Premiere Pro editing workflow for multicam wedding

[John H. Lee]

I recently shoot wedding for the first time as a full time, and just get into the editing for that event. It was 3 DSLR and 1 Camcorder shooting with 2 onboard mic and 2 external recorder (H4n - 4CH with Wireless mic and DR-05 ). Once I collected all the footages and audio files into organized directories, it's quite overwhelming to start edtiting with them. One of my editor friend told me that he usually create one sequence titled DSLR Footage and dump all the footage taken into that sequence, synchronize them first (premiere pro cs5.5 ) Then create multiple sequences with title as Main Edit, Ceremony , speeches, Highlight etc. and copy the appropriate portions into those sequences.
I wonder if there are any other approaches , in an efficient way to save time. Any recommendation will be greatly appreciated!!

[Jay West]

Your situation sounds very, very complex and, from the sound of things, you have not done much in the way of larger projects with PPro.

Seems to me you have file management and a workflow management issues.

Splitting the project up into segments is necessary for workflow management and it creates the need for file managment techniques.

I agree with your editor friend that you need to split your project up into separate segments just so you can manage all the various footages. It seems to me that it won't be enough to create only a single "DSLR sequence" as you understood your friend to say. You will need a bunch of them. You also should create new bins and sub-bins (and name them) as you need them to help keep track of everything.

I want to be sure that there is not a vocabulary problem. Some people say multi-cam and mean only that they had a lot of cameras. I am assuming that you actually were running multiple cameras simultaneously and want to build four-track sequences of the simultaneous footage with PPro's four-way multi-cam editing feature.

First thing is some of your segments are going to be multi-cam and some might not be. (Maybe your titles sequence will not be?) You have to break the project down into manageable bits. So separate out the non-multi-cam bits and make separate bins for them.

I've found that multi-cam sequences on CS 5 that get over ½ hour on my system (I7-950/12 g RAM) tend to make PPro start slowing down as it tries to keep track of all the edits. So, shorter segments are advanatgeous when you can manage it.

Within any given segment, the complication is trying to synchronize six different devices, 3 of which do not have continuous timecodes and two of which may drift out of sync with camera tracks over time .

Your camcorder, shoots continuously through a segment (say, the ceremony). It has audio and video in sync. You use that as your reference clip for each segment. (I am assuming that you let the camera run continuously in the segment. That is, you let it run all the way through the ceremony, and did not have somebody starting and stopping recording.)

When you have placed your camcorder video & audio in the sequence, I recommend you next address your two audio recording devices. Unless you had an accurate synch pulse, you will have to manually sync up the two audio tracks with the camcorder audio. (I'm assuming you were using them together and that you need both of them). You do this by matching the wave forms or by using software like Plural Eyes. If you plan on doing much with your set-up, I suggest you download the Plural Eyes test version and see if it works for you. It could be a big time-saver and well worth the $150.

That gets you the initial sync so that the audio all starts with the video. If you are working in very short segments, say, 5 minutes, then your audio is probably fine and you can go on to syncing the DSLR tracks. When you are working with, say, a 20 minute (or longer) ceremony, you have to deal with "clock drift" to be sure that sync is maintained. The problem here is that digital devices have internal clocks that may vary over time relative to other digital devices. This is more of a problem with some devices than others. For example, I have an mp3 recorder which will go 15 to 20 frames out of sync with my cameras over 8 minutes. On the other hand, I've heard that the Tascam units have clocks that stay very close to camera clocks, so you might not get much drift with the audio from that unit. I've heard varying reports about the H4. Either way, you will need to check the audio at regular intervals to be sure that the devices remains in synch with each other and with the camcorder video. If a device's track drifts out of synch, then you need to make a cut and slide the track back into sync.

So, let's say you've made three or four cuts to get your H4 audio in synch with the 40 minutes of video of the ceremony. You are going to need to use this track again when you go to multi-cam. Easiest way to do this is by exporting the audio as an audio file (*.wav on a pc) which you then re-import and lay over the H4 track so you have continuous audio. (Or, if you got Audition with your CS5.5, send it out to Audition and bring it back as a single file). Alternatively, you also can put it into a new sequence and then nest that sequence over the H4 track in the original ceremony sequence. While this increases the number of sequences and also requires you that do a "render and replace" in order to see the audio waveforms, but it keeps track of the edits in case you needed to go back and adjust something. Which method you use is purely a personal preference.

When you have all of that done with all the audio waveforms lined up, you are ready to start tackling sync with your DSLR segments. I do not use DSLR cams, myself, but I understand most DSLR cams are limited to shooting clips which can be up to 12 minutes before you have to start a new clip. So, in your 40 minute ceremony, each of your DSLRs might have 4 clips that you have to sync to your video. (Maybe your DSLRs shoot longer segments?) Start with the first DSLR cam's first clip. Lay that clip on the track. (You may want to add adiou tracks and move your two device audio tracks down to the bottom. That way, you can bring in your DSLR clip and its audio without accidentally overwriting your device audio). Sync your first clip with the the other tracks. Easiest if you can match photographer's flashes with the ones appearing on the camcorder track but, otherwise you are matching up to audio waveforms. (Hopefully, the room was not so big and the DSLR not so far away that you have time-delay and phasing issues; if you do have those issues, use the audio waveforms to get close to sync and then check mouth movements while playing the audio.)

You repeat this process for each of the ceremony clips from DLSR cam # 1.

Now we get to where you editor friend probably recommended a DSLR sequence. The reason for this is that PPro's multi-cam synch works with whole clips. If there is are multiple clips on a track, PPro only syncs the first clip and does not include any of the clips on that track when you go to the multi-cam sequence to edit. There are a couple of ways to get your 4 clips to show up as one continuous clip (with black spaces between the restarts.) One way is to pull a box with your mouse through the clips on your DLSR track and copy them (Cntl-C or Edit--> Copy). Open a new sequence (which you name so you can keep track of it) and paste the track you've just copied. (Cntrl-V or Edit-->Paste), You now have your clips all perfectly spaced on a timeline. Nest that sequence over the first DSLR track in the original sequence. (Drag the new DSLR sequence over the DSLR track on the previous sequence.) Alternatively, you can export just the DSLR track to an avi or mov file which you import back and then lay over the DSLR track. Again, the choice of methods is personal preference. The sequence method allows for later readjustments and the export method is simpler. Personal preference as to which you use.

You repeat this with each of the other two DSLR cams you use with the seqment.

Now you are ready to sync (Clip--Synchronize), open a new sequence, drop you first sequence into it, enable multi-cam (Clip --->MultiCam--> Enable), and open the multi-cam editing window (Window-->Multi-Cam). When you have done that, drop your two "wav" audio files onto audio tracks in this sequence. (In the multi-cam window, I am assuming that you have not checked "audio follows video." If you have that checked, you probably want to uncheck it.

You said you had "two on-board mic." I'm not sure what you mean. Do you mean two tracks from the on-board audio on you camcorder or on-board mics from two different cameras or did you mean you have a camcorder like a Sony Z5/NX5 which allow you to use the built-in mic in combination with the a shotgun mounted in the camera's mich-holder? If it is one of the latter two, you may need to drop another audio track into your multi-cam edit sequence.

You basically repeat this process for each of the segments you want to have in your project. Because this will produce a lot of sequences, you may want to consider making a bin for squences, and maybe have subfolders/bins within that for particular segments. It is a good idea to rename your sequences. Instead of, say, "sequence 2" you might find it easier to call it "DSLR # 1 Ceremony" or something like that.)

At the end of this process, you have a bunch of discrete segments. I'm assuming you are supplying a disk to the couple. Depending on how you want to set up your disk, you could: (a) assemble all of your sequences together into a final timeline (nesting them in order) and export to Encore (and let it work out the bit rates needed to get it all on one disk); (b) separately encode each sequence and import the encoded files to you could build different timelines and arrangements (assuming you can compute your disk space requirements accurately enough to specify the bit rates for the encoding); or (c) import the PPro sequences into Encore, do your assembly there, and again let Encore work out the bit rates needed to get everything to fit on the disk. Or, you could use variations. Maybe you want to have a documentary-style timeline running from introduction/titles through the end of the reception and then have a separate timeline for you Highlights movie?

However you want to do it, you will find that organizing things in bins and sub-bins, and specifying names for sequences will make this a lot easier to keep track of all the pieces.

[Jay West]

One other thing. When you have collected all your files into folders and directories on the computer, you can use the import "folders" option to in PPro to build your imported bins. For anything that has metadata files associated with it, you will get an error message from PPro saying that it could not import certain files. Just click okay and go on. It will have imported the video and audio and you will have your directory structure.

[John H. Lee]

Thank you so much, Jay, for your in-depth reply!! I've actually printed them out and read multiple times. It is more than I ask for and I am deeply thankful.

[Justin Molush]

Just posting here because the above writeup is great. I was going to multi-cam my next live production and this just about covers all the stuff I knew I would have stumbled on the first time through. Thanks for the writeup, it will save me quite a bit of time later.

[Victor Boyko]

does plural eyes sync up video along with the audio from multiple cams? Or does it just sync bunch of audio sources from different recorders and doesnt touch the video from 2, 3 or 4 video sources?

[Jay West]

If you lay out your camera footage on multiple tracks in PPro, the video follows its linked audio. You have multiple cameras, with each camera on a track plus its linked audio plus addditional second source audio tracks? You export that to a Plural Eyes format, open it in Plural eyes and it syncs the audio tracks with each vieo tracked being pulled along with its linked audio track. That then goes back to PPro for your editing.

You can find links to video explanations, manuals and trial versions here:

Singular Software - PluralEyes for Premiere Pro

If you've got a Windows machine, be aware that some of the documentation seems to assume one is using a Mac even though Singular clearly does have PC/Windows versions.

There are other discussions of Plural eyes in various places here on DVinfo, so do a search. I am not especially knowledgeable about it as I only worked briefly with the downloadable trial. It worked as advertised but I just have not needed it. I shot multi-cam projects with my various audio sources feeding to camcorders which I synch with camera flashes. My few secondary (non-camera) sources have been synching so easily that I have not felt the need to buy Plural Eyes for my work. I'd probably feel different if I were using a lot of DSLR cams and independent audio sources.

[Tim Palmer-Benson]

I am using PP 5.5 in the manner described in this thread. I am using Pluraleyes on a MacPro 3.1 3GHz, 8 core, with Nividia Cuda card. It works great. But one of the problems I have is with PP itself. It seems that it cannot handle multiple tracks of H264 at the same time. In my case I have at least 4 to 5 Canon DSLR cameras I need to synch up in Pluraleyes along with several audio recorders. Pluraleyes handles this just fine but not PP even though I have a CUDA video card. As soon as I start doing color correction, PP chokes and splutters. I have to do a lot of rendering! Do I need to go back to the FCP workflow of making all my clips Pro Res LT?
By the way, I also have a Matrox Mini for HDMI output. Adding this into the mix can freeze PP so that it won't playback at all and one has to reboot. Numerous calls with tech support in Montreal have been unsuccessful in resolving what I am thinking is basically bad software/hardware.

[Marcus Durham]

Originally Posted by Victor Boyko

does plural eyes sync up video along with the audio from multiple cams? Or does it just sync bunch of audio sources from different recorders and doesnt touch the video from 2, 3 or 4 video sources?

It's supposed to. Although I purchased it earlier in the week and it's taking a dislike to my AVCHD footage. Syncs everything else, but messes up the footage from my AVCHD camcorder.

I have a support request in asking why it's not working. It must be some kind of codec issue as it managed to sync everything else.

[Tim Palmer-Benson]

Hi:
I don't know whether this will help, but Pluraleyes doesn't work with the video at all. It works with the audio attached to a video clip and with the audio from a separate audio recorder. As long as the audio recorded with the video is in synch (it normally is) when it comes from a camcorder, it should synch with the audio from other camcorders and with separate audio recorders

[Jay West]

Originally Posted by Tim Palmer-Benson

I am using PP 5.5 in the manner described in this thread. I am using Pluraleyes on a MacPro 3.1 3GHz, 8 core, with Nividia Cuda card. It works great. But one of the problems I have is with PP itself. It seems that it cannot handle multiple tracks of H264 at the same time. In my case I have at least 4 to 5 Canon DSLR cameras I need to synch up in Pluraleyes along with several audio recorders. Pluraleyes handles this just fine but not PP even though I have a CUDA video card. As soon as I start doing color correction, PP chokes and splutters. I have to do a lot of rendering! Do I need to go back to the FCP workflow of making all my clips Pro Res LT?
By the way, I also have a Matrox Mini for HDMI output. Adding this into the mix can freeze PP so that it won't playback at all and one has to reboot. Numerous calls with tech support in Montreal have been unsuccessful in resolving what I am thinking is basically bad software/hardware.

I ran into somewhat similar problems a while back but I am not sure how much my experience will help you. I run PPro CS 5.5 with a Matrox Mini and run multi-cams with at least 3 tracks of AVCHD and as many as four HDV tracks. Unfortunately, I do not use a Mac and do not use any DSLRs. Here goes, anyway.

First, on color correction, PPro has a large number of effects but only some of them are CUDA enabled. If you are using "Color Balance" of "Color Balance HLS," for example, those are not CUDA enabled and can be slow to render and, until rendered, can seriously bog down or impair playback.

Second, playback of multiple tracks of H.264 from DSLRS can be a big load on the system, much as AVCHD (and maybe more so.) In PPro CS 5, before Matrox came out with the new drivers, I was using PPro's own filters and monitoring tools. I found I had better luck running mulitple AVCHD tracks by using PPro's own AVCHD timeline presets for my multi-cam sequences. Now, with CS 5.5 and the latest Matrox drivers, I find it all works pretty well using the generic Matrox 1920x1080 HD timeline setting, although it is best to work with segments of no more than 20 minutes.

My recollection is that, when runing with the Matrox Mini, there are only a generic "DSLR" timeline settings. I am not familiar with them, but are they supposed to be H.264?

If not, that brings me to my third point, which is actually a response to your question about ProRes conversions. When I am doing many-camera multi-cam mixes, my preference --- PC based, obviously --- has been to use Cineform NEO, which I suspect would be a lot like doing a ProRes conversion. What Neo and ProRes do is basically convert your highly-compressed original footage to a much less compressed avi or mov format. This results in much large file sizes but much easier editing/playback because PPro does not need to try to decompress on the fly (which is otherwise required when editing AVCHD and H.264.)

I find Cineform Neo particularly useful for color -correction and matching tasks. I can run Cineform's "First Light" application alongside PPro. First Light operates in metadata so there is no real rendering required and changes appear almost instantly in PPro. Basic color matching etc. is thus easier on me and PPro. Cineform does hav e Mac versions. I don't know much about Pro Res, so can't say much more about that.

[Dan Asseff]

Jay,
I am new to PP5.5 and am loving it. My problem is trying to set up a multi-cam edit. I sync up the cameras on the time line then right-click on the time line and select nest. The I load the nested video on the time line then right-click then enable. Then go to window drop down and select multi-camera monitor but only one video shows, what am I doing wrong. Hope you can help. Thank you

Dan

[Adam Gold]

Hm. Sounds like there are a couple of new commands in 5.5. The right-click > Nest is new to me and doesn't do what I'd expect it to. I just did what you said you did and it didn't work for me either.

I stack the clips and sync them in sequence 1, then create a new sequence (File > New > Sequence). Drag sequence 1 from the project panel into the sequence 2 timeline track 1, right click on the track and choose Multicam > Enable. Then Window > Multi Camera Monitor. Should now show a quad split on right and Preview Monitor on Left if you have Show Preview Monitor checked. Make sure the CTI is in a part that actually has all four tracks in it.

[Dan Asseff]

Adam
Thanks for the reply. I did what you said to do but it didn't work last night, so I gave up. I closed the program.
This morning did what you said and it worked, go figure. I guess there is some kind of bug. Thank you again for the help. What is CTI?

Dan

[Jay West]

CTI = "Current Time Indicator."

Basically, the cursor.

I think Adam was referring to sequences where you might have one or more nested multi-cam segments mixed with some stretches of non-multi-cam material. If you have the CTI in a part of the sequence that is a nested multi-cam sequence, you will see the four-way split window. If you move out of the multi-cam section to a single cam section, you will see only a single large image. Also, if some of your original tracks are longer than others, the others will drop out of the multi-cam display when their video runs out.

Something similar happens when you layer non-multi-cam tracks above your nested multi-cam sequence. This is something I do regularly when working with more than four camera views. (I put the four main cams together in a sequence, sync them. and then nest them into a second sequence with multi-cam enabled. Over top of the nested sequence, i layer in the additional tracks, which are mostly going to be cut-away shots that I will need only occasionally. I only activate them --- that is, click on the "eye" icon in the left corner of the track's control box --- when I need a cut-away shot.) The uppermost activated video track is the one that gets displayed in the monitor windows. This is true for the timeline playback monitor window and the feed through the Matrox Mini if you are doing that. Interestingly, when you activate one of the upper tracks, your multi-cam display can still display the quad-windows while your timeline playback monitor will show the uppermost activated track. This can be handy when you have five camera tracks you want to monitor.

Another thing to be aware of with regard to the nested multi-cam sequences: the sequence from which you are nesting has to have the "eye" icon activated for each of the tracks you want to multi-cam. If you go back to the original stack of sequences and deactivate one or more of those tracks --- which you might do if you wanted to apply a color correction/balance filter to only one of the tracks --- you have to remember to reactivate the others when you are done adding the effect to that track. If you don't reactivate, then you won't see the others in the multi-cam window for your other sequence.