[ Product Disclosure. This tutorial uses the Griddy plugin from FXFactory. I used the free trial version to create this tutorial which is why there are lines through the images. ]
Dick Osso wrote: I’ve been asked to create a Video Dance Recital, where each dancer performs at home, but needs to be combined into the entire performance. Suggestions?
This is a great question. These projects are fun to watch, but they are really tricky to build. In this article, I want to discuss the workflow necessary to create these.
As part of this tutorial, I’m using a brand-new FCP X plugin from FX Factory – called “Griddy” – that helps with on-screen positioning.
Here’s my overall workflow to create a performance video wall:
Feel free to add more suggestions in the comments below.
– – –
The biggest logistical challenge is to get audio to the performers for them to perform with. For example:
In all three cases, add a cue – like a beep – two seconds before the performance starts. This gives you a cue you can use to bring all these different elements roughly into sync during the edit. Without it, you’ll spend far too much time getting everyone aligned for time. However, given the vagaries of performance, you’ll still need to allow time to manually tweak the sync for every clip.
NOTE: Singers and musicians will listen to the soundtrack on headsets, so determine a method for them to accurately indicate when they hear the cue. Your goal is to get this cue close, but it doesn’t need to be perfect.
Also, you should stress to your participants whether they should shoot horizontal or vertical video. From experience, most people can handle this question. There is no way you can get them to standardize on a codec, frame size or frame rate. You can ask, but most people won’t understand your question. Just hope for the best, and allow time for transcoding.
BEFORE EDITING STARTS
Gather all the media, clearly label it and verify that cue points exist. Then, spend time thinking about the look you want on screen. While you can change your mind once editing starts, a little planning at the beginning will save a ton of time later.
One of the things you want to check, before editing starts, is that you have enough elements to create a finished performance. It is very hard to edit a quartet if the tenor forgot to send their file.
CREATE THE AUDIO MIX
Before doing any video editing, create an audio mix of the entire performance AND INCLUDE THE SYNC POINT in the mix. Retaining the sync point is essential for the video edit. By mixing the audio first, you can clean up problems and get exactly the right sound you need. This will simplify the video edit.
NOTE: The easiest way to create an audio edit is to load all the clips into an audio editor, such as ProTools or Audition so that each singer has their own track. This allows easy adding of EQ and dynamic level processing. It is beyond the scope of this article to describe the best way to do an audio mix.
Export the audio as a stereo WAV file and import into Final Cut Pro X to begin the video edit.
CREATE THE VIDEO EDIT
This is not a multicam edit, though it may seem so at first blush. Multicam edits create one and only one image on screen at a time. A video wall requires multiple images on screen at once.
I strongly recommend transcoding all files into ProRes 422 with proxy files. Mobile devices shoot highly-compressed video (HEVC or H.264) which is really hard to edit when using multiple streams. Transcoding will simplify your life.
I also recommend proxy files to decrease the bandwidth you’ll need from storage. In fact, because each image is getting significantly reduced in size, you may be able to use proxy files for final output. Just a thought.
NOTE: Here’s an article on how to create and use proxy media in Final Cut Pro X.
Import all media and edit the finished audio track into the Primary storyline. This prevents it from moving as you add and adjust video clips. Be sure the sync point is both visible and audible in the timeline.
NOTE: It may simplify syncing the video clips to add a five second gap (black video) at the start of the audio. (Edit > Insert Generator > Gap)
Stack all the video clips above the audio track, aligning each sync point with the master audio track. The precise vertical order doesn’t matter, because each will appear in its own section of the grid.
Then, and this is the tricky part, adjust each video clip to get the best sync possible with the audio track. When all the video clips are aligned, put a marker on each clip that aligns with the audio sync, so that if you accidentally move a clip, you can quickly get it back into sync.
When all clips are aligned and marked for sync, select all layers except the top video clip and the audio track and type V; to make them invisible.
Apply the Griddy plugin to the top clip. Griddy is a new plugin from FXFactory that makes creating video walls much easier and more flexible. Use the Columns and Rows slider to determine how many images you see at once.
Price: $49.00 (US) with free trial available
NOTE: For this tutorial I used the free trial version to learn the software, which is why you see diagonal lines through the images. Before creating your own projects, it will also help to watch the Griddy tutorial here. There are a LOT! of controls to tweak.
Griddy supports up to 400 simultaneous images, 20 columns by 20 rows, plus animation, titles, borders and highlights; far more than I am illustrating here.
With Griddy set to “Layout Mode” create the overall look you want for the project using the top clip. While it is easy to change positions during playback, now is the time to create the overall layout and look for your project.
When you have the look you want, turn off Layout Mode, then copy the effects from the top clip (Edit > Copy), then paste the effect into all the lower clips (Edit > Paste Attributes).
NOTE: Pasting attributes will retain any effects, such as color grading or scaling, that you may have already applied to individual clips.
Adjust the Location slider until each participant locks into its own grid section. Make sure, before moving this, that you don’t create keyframes accidentally.
NOTE: The Location slider is keyframable, which means you can animate a singer or dancer moving from one grid to another.
Tweak until you are happy – or the deadline arrives.
You export this project the same as any other, except that you need to set an In immediately after the sync point. This creates the look that everyone starts seamlessly.
Final Cut will use the In as the starting point for the final export.
There are two hidden “gotcha’s” you need to make this work:
Video wall edits are a fascinating way to harness the power of editing to create a group performance – even when each member of the group is isolated. These can be very effective ways to create compelling performances under very difficult performance conditions.
16 Responses to Create a Performance Video Wall in Final Cut Pro X
WELL DONE Larry…..lots of step by step…appreciate your following up on this NEWEST client request, given the world today. This gives me a huge workflow advantage I can now deliver to my clients…..who fear their clients will leave or stop….but having a Video Wall with remote clips…is safe in clients minds. Great Job Larry !!!
Ha ha – right in the middle of one of these projects in Premier Pro.
I have decided not to edit the audio until complete, and here’s why.
I have 4 musicians, and 11 singers. There are only two occasions (beginning and end) where we will have all singers heard/seen) simultaneously, for very short bursts; otherwise solos, duets, triples, and quads. I am passing the sound sweetening to another specialist, and only want him to have the actual edited groupings to work on, as this will give him the best insights as to how to work with the sound.
So I think it matters what style will be utilized. As a note, each singer sang the same lyrics, without knowledge they were going to be broken into smaller sets.
Your comments make sense. The bigger point, which you reinforce, is to handle the audio mix separately from the video. Don’t try to do both at once.
I wish you success with your project.
This is a keeper! Thank you Larry
There’s also a free plugin available which is great for these projects. GridX from Digital Heaven https://www.digital-heaven.co.uk/freeware/
This generator is fully customisable for number of cells wide and high along with the colour and thickness of the grid lines. I used it on the top layer to aid layout and tidy up non-precise layer placement.
You can have multiple instances of it with a mask or crop to allow different aspect ratios or sizes of grid in one frame.
Thanks for the reminder about GridX. I had forgotten about it.
Larry’s talked about similar setups before, including one for my uke club sing-along projects. Pretty much everything Larry says is what I do, though, I shape and “produce” the audio in Garage Band, not Logic (baby steps). Once I’ve aligned and sweetened my audio, as Larry says, I export that as a guide track to the edit, and line the tracks up by ear and waveform, then after adding synch markers to the video tracks I mute the multiple track audios so as not to muddy things, and reduce processing overhead again.
I will suggest it is smart to batch-conform all your incoming clips to the same sample rates as a preliminary step so your lip synch doesn’t drift. I also do preliminary color correction and framing during the ingest stage for each clip as they come to me… reducing processor overhead on the timeline is my goal. Proxies are indeed a useful way to work once you get to more than, say, six tracks on my older system (though my son just gifted me with a RAM upgrade to the ole 27-incher for Father’s Day, so I may see better handling now).
One other thing I will do when the project gets huge is render out sub-groups and then work with those “flattened” pieces.
Doing a Dance recital video sounds fascinating in that you might be able to use this new plug-in to make a dancer seem to dance all the way across the grid, passing “in front of” the other dancers. My head hurst just thinking about the work to keyframe that:-)
Hi Larry. Thank you for that great explanation. I just have the one question at the moment. Would it be possible to use the Synchronize Clips feature in Final Cut Pro X (Clip>Synchronize Clips), rather than manually lining up each clip as you described?
I’ve never used this feature, nor done Video Walls or Multicam editing, but have a project coming up soon that may require synchronizing clips or using Multicam.
Thank you for any advice that you are able to provide.
Maybe. If they are listening on headsets, you won’t have source audio to sync to the master track. And, if an instrument it playing its part, it won’t sync to other instruments. Both of those situations would render Synchronize Clip useless.
However, for a dance recital, where all performers are dancing to the same music – AND they record that music with their video – then synchronize clip would work.
Thank you Larry. Yes, I thought that may be the issue – musicians listening through headsets. As you indicate, the recordings captured by their individual instruments would be difficult to synchronize to the master track using the Synchronize Clips feature, given that the master track would clearly not be recorded along with the instrument they are playing.
Will Griddy run on iMac (Retina 5K, 27-inch, Late 2015) running FCPX2.3 / AMD Radeon R9 M390 2048 MB, 3.2 GHz Intel Core i5, 24 GB 1867 MHz DDR3??
Your system is fine, but FCP X 2.3 may be too old. Check the FXFactory.com website for more details. FXFactory (which is free) also needs to be installed on your system for any of their plugins to work.
Hi Larry, Thanks for the reply. I have now upgraded appropriately and am running Griddy in FCP4X. I have another question – I want to set up for 56 clips – 7 x 8 – it’s a vim spill out of the 16:9 fcpx window and can’t be seen. What am I doing wrong here? The top level clip I am laying down is 1080p 23.98… Is there a comprehensive user manual for Griddy anywhere yet?
My best suggestion is to contact the support team at FXFactory.com. They publish Griddy and will better understand how to get it to do what you need done.
Thanks Larry. This column was a big help. I wrote this down for a friend, so here’s my workflow for whomever it helps, and if the error of any of my ways can be pointed out:
Everything was done in Apple Compressor, Apple Motion, Final Cut, fxfactory’s Griddy, and some work on the title card in Affinity Photo.
Preserving image quality and color depth was a priority
The audio came from a CD we recorded in 1995. (I played bass, my wife, flute)
The CD audio was converted from 44.1khz to 48khz with Apple Compressor. That was the audio master. a 4 second front was added to the front for time for the title card.
I always obtain a score or lead sheet, or conductor’s score, orchestral reduction-whatever is available that has the most information musically-it helps to lay out a game plan for the whole piece.
It helped that we had a finished audio track of the piece-the musicians perform it yearly. All the choir and orchestra were sent a copy of the CD audio to sing along with, but none of their audio was used.
The instructions sent to the choir and orchestra we kept as simple as possible-there’s only so much you can ask -it helped that the piece was light-and they were asked to have fun with it.
I watched a bunch of tutorials and finished products, and corresponded with a few who had already been down this path-I’ve been an editor a long time (started on a Steenbeck!), but virtual choirs are a unique beast.
Everyone was given a 1-week window after Thanksgiving to submit their clips. I allowed 2 weeks to assemble the footage-and should have planned on twice that.
Conforming all submitted Clips: 2 steps
1st Compressor, then FCPX
All clips went through Compressor, to arrive a general consistency-because frame rates, file formats, clip sizes and colors are all over the map
Compressor output settings:
Video Colorspace: Rec.2020
This step was easy. I could put all the clips into a queue and let it run overnight. This is why I went uncompressed: I was concerned with the computer’s ability to play 62 clips all at once, and didn’t know what would happen when all the tiles were assembled. A past colleague from the 90s had a theory that compressed formats slowed a computer down unpacking and repacking compressed video. For this project I was using a 10 core iMac Pro with OWC SSD external storage, so I went uncompressed and was happy with the results.
Finishing that, In FCPX I created a 3840x 2160 rec.2020 library for the project, and a project file for every clip. I didn’t wan’t to be syncing or color correcting or stabilizing in Griddy-I wanted finished clips when I got to that stage. So, in FCPX:
Retiming for sync to the master 48khz wav.)
Reframe and crop: output square 1080×1080 tile.
Video noise reduction
Stabilize when necessary
4 second slug at the front of each clip for title card
Organizing and tracking the footage:
In the computer’s download folder clips were tagged according to dates they arrived (they were submitted to dropbox, & downloaded from there).
A spreadsheet was kept with original filenames, formats, names, download dates, whether they were choir or orchestra, & columns to check off completed steps for each.
I tried to keep original filenames to the end, except when clips came in without performer’s names, which were added.
Overkill?, Maybe but it was necessary for me to keep a handle on it all. Our social media people would need to know who was in the piece for tagging.
Daily backups, of course,
One thing I didn’t know who to do: make a bunch of individual compound clips (required by Griddy) without doing them one by one
Each full page (orchestra, men, whole group) was output for the entirety of the song, and then brought in and a finished product was cut together at the end. The frame and flashing Christmas lights at the end were built in Apple Motion, rendered out with an alpha channel, and keyed in.
here’s a link to the clip:
The video is a delight to watch and your idea of using an existing audio track was inspired.