multiple audio tracks sync

Hi
I am an independent filmmaker from india
I am editing my indie feature in fcp x 10.4.

I have one video track and three audio tracks.
One from boom and other 2 tracks from lavaliere recorded in Tascam dr 10 L
I use plural eyes to sync it

For every day of shoot, I had created an fcp project with sync files from plural eyes xml

My question is how do I get the synced ok takes to my final edit?

Do I have to create a compound clips for every shot and then add them to another time line?

There are many videos about multicamera sync in plural eyes
I couldn’t find any tutorial which tells about the workflow for single video sync with three audio tracks

Is there anyother way to do this?
Please help
Thanks
biju

Is there a reason you use Pluraleyes and not just make Sync Clips inside FCPX? Just curious.

biju wrote: ...I use plural eyes to sync it...For every day of shoot, I had created an fcp project with sync files from plural eyes xml...My question is how do I get the synced ok takes to my final edit?...Is there anyother way to do this?....

Plural Eyes does not support very well the normal FCPX workflow of syncing clips in the Event Browser. It is designed for a conventional track-oriented NLE.

Thus you have to either (1) Put your clips on a timeline and round trip to Plural Eyes, or (2) First import all "sync-able" media to Plural Eyes, sync there then import to FCPX.

I have requested Red Giant improve Plural Eyes to support syncing from the FCPX Event Browser, but I'm not optimistic they will do this.

While Plural Eyes is very powerful and the audio sync algorithms very good, FCPX can also do this quite well and it avoids the "jumping through hoops" required by Plural Eyes.

As FCPX.guru said, you can easily sync this within FCPX. Some cameras emit metadata which provides a camera name upon import to FCPX but many do not. Therefore you must usually label each group of clips from each device. This is easy and fast. Procedure:

(1) Set Event Browser to List View to better see the clip data. Pick menu View>Browser>Toggle Filmstrip/List View or press OPT+CMD+2.

(2) In the Event Browser, select all clips from a given audio recorder or camera. If the Inspector is not on, use CMD+4 to show it.

(3) In the Inspector, click the "i" button to show information about the clip. This will reveal a "Camera Name" edit box at the bottom.

(4) Enter a Camera Name for that group of clips and press return. This batch labels all selected clips with that camera name.

(5) Repeat for each camera or audio recorder. Note all clips from a given device can be labeled in one step, so this is quick.

(6) Select all clips in the Event Browser you wish to sync

(7) Right-click and select either "New Multicam Clip" or "Synchronize Clips". Some experienced editors like Sam Mestman suggests using multicam clips even for single-camera situations: www.fcpworks.com/sync-or-multicam-clips/

8. For either multicam or sync clips make sure the checkbox "use audio for synchronization" is checked.

This will create a new multicam or sync clip which is synchronized using audio. You can right-click and pick "Open Clip" or "Open in Angle Editor" to verify or adjust the sync.

thank you ,joe and fcp guru
really appreciate your valuable time.

i tried the methods you mentioned,
i chose the all the video footage from one day
i have three audio tracks
and tried batch synchronising
i am using a macbook pro 2015 with 16gigs of ram
its taking forever

whereas the same amount of clips in plural eyes i was able to sync them in less than ten minutes.

i am editing a indie feature.
in the video tutorial of sam mestman ,his method of syncing multiple audio as if they are multicam clips was interesting
i think this method works for documentaries
as you know i n feature film we have numerous clips with shorter dialogue duration


i have approximately 80 video clips and 240 audio clips (because we have three mics running for every shot)

please let me know if i understood the whole workflow correctly
my question is can i sync numerous clips ? like 80 videos clips and corresponding 240 audio clips
thanks

biju wrote: ...i chose the all the video footage from one day
i have three audio tracks...and tried batch synchronising...a macbook pro 2015 with 16gigs of ram its taking forever...the same amount of clips in plural eyes i was able to sync them in less than ten minutes...i am editing a indie feature.... in the video tutorial of sam mestman ,his method of syncing multiple audio as if they are multicam clips was interesting...i think this method works for documentaries...as you know i n feature film we have numerous clips with shorter dialogue duration...i have approximately 80 video clips and 240 audio clips (because we have three mics running for every shot)...

FCPX audio sync is really intended for use on a shot or scene basis. It is not designed to create a single long multicam or sync clip of all your shots and scenes from an entire day. In your case that's 320 clips.

A multicam or sync clip is generally for one shot, one take or (at the very most) several takes comprising one scene. In documentaries it's used to created a synced multicam clip for one interview, not to create a long MC clip of all the interviews shot in one day. My documentary team has shot over 20 multicam interviews per day -- each of those are synced separately to create a multicam clip per interview.

It's not only an issue of the technical ability of FCPX to sync this. Putting multiple scenes or multiple interviews into a single long MC or sync clip doesn't make sense from an organizational standpoint. The editor generally wants each shot or interview as a separate synced MC clip, and labeled with metadata in the Event Browser. They don't want a gigantic long assembly of everything shot that day.

What you seem to be trying to achieve is the functionality of sync via timecode where all cameras and recorders are automatically in sync -- except you aren't using SMPTE timecode. Audio sync in either FCPX or Premiere is not designed as a generic substitute for this. Probably the cheapest way to add timecode sync to all devices is Tentacle: www.tentaclesync.com

This way you can stop/start your devices many times and they will all be in sync via timecode. On devices that don't support SMPTE timecode they use one audio channel for linear audio timecode. Another device is LockitBuddy. I haven't used either of these: www.lockitbuddy.com/website/index.html

If you are shooting with similar cameras that support free run time of day "timecode" and respond to an IR or button reset, you can sometime achieve a "poor man's timecode" by resetting them to zero at the start of a shoot. Even if turned off they will retain the same time-of-day timecode, and usually don't drift very much in a few hours. Dave Dugdale explains this method here:

However many field audio recorders do not have this feature, and lower-end recorders display a pseudo-timecode which isn't SMPTE timecode. However if all audio recorders and cameras supported free-run time of day "timecode", and they all responded to jam sync to zero or a fixed value, in theory they could be used for several hours without drifting much. Even if touching up the sync was required, that would be faster than syncing by hand, and it might make audio sync faster, depending on the NLE.

If you still want to sync everything shot in a single day in one operation, your best option is to first import all A/V media to Plural Eyes, sync it there, then import that XML file to FCPX.

Thanks

Dear Joe
You wrote :

Putting multiple scenes or multiple interviews into a single long MC or sync clip doesn't make sense from an organizational standpoint

I think my message didn't convey my requirements properly. Sorry for the confusion
Just to clarify, this is feature film, i have no intention to sync multiple scenes into a long clip.

I had found a work-flow which suits me.
Thanks for your valuable time

It amazes me that syncing, which is such a fundamental step of the video workflow, is still a big PITA in all workflows that are non-professional (i.e. without timecode and metadata labels entered on set by the audio guy).

Indeed if you cut in FCPX, Pluraleyes is just completely useless since what you want are video files with multi-channel audio and Pluraleyes does not give you that. (It seems that it could but it does not work.)

Syncing in FCPX can be done by synchronized clips or multicam clips (even when having only 1 camera). And as Sam Mestman demonstrates in the video joema lined above www.fcpworks.com/sync-or-multicam-clips/ multicam clips are the way to go 99% of the time. You will only go with synchronized clips in a professional workflow where you have TC and proper labeled audio BEFORE importing in FCPX and use sync-n-link to generate synchronized clips.

Moreover, there is one remaining identical problem with Synchronized clips and MC clips is that you cannot reassign their tracks' roles without opening them. This is just a pain.
And for synchronized clips you even cannot switch on/off their internal channels without opening them up. Bigger pain.

But the most annoying thing in FCPX is that you CANNOT BATCH SYNC !!! and this is really frustrating when you know what Pluraleyes (which is useless for FCPX) is capable of.

Apple really needs to offer batch sync and the possibility of changing multicam/synchronized clips' tracks roles without opening them.

In the meantime, the best workflow I found when not using TC is the following.

1. Be sure all your audio is in sync and lives as one multitrack audio (therefore of the same length) for each section that happened at the same time. Usually if you recorded everything on the same multitrack recorder, that will be the case. If it's not, it will be a pain. (Unless somebody knows a great way to do it for batches).

2. Sync in DaVinci Resolve. Throw corresponding audio and video in the same bins by packs (i.e. one bin per scene or interview, something like that). Select all, right click > Auto Sync Audio based on waveform (and append tracks, if you want to keep the camera audio).
The whole scene is now synced. Repeat for each scene.

3. Place all your clips in a timeline and export them as individual clips. Using the codec of your choice. You will end up with videos having multitrack audio.

4. Import into FCPX and add the necessary metadata to your clips. There are also several great videos from Sam Mestman and FCP Works about that.


Why this workflow is the best I've yet found when not having TC.
- You get video files with proper synced multitrack audio for you can easily change roles as you want for each track at any point of your workflow in FCPX.
- Resolve can handle almost all video formats you can think of. FCPX not so much. In particular, if you are dealing with Blackmagic RAW files, you will need to go through Resolve at first anyway to generate ProRes proxies so why not do it in a way that facilitates your workflow too.
- At the same time, you can also generate dailies straight from Resolve if needed. It has some capabilities for doing that are still not ideal in FCPX.

Still, if somebody has a better full workflow (i.e. sync, dailies, metadata, ...) for FCPX than the one I described when dealing with a project where TC wasn't used, I am listening carefully

Cheers

R.

My documentary team has shot 40 multicam interviews per site, and in general FCPX audio sync works pretty well -- provided our equipment is set up and used correctly.

A key issue is all cameras and recorders absolutely must have the same time of day clock so they will sort adjacent to each other in the FCPX Event Browser. That makes it much easier to pick out which clips go with which interview.

Another issue is all cameras must record decent-quality scratch audio. It is very easy for cameras with external mics or external audio controls to get turned off or misconfigured, then you have no audio to sync with.

Properly using a head or tail slate helps cover cases where scratch audio fails. This requires all camera operators to frame the slate, which sometimes doesn't happen. This is such a common problem that some smart slates have a bright stobe light which flashes to get the attention of the camera operators.

All cameras and recorders which can be programmed to use single files vs segmented files will reduce the number of files to sync.

Where possible use a field recorder that makes a single polyphonic WAV file, not separate files per track.

Where possible use a field recorder (like the Zoom F4 or F8) which has iXML support. This makes identifying the channels much easier inside FCPX. This was discussed in MacBreak Studio 384:



If after production wraps it becomes apparent some camera/recorder time of day clocks were set incorrectly, seize those devices before they are adjusted, boot them up and obtain the differential time. E.g, if local time was 3PM and the recorder was set to 1:45PM, the correction offset is + 1:15. All the files from that device can be batch-updated in a single step in FCPX. This procedure is shown in MacBreak Studio #223, starting at 01:50:



If the cameras support customizable in-camera filenames, configure that so the files from each camera are distinguishable. If this isn't possible, make sure during offload that the cards from different cameras don't get mixed up. Ideally offload them into a folder by production day, with sub-folders for each camera and having a unique name.

During offload, non-unique filenames should ideally be renamed for global uniqueness, and this is required to avoid an XML import bug that can create spurious duplicate clips. During import to FCPX use the option "assign keywords from folders". If all the files from a given camera and production day are in a named folder by camera name, this automatically tags the clips so they can easily be labeled in Event Viewer with a camera or angle name.

If the files are properly named, tagged and dated, and if all cameras have decent scratch audio, then sync is a lot faster and easier. Even if one camera or recorder starts before the others the audio sync will handle that.

Re batch sync, I don't see how that could work since in FCPX the parent clips comprising the various multicam clips must be identified manually. I don't see an obvious way FCPX could reliably automatically identify each separate multicam clip group. Plural Eyes can "batch sync" a large number of clips but it produces a single gigantic multicam clip which then must be cut up or extracted into smaller clips. It's not perfect, makes mistakes so you then have to check each sync and fix them. Unlike syncing within FCPX you can't get Plural Eyes to re-sync a single mis-synced clip within a longer timeline.

So it's not the presence or absence of batch sync of many multicam clips that is laborious -- it's the prep to get ready for the sync, whether individual multicam clips or a whole batch of them. If Plural Eyes (or FCPX) could batch sync a group of previously-identified clip groups in the FCPX Event Browser, that would help a little but you'd still have to do the manual identification and labeling.

Plural Eyes is really designed to sync an individual shot or interview, not all the clips you shot all day. The waveform-matching algorithm is so good it can often sync 100+ clips but the UI and workflow design is not optimized for doing batches of scenes, unless you place them all on a timeline which is not the preferred FCPX organizational method. IOW it forces you to essentially bypass the Event Browser and use FCPX as if it were Premiere.

Hey joema,

All of what you described is indeed good practice which can help many users in trouble.

joema wrote: Re batch sync, I don't see how that could work since in FCPX the parent clips comprising the various multicam clips must be identified manually. I don't see an obvious way FCPX could reliably automatically identify each separate multicam clip group. Plural Eyes can "batch sync" a large number of clips but it produces a single gigantic multicam clip which then must be cut up or extracted into smaller clips. It's not perfect, makes mistakes so you then have to check each sync and fix them. Unlike syncing within FCPX you can't get Plural Eyes to re-sync a single mis-synced clip within a longer timeline.

So it's not the presence or absence of batch sync of many multicam clips that is laborious -- it's the prep to get ready for the sync, whether individual multicam clips or a whole batch of them. If Plural Eyes (or FCPX) could batch sync a group of previously-identified clip groups in the FCPX Event Browser, that would help a little but you'd still have to do the manual identification and labeling.

Plural Eyes is really designed to sync an individual shot or interview, not all the clips you shot all day. The waveform-matching algorithm is so good it can often sync 100+ clips but the UI and workflow design is not optimized for doing batches of scenes, unless you place them all on a timeline which is not the preferred FCPX organizational method. IOW it forces you to essentially bypass the Event Browser and use FCPX as if it were Premiere.

About batch sync, yes you need a little prep. Ideally this should only be a few groups, like all scenes. But FCPX should be able at the very minimum to batch sync all selected files for a scene shot with a single camera. And as of now, it cannot. Resolve can do this very well. And that's what's really missing in FCPX IMHO. I don't want to select EACH and EVERY shot and the corresponding audio files and hit cmd+opt+G and REPEAT for ALL of them. Come on, it's 2018. Computers should work for us, not the other way around.

ronfya wrote: ...Sync in DaVinci Resolve. Throw corresponding audio and video in the same bins by packs (i.e. one bin per scene or interview, something like that). Select all, right click > Auto Sync Audio based on waveform (and append tracks, if you want to keep the camera audio)...The whole scene is now synced. Repeat for each scene....

3. Place all your clips in a timeline and export them as individual clips. Using the codec of your choice. You will end up with videos having multitrack audio.

4. Import into FCPX and add the necessary metadata to your clips....
...About batch sync, yes you need a little prep. Ideally this should only be a few groups, like all scenes. But FCPX should be able at the very minimum to batch sync all selected files for a scene shot with a single camera. And as of now, it cannot. Resolve can do this very well. And that's what's really missing in FCPX IMHO. I don't want to select EACH and EVERY shot and the corresponding audio files and hit cmd+opt+G and REPEAT for ALL of them. Come on, it's 2018. Computers should work for us, not the other way around.

I don't understand what you're saying. Already, FCPX can "batch sync all selected files for a scene shot with a single camera". You just have to select them, the same as in Resolve when you implicitly select them by copying those to a bin.

As you described in Resolve, "repeat for each scene", you'd do the same thing in FCPX. For each scene you identify the clips and sync them.

In FCPX, either the camera provides the clip metadata or you must add that. In a multicam scene this keeps each angle unique and consistent. For our Sony Alpha cameras I've spent lots of time looking through the video header with various tools, and there's no camera name or angle name in the metadata. Somebody -- the DIT or sound guy -- must add this before import or it must be added after import by the assistant editor.

Adding this metadata in FCPX takes a couple of minutes for hundreds or thousands of clips. You simply select all clips from a given camera over an entire shooting day, week, etc, and batch name them in Inspector. You do this for each camera and recorder. If you have have 2,000 clips from three cameras and one audio recorder, that's just four batch naming steps. It's five minutes of work on a 1,000 hour edit.

After assigning metadata, syncing each multicam requires selecting those clips for the interview or entire scene -- if you want a separate multicam object for the interview or scene. In the FCPX Event Browser workflow you normally want separate multicam clips for each one, not a gigantic timeline.

The procedure you describe involves importing to Resolve, manually segregating all the clips from each scene into a bin, syncing them, placing those on a timeline, exporting it as separate clips, then importing those FCPX. If you are assuming the Resolve multicam sync will work properly without adding camera name or camera angle metadata, I'd like to know how that could possibly work with multiple similar cameras having similar filenames.

If you are assuming you add the camera angle or camera name metadata in Resolve or in an external workflow, that's no different than FCPX.

Like Plural Eyes, Resolve could sync them by waveform but without camera name or angle metadata, how could it consistently keep all angles from a given camera on the same track? That is, unless somebody adds the metadata -- just as required for FCPX.

For our documentary workflow this would be lots of additional work. It's an extra transcode for each clip, and still requires assigning angle or name metadata, still requires manually identifying the clips for a given scene and copying them to a bin. I don't see how it saves any work, in fact it seems slower.

I also don't see how Resolve could export a rendered synced multicam file which FCPX would import as multicam. If you are only talking about single-cam situations with external audio that is a small subset of cases.

Also, either in Resolve or FCPX each sync must be inspected and touched up if needed. Often the sync is OK but not always, so the ability to manually adjust the sync will always be required.

There are likely ways the FCPX audio sync could be improved but I've yet to see a description of how that could work when considering (a) the required metadata issues and (b) the FCPX Event Browser workflow.

One possible way is to add scene metadata at acquisition time via Lumberjack or similar logging methods. This would tag each group of clips as belonging to an interview, scene or take. FCPX could be enhanced to batch sync all of those, assigning the multicam name to each synced clip.