The Art Of Music Video Editorial

by Alan Chimenti

An award winning editor, Alan Chementi’s love of the blend of effects and storytelling brought him to visual effects powerhouse Radium in 2002 after a 12-year tenure at Western Images. Chimenti has nearly a 100 music videos to his credit, including Green Day, NSYNC, Soundgarden, Dave Matthews Band, Cake, and No Doubt and at least twice as many commercial spots for clients such as Honda, Lexus, Gap, and Intel.

MVWire kindly asked me to provide some of my musings and techniques on the “art” of music video editorial. It CAN sometimes be “art”, sometimes merely commerce, but usually it’s a happy medium in between. Having cut somewhere around 100 music videos starting back in 1991 on early incarnations of AVID, as well as linear based tape systems, I can offer up many thoughts. As I still cut primarily on an AVID Media Composer, I will reference AVID concepts and terminologies throughout the following paragraphs, (but most of the terms and techniques apply to AVID Xpress Pro, too). Of course, the editorial software of choice should never matter, be it Final Cut Pro, Vegas Video, Adobe Premier, or even iMovie. The platform is irrelevant. Obviously some software can handle more diverse formats than others, or has more bells and whistles, but what music video editorial (and all editorial, for that matter) ultimately comes down to is the storytelling.

A few things to keep in mind, not only when starting a video, but also throughout the entire process are:

Concept – What is the story that the director wants to tell? The video is ultimately the director’s vision and it’s up to the editor not only to execute that vision, but ideally to bring something to the table that the director didn’t expect. Of course, if you direct and edit your own work, then the process becomes streamlined. If you’re editing for someone else, be sure to repeatedly reference the director’s treatment throughout the process to be sure that you are adhering to the story and ideas that they are trying to present.

Components – How will the various elements (band performance, narrative action, graphic implementation, etc.) blend together to create an entertaining and cohesive video? Where might each of these components fall in the grand scheme of the piece? Are they constructed in a way that not only builds throughout, but also punctuates and supports the music?

Execution – What is the idea for the final look for the piece? Is it graphic? Gritty? Animated? Will there be CG elements? On what platform will it be completed? Discreet Flame? After Effects? On Final Cut Pro itself? Keep in mind how each of the judgments that you choose might not only make the piece better, but could have a drastic impact the final budget and delivery schedule. You want to be sure that you can deliver what you promise, and hopefully it is above and beyond the call of duty. In this industry, it seems that you’re only as good as your last video.

Personally, I approach all music videos the same way (granted that they are performance based). The majority of music videos that I cut are still shot on film (as opposed to tape or DV) so I have the added luxury of having all of the music synched up with the performance in the telecine process, as well as having both a clean set of final DigiBetas, and a set of “work tapes” (Beta SP, 3/4” or DV) with windows burned in. The windows reference the Address Track timecode of the master tapes (for the final assembly or “conform”), the Keycode™ of the film (in case I need a specific shot re-transferred), and the timecode of the song from the playback DAT (so I always know where I’m at in the song). These are the normal steps that I follow upon receiving the final transferred film:

1) Load, or “digitize” all of the window-burned material into the AVID, viewing and occasionally making notes as I go.

2) Break out “subclip” all of the performance and narrative takes (eliminating the sync audio track at this point).

3) Sync the “subclipped” performance takes together with what will ultimately be the final track using a common sync point between the takes utilizing the “Group Clip” function on AVID. The result of this is that all of my performance takes are now tied to a single master audio bed, and it actually allows me to toggle through all of my takes at any point in the song and see what my performance choices are. When I start editing, I use the “group” audio bed as the base audio bed in my master sequence. This allows me to “match back” at any point in my master sequence to my source “group”. I’m not sure that there is a comparable function to “Group Clip” in any of the other edit systems. It is something that AVID perfected way back in the ‘90’s, and has always been a lynchpin of music video editorial. It allows you to FLY through performance editorial. Grouping various performance takes to a common audio track also ensures the playback levels will be consistent with whatever imagery I’m viewing, so I’m not constantly tweaking my speakers for a comfortable playback level. Of course, for other NLEs that don’t have such a function, the next best trick is to lay out all of your performance takes vertically (separate takes on individual video layers), and “cross-patch” the video layer that you want to your master sequence for that particular shot.

A side note about synching tracks:

In AVID you also have the ability to “gang” the “group clip” in the source monitor with the master clip that you are cutting, so that wherever you are in the sequence, the source will follow. I believe this is “Linked Selection” in FCP.

For synching video and audio to a SmartSlate without a record label produced DAT tape, there is always the tried-and-true “Mike Sloat Method” (as lifted from the MV Wire Community pages):

Mike’s preferred playback method takes the music audio track into Final Cut Pro on a laptop and put it on audio track 1, starting at 01:00:00:00 (your time line should have a good amount of pre roll as well; in the user preferences menu, set up the time line to start at about 00:59:35:00). Then set up a SMPTE time code file* on audio track 2, starting at the beginning of the time line. From the laptop, send the left channel (track 1) to the audio playback device, and the SMPTE time code (audio track 2, right channel) to the smart slate*. This way, you have a visual cue in the FCP time line as to where you need playback from, as well as an audio cue. In post, you’ll get rid of the SMPTE track, leaving the music audio track right where it is, parked at hour 1. You can also put markers in the timeline for a quick way to get to different sections of the song.

**Stereo Mini OUT from the computer to quarter inch splitter; right channel to smart slate, left channel to the audio

One other note about syncing music with video: If you don’t have the luxury of having a “SmartSlate” or the “Mike Sloat method” on set to display playback timecode and have to resort to manually synching up the performance, remember to look for defining actions such as drum hits or vocal “percussives” such as “P”s or “B”s, to assure accuracy.

4) Break out all of the narrative elements (and break those down into subsets – using “locators” to mark the beginning of each of the takes for a scene as well as the really sweet/interesting spots). Colored “locators” in AVID allow easy cataloging, sorting, display, navigation and retrieval of said material.

5) I often create a “meltdown” sequence of interesting, or “must-have” shots that I come across as I review the footage. I keep adding to this sequence throughout the entire editorial process as you inevitably find more and more interesting shots through your regular workflow. I can easily call this sequence up if I need a unique shot. Sometimes, I will actually cut small stand-alone stories with the footage and later insert sections of it into the performance cut to begin telling the story.

A side note about organization:
Though this entire organizational part is time consuming, it is the backbone that allows the rest of the project to evolve more quickly. I cannot stress enough that organization is one of the most important aspects of any type of editorial. You have to know the location and status of all of your assets, especially when you are editing with clients in the suite. This goes not only for raw footage, but also for graphic elements, sound effects, titles, effect set-ups, etc. Your knowledge of the material and the ability to be able to call up client requests immediately upon their command, displays not only editorial knowledge and prowess, but also an understanding of the process, and a professional work ethic.

6) I’ll usually then take a day or so and cut together a loose approximation of the performance (hunting and pecking as I go), which, for me, performs several functions:
a) I begin to become familiar with the material
b) I start to give the video a pace (usually cued by the music), altering the structure to see (or feel) where things should move quickly or languidly, letting the edits reinforce the song’s pacing.
c) I have a performance bed, which is now the basis to build a narrative upon.

I always make sure that no matter what, the first line of the first verse is on the singer (as well as having most of the first verse feature the band). This is the most effective way to make sure that your viewing audience knows who the components of the band are, and establish their presence in the viewer’s mind. Though we all seek to create pieces of art with our work, music videos are more often than not marketing vehicles to sell the musicians, so it is important that the viewer is left with a favorable impression of the performers.

As I edit, I begin formulate an idea where the narrative might start to come into play, using verses, choruses, and musical cues to best dictate where plotlines should appear. I usually utilize other video layers in my timeline to build the narrative. This makes it simple to locate these sections and easy to move them around to other parts of the song without disrupting any of the existing performance bed.

From there its just going back and forth over the video, concentrating on the different sections, all the while keeping the entire piece as a whole in mind. Using feeling to add bit upon bit of performance to highlight the subtle nuances of the music (a drum fill, or bass accent that may normally go unnoticed is suddenly brought to the viewers attention), creating bits of interplay between band members, and making little sub-stories of the main plot line. It’s incredible how creating these little back-stories to the action end up reinforcing the narrative. Director Nathan Cox (also an incredible editor in his own right), once told me when I was cutting for him about his theory of the “power of three”. Basically, to tell a strong story, you need three shots: a set-up, a reaction, and a conclusion. I forget Nathan’s exact terminology, but it’s something like “one shot is an idea, two is a mistake, and three is a concept.” I’m not saying that this is etched in stone, but the human mind seems to easily link 3 ideas together. Use shots that either build or juxtapose action in order to reinforce the storyline. Apply concepts like motion vectors to help move the action along and direct the viewer’s eye where you want it to be. Utilize techniques such as flopping an image or slowing the motion to see if it heightens the drama that you are trying to create; or distort or speed up an image if you are trying to convey comedy. Look for “quirky” shots, too. Ones where a performer’s or actor’s guard is down just enough to let an inner innocence shine through. Quite often this is at the roll-up or roll-out of a take, when they are at their most human. It’s amazing how much this connects to a view at an emotional level.

As I feel that I’m getting close to having a pretty good piece, I then sit down and re-watch the performance takes to see what I might have missed. Does this close-up vocal line work better than the one that I already have there? No. Well, can I slip the sync and use that vocal in the next chorus? Stuff like that. Where can elements that are not necessarily in sync with the current place in the song be utilized in other parts of the song, be it a cymbal crash or guitar lick? I do the same with the narrative takes. There will often be more good shots in the various performance takes than can fit in the time allotted for the video, so at some point, some of your favorite shots will have to be jettisoned for the good of the entire piece.

One great thing about non-linear edit systems is that it’s very easy to move around the timeline and work on different portions when you become bogged down in one. I find music videos much like painting or drawing a picture – you move around the various areas and add a little more “highlight” here and a little more “shadow” there – all the while attempting to create a balance of the piece as a whole. And, of course, the beginning of the video is always the hardest part. I usually just skip this when I’m starting out and come back to it when I feel that I actually have some momentum going on in the video.

Non-linear editors also make it easy to create a copy of your sequence at any point so that you not only have a back-up of various stages of the video’s evolution (should something drastic happen to your current sequence that you cannot “un-do” out of, but you also have a track record of progress and are able to go back and review if what you had at a particular point may have been better than additional changes that you might have made.

Another thing that I often do is roll off work in progress to a VHS or DVD and look at it in another environment. This really helps you feel what is working and what is not. I also let other people take a look at the video when I’m nearing completion to see if they are reacting the way that I want them to, and, if not, why? (Quite often, since it’s still work-in-progress, you have to take responses with a grain of salt).

At some point, it’s just done – I don’t know how or why, but it is. You step away and send it off to the record label for comments. The whole process usually takes me between 4 to 6 days (12-15 hour days), sometimes more (in the case of an effects-heavy piece, or a “big-ticket” performer where feedback comes in layers).

Finally, how will the final elements come together for the completed piece? It’s important that proper preparation is made if the project is being handed off to a third party for the finish. Checking things like proper EDL (Edit Decision List) formats, and providing detailed notes for effects, the final audio track (especially if sound design is added to the mix) additional graphic elements that will not “auto-assemble”, and a rough cut on a suitable format will all help expedite a smooth finish. Also, be sure to have correct slate information that at least includes the artist, song title, total running time, Production Company, Director, Producer and Editor names, and a date. Quite often these slate guidelines are provided by the record company, and vary from label to label. Also, be sure to add a copyright (about 4 seconds with a quick fade up and out) a few seconds after the end of the video to at least safeguard your work.

I hope that the previous paragraphs, as meandering as they may be, shed at least a little light on how one editor deals with editing a music video from start-to-finish. Again, my techniques and suggestions are strictly personal. With editing, there are a dozen different ways to arrive at the same destination. I hope that you as a reader are able to take something positive away from this text. Of course, you should also work in whatever style or method you feel most comfortable. Ultimately, the video should take a viewer on a ride. It should speed up, slow down, twist and turn around the music, and leave the viewer with a conclusive ending.

Lastly, one great book that I can recommend for all types of editorial is Walter Murch’s “In the Blink of an Eye”. There is some great theory that is presented by this incredibly talented Oscar winning editor.

Most of all have fun with what you’re doing!


Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>