AMA Linked Media Has Different Timecode

[Michael Phillips]

So I went an un-archived some 2.5K DNG files and downloaded some .ari 2K files as well just to have two different sources. I made OPAtom1a files with DNxHR HQX codec and used the MXF AMA plug-in to link to those files. I still have no problem with timecode.

In your last post you say MXF OPAtom and not OPAtom1a. Which ones are they exactly to be sure to replicate the same test. But I would go back to your support person at Avid and let them know that MXF from Resolve works and need to look elsewhere in the workflow as to what might be the issue. No cross frame rate stuff here is there?

As far as Scott’s findings, I don’t know what the exact process was in his workflow to be sure. I have been making both types of MXF files in Resolve 11 and now 12, and worked with them in the past several versions of Media Composer (DNxHD) and DNxHR with v8.x.

Michael

[Michael Phillips]

One of the primary problems I see in the above screenshots is that TAPE is listed at mmob etc. that means that in Resolve the REEL was never properly set and was left blank. When it comes into Avid, it randomly numbers them mmob.1. 2 etc. The proper file name is only in the Name column which can be fragile if not duplicated elsewhere for backup. But this does not mean other issues might not crop up – first step it to properly set REEL in Resolve before making native MXF dailies.

Michael

[Tim Wilson]

Greetings all,

Hopping in with a brief explanation, and a couple of reminders.

1) For our first 19 years, we didn’t allow members to delete their posts. When we opened up that option as an experiment, chaos quickly ensued. Even when members removed their posts for the purest of reasons, threads became difficult to follow, good stuff went missing, etc.

This thread going off the rails in part because of members removing their posts is an example of why we returned to our previous policy of, if you want a post removed, contact us using the Contact link in the upper right of the page, and we’ll consider it, but generally, no. Accountability depends on the knowledge that if you’re going to say something in Creative COW, you’re going to have to stand by it.

——- This is ENTIRELY in keeping with policies of forums across the web, in the widest possible range of areas of interest. Threaded conversations depend on structure. Our brief experiment with users having unlimited access to deleting their post was a deviation from standard web practice that we have amply demonstrated is simply not at all sustainable.

2) We have always had a policy against posts about other posters. This was originally created to prevent personal attacks, but also more generally, to just keep threads on topic, and to keep from going down rabbit holes about motivations, intent, and other non-answers to important questions that bring people here in the first place.

3) So, while acknowledging the unpleasant irony of deleting posts about deleted posts, and acknowledging that there are no satisfactory outcomes to the “what the h— happened here?” detour that the thread took, I have pared the thread back to the topic.

4) Any ON TOPIC additions to this thread are of course welcome.

5) In keeping with other longstanding policies at Creative COW, arguments about administrative actions are not. This post is thus read-only. If you have issues with my handling of this or any other issues, please take them up with me directly, using the Contact link at the top of this page. I’m very much up for phone/Skype calls as well as emails, and will happily reach out to you on my dime.

Thank you for your understanding, and as always, thank you for continued support for Creative COW. It pains me to fall short of your expectations and our own, and I welcome any direct conversations with any of you about how we can do better.

Best regards,

Tim Wilson
Editor-in-Chief
Creative COW

[Scott Freeman]

Michael Thibault,

An automated tapeless reconform should have a strategy in-place for the reconform. Several strategies are available for the DaVinci Resolve / Media Composer User.

For this reconform you presented to me, it looks like the strategy for the reconform is a sequence of master clips using Source File column metadata as the metadata you would like to use for the automatic reconform in DaVinci Resolve from an avid sequence AAF? Maybe, perchance, this is why the AMA plug-in path was used here in the first place? In DaVinci Resolve to initiate using the Avid Master Clip Source File column metadata, the DaVinci Resolve user when importing the Avid AAF places a checkmark next to the words “Link to source camera files” .

Avid AMA Plug-in Ingest reconform back in Resolve problem seen here:
one issue I have noticed with DaVinci Resolve, if the user renders out MXF OP-Atom files and the Avid Media Composer user uses the AMA MSP_MXF Plug-in as the ingest method, I have seen the Avid assign wrong timecode values upon ingest for those master clips.
If the Avid user ingests those same MXF OP-Atom media files by importing a media database into the bin, then the timecode is correct for those master clips.

Below are steps for a possible ingest reconform in Resolve solution:
Because you already created a sequence with master clips that have the wrong timecode values, here is my attempt to update the metadata for the master clips linked to the sequence for Resolve using Media Composer as the metadata manager.

Here are the steps I am taking to automatically perform a reconform in Resolve with correct timecode values and keeping the Source File column metadata intact for use as the tapeless reconform strategy in Resolve.

Thank you so much for presenting this ingest path issue. From what I can tell to create an automatic reconform in Resolve from an Avid sequence for this scenario, this is my best attempt to update the metadata for an automatic reconform using Source File Column metadata with correct timecode values that match your original tapeless master files.

I looked at the bins you sent me and these are the steps I did to update the bin metadata for the master clips linked to the sequence, if the strategy for tapeless reconform is to use Source File column metadata with accurate timecode in DaVinci Resolve.

1) Open the Source File Mobs bin (AMA Plug-in Ingest bin), duplicate Source File column metadata to a new Source File7 custom column. The custom column Source File7 for this procedure I am populated to rectify that metadata and then in Avid is used as an auto-merge to the correct timecode master clips.
2) In the Project Window, Click on the “Info” Tab, and Click on the “Clear Bin Memory” button. If the MXF OP-Atom media is linked to the master clips, now go to the located storage device > Avid MediaFiles Folder > MXF Folder > Numbered Folder > clear the media databases.
3) Open up the Tape Mobs bin (msmMMOB.mdb Ingest bin), and make it the Start Timecode Master Clip Metadata Leader Bin by duplicating the master clips start timecode and populating the Auxiliary TC2 column. This is amazing metadata control where avid can now use this master clip timecode metadata and ripple it across the project.
4) Keep the Start Timecode Master Clip Metadata Leader Bin Open, and then open up the Source File Mobs bin to automatically update. The Avid automatically merges the leader and follower metadata giving the user some metadata merging control.
5) In the Start Timecode Master Clip Metadata Leader Bin, clear all headings except for Name and Source File7 columns, then select all, and then export an ALE of that bin.
6) Open up the ALE in a text editor and:
change the column heading “Tape” to “link”
change the column heading “Source File” to “source file9”
change the column heading “Source File7” to “Source File”
I noticed that “_A01” is in the Source File column yet the Name column omitted the “_A01” so if the name metadata is the one for you, then I would rename the “Name” column to “Source File” and “Source File7” to “Name”
7) Save the ALE
8) Create a new bin and import the ALE
9) In the Start Timecode Master Clip Metadata Leader Bin duplicate the “Tape” column metadata into the “link” column.
10) In the Project Window > Info tab, click the clear memory button.
11) Select the sequence and the ALE master clips at the same time and then select the relink command
with the following selections:

here are before and after snap shots of the bins you provided me:

before:

wrong start timecode source file mobs bin with source file column metadata – this is the AMA plug-in bin

missing source file metadata with correct timecode – this is the imported media database bin and is the tape mobs bin. to note because in Resolve there was not a Reel Name assigned at the time of render, the Avid went into a mode where tape gets assigned the name of the file used for the ingest with appended dot number as the tape name making each of these master clip unique. I have noticed if the user would of used the Avid Media Tool for ingest into the bin, the avid would of had blank tape names.

after metadata procedure of steps listed above

updated corrected metadata bin and timeline

After this metadata management for this example, the Source File column populated metadata is also a metadata match to the Source Icon Name Metadata and thus default Reel Name Metadata for Events in the Resolve Edit Index when Avid creates the AAF from the sequence. Now there are two automatic reconform strategies in place for the DaVinci Resolve user to reconform automatically the Avid Sequence AAF with now corrected timecode values and Souce File column metadata.

to repeat:

I have seen this issue before: I attempt to test out codec wrapper combos and all ingest paths into avid that I can think of. I noticed that if creating MXF OP-Atom files with Resolve and the Avid Media Composer user ingests those MXF OP-Atom files with the MSP_MXF Plug-in the Avid does seem to pull those files into the bin and I have noticed a timecode discrepancy.

tested with Resolve v12.0.1 and Media Composer v8.4.1 (Mac OS 10.10.5)
I will keep testing and retesting. And I recommend that users always test out the ingest path to make sure you are getting the desired results before diving in.

From memory: I noticed that if the fps of the mxf op-atom file is different from the project fps and ingested into that project then I was able to recreate the timecode problem. If in a 59.94i project and the mxf op-atom file was also 59.94i then I was able to recreate the timecode problem.

As a workflow note: I currently would not recommend importing Resolve created MXF OP-Atom files using the MXF AMA plug-in. I also currently do not recommend importing an AAF from Resolve into Avid, because I have seen the Avid lock the user out of metadata modification commands. For your project it looked like for at least 10 of the files, you imported an AAF from Resolve into Media Composer and for those files I had to move those into their own bin and manually paste the timecode into Auxiliary TC2. The rest of the master clips I was able to select the start timecode heading and use the duplicate command to populate the Auxiliary TC2 column.

Also to note, when using Resolve to render individual MXF OP-Atom files for Media Composer, before rendering, please enable Reel Names for display in the Blackmagic Design DaVinci Resolve Media Pool. This Reel Name Extraction opens up more reconform strategies for downstream metadata use.

I hope this is helpful

[Michael Phillips]

Thanks Scott for the detailed response. I will just add two a few points of precision. There is a difference between MXF OPAtom and MXF OPAtom 1a.

MXF OPAtom exports one file for every essence track in the file. If the input clip is VA1A2, then you will have three separate MXF files (this is also dependent on your export settings on the deliver page whether audio is being exported or not.

MXF OPAtom1a is a single MXF file with all tracks included. Most cameras, if not all, that record MXF use 1a.

Supported Worklow:

Avid MediaFiles folder structure only supports OPAtom
Avid’s AMA MSP_MXF Plug-in supports OPAtom1a

Not supported:
You cannot put OPAtom1a files into the Avid MediaFiles folder structure
You cannot link to OPAtom files via the AMA MSP_MXF Plug-in

If for whatever reason media did get ingested this way, then there could be issues with the metadata. But usually Media Composer kicks up an error message saying not supported or being quarantined.

When creating Avid ready media via any of these methods you need to know that:

Media Composer gets REEL info from OPAtom files via the TAPE field embedded in the MXF files. This is done by correctly setting assist conform settings in Resolve to whatever the method is going to be. The user can see in the REEL column in Resolve what that is going to be. It should not be empty. Or if it does have a value, is it the one that you want to use? Formats like RED, Arri, DNG, etc. carry an embedded REEL that is not necessarily the same as the file name. It is usually an 8 character version of the filename to support legacy 8 character limitations in CMX EDLs. When you see a Tape name like I saw in a previous screen capture of MMSMOB.01, etc. it is because REEL was not properly set during the transcode process. For most workflows these days, selecting “use clip file name” will never steer you wrong in an Avid based workflow.

If exporting MXF OPAtom1a files, then the AMA uses the actual name of the file as seen at the OS level. So this naming is important for roundtrip conform and is done on the Deliver page in Resolve in the section for naming.”Source Name” is what you want as well as ensuring the individual clips is selected, and not timeline at the top of the page. When AMA linking, Media Composer uses the file name in Source File column for tracking, relinking, EDL’s, AAF and conform processes.

In keeping the above in mind, and having created hundreds of hours of dailies using both MXFOPatom and MXF OPatom1a, I have never had a timecode discrepancy. Not saying it can’t happen, but I have never come across one. The issues I see on these forums, other forums, and direct contact has been not properly selecting how REEL was going to be defined and settling on OPAtom or OPAtom1a.

I have recently come across one issue using MXF OPAtom1a with DNxHR – not a timecode issue, but a codec issue. DNxHR is a resolution independent codec allowing it to scale from SD to 8K if needed. I was exporting OPatom1a 1920×1080 as DNxHR SQ and it would not play in Media Composer – it gets a decode error. The very same file will play fine in Premiere Pro, Resolve, Media Encoder, VLC, etc. Making a 1920×1080 OPAtom DNxHR played back fine. Making a 4K UHD OPatom1a file decoded and played fine.

Long time requests for Avid, and I mean loooong:

1. Allow users to drag and drop OPAtom files directly into a bin and prompt user to copy/move to Avid MediaFiles folder directory.
2. Allow QT and MXF AMA plug-in to support embedded REEL metadata and track it in a column of my choosing.

Other tidbits – Resolve also embeds project name into the MXF OPAtom files whereas that field used to be blank. So if you were using MediaTool to find media for a given project you would not know (Avid DMF still has this problem). I believe that starting with Resolve 12 whatever the project name is in Resolve will be the project name in the file, so if that is important, make sure that is declared up front in the process.

Scott and I have had this conversation, I am a fan of redundant metadata – I often duplicate columns into other columns. For example, when importing BWF, I will duplicate the TapeID column into the SoundRoll column because TapeID carries the actual filename. If BWF was recorded with a TAPE name then that would appear in the Tape column. But this way I have both, and EDL manage does not support TapeID as source but does for SoundRoll. I also duplicate the START TC into the SoundTC for redundancy purposes as SoundTC has a defined frame rate whereas AuxTC can change based on the project type it is in.

Michael

[Simon Lytting]

This is only a bug in Avid MC8.x.

Fortunately there’s a solution if you found out a long time after you started editing.

First of all you need access to a Media Composer 7.x

Open your project in this MC 7.x.

Then bring your original MXF files into the project and you will see the timecode is correct on these clips.

Go through all of you original clips and type in the correct starting timecode in the AUX TC1 column.

Then delete the media with “media only” on the clips with the wrong timecode. (but which now has the correct timecode in Aux TC)

Relink all the clips again choosing the Aux TC1 setting on original clip and Start TC on target Clip.

-Simon

Voila! All your clips are now online again but from the material with the correct timecode AND you get to keep your already edited material.

Be aware though, to do this on a copy of the project, in case something goes wrong.

Good luck!