tested to see if a digital signal was delivered without modification through the system. This is NOT testing to see if the signal was properly marked, it would just show if a digital signal was touched. We used the D-Box digital signal as a test on channel 13.

Most systems successfully responded to the all on/all off KDM flags. (One system did not turn on the forensic mark - since it worked before, there was a bug to be discovered in this beta software). Only a few of the systems responded to the 1/2 off flag. But since we didn’t test the lower channels we can not be sure if they were marked - something to do for next time.

Naming Metadata / Markers

A few of the sample files we used populated the CPL Metadata fields and marker data. All played without issues. As expected, no system did anything with the data, but no problems in playback.

Jerry

Chairman ISDCF

ISDCF Chairman’s Meeting Notes - January 16, 2013

At the meeting I was encouraged to create these cryptic meeting notes. These are NOT the same as the once-in-a-while Chairmans notes - these are just my summary of the past meeting delivered soon after the meeting. Full meeting notes will follow. THIS IS A DRAFT AND MAY NOT BE ACCURATE. Notes are from memory, not from the transcription.

Audio Adventures

In the past month a movie was released that contained a short segment of a high pitched sound effect in the center channel. (7 seconds at 0db.) A number (over 100) high frequency speakers did not survive the sound effect and needed to be repaired/replaced. The auditorium needed to be off-line until the repair was completed.

The general consensus was that many auditoriums do not have the head-room in their sounds systems for sounds of this nature. Mixing rooms and QC rooms are designed for higher head-room so the danger was not detected in normal review.

In the DCP world there is a direct delivery of audio from the studio to the theater so mixers need to be aware that they can deliver full range sound that might not survive in the theater.

The suggestion is for the ISDCF chairman to write up a generic document to make mixers aware of the opportunity/problem. Also to suggest that audio hardware include some protection for the speaker systems.

Hobbit-2 HFR

In December the latest episode of “The Hobbit” was released. Not only did it do very well in the box office, it also is helping pave the way for SMPTE-DCP and High Frame Rate.

1,486 screens in the United States played the Hobbit in High Frame Rate. All of these were in SMPTE-DCP format (audio was in wild track format in the SMPTE package - sort of cheating IMHO). There were only minor issues and the introduction was considered a success. So many sites are capable of playing 3D48 using SMPTE-DCP packages.

Europe released in Interop-DCP, but most felt that there were probably no reason it could not have been released in SMPTE-DCP.

Some other Hobbit stats:

– 5 US image versions (2D24, 3D24 4.5ftL, 3D24 7ftL, 3D48 4.5ftL, 3D48 7ftL)

(unknown how many image versions worldwide)

– 260 CPL versions worldwide total (language and burned in subtitles)
– Separate SMPTE-DCP disc from Interop-DCP content

PlugFest February 2014

We decided to have the next plugfest on February 24/25 (Monday/Tuesday) at CineCert prior to the main ISDCF meeting on February 26. There will be Sony, DLP series 1, and DLP series 2 projectors and a number of servers - over 4 have said they were coming at this time. As we plan for the event many topics were discussed and these are mini-topics of their own:

SMPTE-DCP

We continue to help in the move to SMPTE-DCP from the current Interop-DCP. We have described at the last meeting the authoring constraints for audio and subtitles that we believe are necessary to facilitate the wide use of SMPTE-DCP in distribution. We have defined a set of test material that may help show that a system is ready and able to properly play SMPTE-DCPs. We hope these will be ready for the plug-fest.

4K PNGs

We will work with Japan Digital Cinema Forum (JDCF) in testing of new content - especially with regard to the issues with 2K content, 4K content, 2K PNGs, and 4K PNGs.

TKR

While there is interest in TKR - Theater Key Retrieval - there are some studios that feel it is insufficient. They feel there needs to be TDL - Trusted Device List - support to make a complete eco-system. Others want to move forward with this first part of the system.

At the moment TKR has not been implemented by any studio, although at least one studio is willing to provide content. Authoring tools have been updated to allow including of the URL in the CPL. We hope to include TKR in the test material for the plugfest.

Markers

We will again include content with markers (mainly for control of “lights up” at the beginning of the movie credits). Currently this is provided by paper or pdf delivery of information that must be manually added to the play-list.

Hard Drive mixed mode

We will have content in both SMPTE-DCP and Interop-DCP as part of the SMPTE-DCP test material. We will include both file types on the same hard drive using the protocol that has been suggested to SMPTE.

Selective Audio Forensic Marking

We wanted to again test the ability of creating KDMs that selectively enable the audio forensic marking for the audio channels. We are going to use a digital signal to test the marking (all on, all off, lower channels on). This fast test will not confirm the presence of a forensic mark, but it will confirm no impact on the audio channel. We will invite the forensic marking providers the opportunity to test their systems as well.

Aux Data Tracks

We will do some testing of ingesting of unknown Aux Data tracks (Atmos, MDA, D-Box, CPL Metadata, etc) Ingesting content in general with Asset map changes.

With all testing, the intended behavior is:

1)  Don’t Crash (aka Don’t Panic)

2)  Ingest anything but don’t play unknown assets

3)  Playable assets are defined in the CPL

Should there be playback constraints related to unknown assets? Any additional information in CPL be useful (now is a GREAT time due to upcoming meeting and FCD ballot resolution on CPL metadata document)?

Festival DCP problems

A number of software providers have entered into the marketplace to create DCPs. Due to their low cost, many content providers are using them especially in creating content for festivals. This has been problematic with a fair amount of “finger pointing” in trying to determine the problems in playback with different servers and software creating tools. We are going to invite tool makers to encode simple 2D24-2K-5.1 content for playback on a number of servers/systems. We hope this will help the problem in festivals. Bill Hogan has stepped up to wrangle the content and service providers.

Multiple KDM

We again recognized the need for a better understanding of how to deal with multiple KDMs in an auditorium - both for multiple projectors as well as immersive audio solutions. A subcommittee may work on this topic. (Dean (chair), Jim W. Kevin D, Nick M. Wade H, Mike R, Andre, JP, Anthony, Pierre, Andrew, Brian)

New CTP Subtitle Reel under development

We learned that there may be a new compliance test plan subtitle reel under development. No additional information was available.

HI Track adventures

We briefly talked about the issue with the HI track in the environment of pre-show, trailers, and the main feature. No real conclusion was reached on how to help with this issue.

Jerry

Chairman ISDCF

=====

ISDCF Chairman’s Meeting Notes - November 2013     Jerry Pierce

The last ISDCF meeting covered some great topics. I’m going to summarize the important topics in this short note - full meeting notes will follow. Just so we are all on the same page, THIS IS A DRAFT AND MAY NOT BE ACCURATE. Notes are from memory, not a transcription. Let me know if this format helps, maybe I’ll make it part of the monthly process...

SMPTE-DCP  / Interop-DCP

Aren’t we done yet?

We have shown in the lab that deployed hardware and special authoring constraints can be configured to play SMPTE-DCP -- the real world may be a different matter.

So we have a plan:

Constrain the subtitle files so they will play on all projectors (including Series 1 DLP).

In particular:

Use EntryPoint and Duration with knowledge of the timeline

Do not use an EntryPoint that you know will be beyond the current point on the server provided timeline.

EntryPoint for reel1 could not be larger than 00:00:00:00.  EntryPoint for reel2 could not be larger than Duration of reel1. EntryPoint for reel3 could not be larger than Duration of reel1 + Duration of reel2, etc.

Use audio channel labeling to create packages that resemble Interop layout, but confirm to the new SMPTE standard. This way a non-routing server will play right and a properly routing server will also play right.

Proposal: Layout channels in MXF track file according to Interop order and use labeling in 377-4 to identify channels

The general feeling (and hope) is that if a test package plays at all, it will properly play both Interop and SMPTE packages. So the follow up is to better define the constraints for mastering and create a new set of test DCP’s that confirm that it works.

2) HI Configuration

We are finding that getting the right HI track to hearing impaired patrons is not easy. There are 3 problems: (1) There are three sources of content for patrons - preshow, trailers, and features. The goal is to have the proper HI at all times - at least the best that can be expected. (2) Some HI systems require analog in (mainly IR systems) and others require digital AES in. (3) It is not completely clear what HI patrons want - the dialog only or a dialog enhanced track. A dialog only track would assume music and effects would leak through the (cheap) headsets. A dialog only track would be less appealing to those who steal the HI audio track.

Most (all?) external audio processors generate an (analog) HI track.

A suggested “magic HI” box could be used to always generate something for the HI track and have the track the best that is available. This concept is shown below.

More to discuss on this issue.

Drive formats for Interop/SMPTE DCP files

There has been a problem in creating USB sticks and hard drives with both SMPTE and Interop DCP packages. There has been confusion and not clear standard methods defined in SMPTE. An ISDCF subgroup met and proposed an approach that might work. The information is:

To be submitted to SMPTE from ISDCF working group

IF YOU WRITE OR READ DCPs — THIS IS REALLY IMPORTANT!!

Ext3 remains as file structure

Always show file extensions

All servers needs to look at assetmaps (assets not referenced by an assetmap should be ignored // mastering should not have assets that are not referred to by an assetmap)

Allow for hard links

Do not allow for sym links

THOU SHALL Not ../   (don’t allow a fragile link - stay within the tree)

Assetmap in root - also look in directory for additional assetmaps in directory - both are allowed

Assetmaps found only one directory down

Recommended that mastering should not have both assetmaps in root and subdirectories

Servers need to look for root & first-level directories for assetmaps

Ingest valid assets as much as possible (see an error, skip on to next asset)

If two assets with same UUID. If hashes match, no problem; if hashes don’t match, big problem

Think before ingesting … e.g. look at file size first, then hash

vol index in a given directory is associated with assetmap in same directory.

There should not be both SMPTE and Interop asset maps in same directory

What mods are needed for 429-9?

Multiple assetmaps allowed

Can’t leave the tree defined by assetmap

Assetmaps only one directory down

mastering shall not have both assetmaps in root and subdirectories

Ingest valid assets as much as possible

Fix URL scheme example. in 429-9 (do not use “file://“ in relative paths! The syntax of the path shall be “file”)

Remove clause referencing /.. in 8.1 of 429-9

vol index in a given directory is associated with assetmap in same directory. vol index is marginally optional.

Should examples be included in 429-9?

4K PNG subtitles

We revisited Interop-DCP performance with PNG subtitles. It seems we all can handle 2K images with 2K pngs, the issue is for a 4K image with 4K pngs. Especially a 4K image playback on a 2K projector with 4K pngs. The most reliable 4K image presentation has been with 2K pngs, but even that can cause problems.

(Is there any way we could make this more confusing? I’m looking for ideas -- how about 3K pngs? [joke])

One suggestion is to create two set of pngs for all 4K movies - one in 2K, the other in 4K and create 2 CPL’s for each - assuming we could figure out the naming convention to make this happen.

Naming Convention

Minor language changes were proposed for the naming convention as well as addition of “NR - Not Rated” to a all region category (for festivals), and to US & Germany.

6) TKR - Theater Key Retrieval

There has been essentially no progress on TKR - mainly due to lack of including the URL in the CPL.

Other topics were covered, but this is my list of the important stuff. I realize this is somewhat cryptic and not fully accurate - hopefully there will be discussion to make it better and more accurate.

Jerry

Chairman ISDCF

New Test Content for ISDCF-SMPTE/Interop DCP Testing

Audio Interop WTF channel identification - all - repeat HI / VI three times [HI/VI/HI/VI/HI/VI]

Audio SMPTE Layout channels in MXF track file according to Interop order and use labeling in 377-4 to identify channels - all channels - repeat HI / VI three times

Audio SMPTE layout 1 (5.1 with HI / VI) repeat HI / VI three times

Audio SMPTE layout 5 (7.1 with HI / VI) Repeat HI / VI / LRS / RRS three times

Subtitles Interop 1.5 minutes of subtitle timing with CC

Subtitles SMPTE 2.5 minutes of subtitle timing with CC using constrained time line (EntryPoint for reel2 could not be larger than Duration of reel1)