FCPX Audio Magic–matching tonality between clips

FCPX Match Audio

I knew there must be a way. I knew it, I knew it, I knew it.

I won’t tell you how many hours–I mean DAYS– I spent trying to match up tonality between two different audio takes.

I had a 2 hour interview to cut down to about 12 minutes.

The first thing I did–before cutting anything–was to give a rough mix to the whole interview audio track. That’s so no matter what order the pieces wound up in, they’d all be the same mix.

Then I fine tuned it to a final narrative at which point I copy-pasted the audio track into a separate project where I mixed it and then exported  it back into my original project as one 12 minute piece. That way, any further mixing could be applied to the whole track and not have to be applied to dozens of little pieces.

I forgot one very important thing–to keep a copy of that first mix saved as a separate project I could go back to if need to. I had called that project “work” (meaning it was a temp project just to be able to do something out of the main project). I later used the ‘work’ project again, this time over-writing it to do something else.

Well, that was dumb.

After I had crafted the narration and added the B roll for my rough edit with my mixed audio I realized I needed an additional piece of narration to cover a particular subject.  So I found the piece in the original footage after realizing I had over written the project where I had mixed it and exported it from. You see, even if the piece I wanted to use wasn’t in the original project, I could have either dragged out the clip nearest it to find it (at which point it would have the same mix as everything else, OR I could have simply copy-pasted the audio effects from the mix onto the new un-mixed clip.

Now, try as I might, I could not duplicate the original mix to achieve the same tonality with the new clip. I tried all kinds of different EQ tools, studied the frequency response to try to match it, but all to no result. The original mix dealt with ambient noise removal, making the voice more present and fuller. No matter what I did, I could not match the new clip to the mixed version. I couldn’t remember the order I did what in (and that does matter)–and there was a looming deadline.

Finally I must have googled the right question.

Right there at the top of the list was a Larry Jordan article on this very problem: matching tonality between shots.

Turns out the solution is right there in FCPX and is idiot simple.

With one click I suddenly had a very near match that I was able to tweak a bit and all in a matter of a couple minutes (after wasting untold hours) doing other stupid things.

The main difference now wasn’t really tonality, but the talent’s emotional tone and pace, but I was able to make it close enough to not be really noticeable.

Here’s the Larry Jordan link that explains where to find the magic tool and how to use it. It will take you about 30 seconds to learn it and not forget it. Larry’s article on matching tonality in FCPX.

Notes to self:

1) ALWAYS, ALWAYS, ALWAYS save every project for exactly what it is, even if a temporary work project that’s been copied pasted out of your main edit for the purposes of being exported or copy-pasted back to it.

2) Periodically duplicate your project (particularly after completing a significant phase) and number them all in sequential order while continuing work on the new duplicated project which is now your current version. Leave the old ones alone. If you ever need to go back to cut and paste an audio or video effect to your current version of the edit, all the information will be there.

B Roll, What it is and how it will save your bacon

B roll

First of all, what is it?

For a comprehensive definition and history, here’s the Wiki.

In short, in the early days of linear film editing (and tape editing) of documentaries, interviews or dramatic films, the main scene, master scene or main interview was strung together on one roll of film or tape called the A roll. Wherever that scene was to cut to another shot (a cut away, an insert or simply another scene not from the main scene) the A roll would get a piece of opaque leader and in the corresponding blank spot, the appropriate shot would be synced up on the ‘B roll’, itself separated by opaque leader which corresponded to those parts of the ‘A roll’ that were meant to be seen.

Later it would all be married up. This procedure is what made editing ‘seamless’ (otherwise you’d see constant flashes every time two pieces of film were joined together with editing tape). It also allowed for seamless transitional treatments (such as dissolves, fade to black, etc.)

These days it’s all done on computer—even features shot on film. The film is transferred to video where it is edited and then output back to film. Computer editing programs, of course, are called non-liner editing systems where you can simply drag and drop shots, transitions, effects, etc. with impunity. It’s intrinsically seamless.

Anyway, while B roll retains its original meaning, it has also modernized itself through the magic of Time and Slang.

Today in corporate video work, wedding videos, events we generally mean B roll to be any shot that’s not part of the narrative (such as an interview) or story, such as a scripted film. In film they generally call this ‘coverage’, which means you shoot a bunch of detail shots (the clock on the wall, the leaves blowing on the trees, birds in flight, people walking down the street, etc, ad infinitum) that might come in handy for the editor if he’s in a tight spot. Great film makers with large crews usually plan their films to the nth degree and specify all these things. Even then, smart cameramen shoot more than was asked for even by a detailed script.

In the run and gun category, when you don’t have crew support (assistant directors, script supervisors, etc.) B roll is essential. It will be, in fact, the main pictorial content of your video.

Let’s say you have a 15 minute interview. Or even a 3 minute one (in final edited form).

Who wants to see a talking head for 3 minutes or 15 minutes straight?

More to the point, most interviews of any length have to be edited. How do you cover up the cuts?

B roll.

But the purpose of B roll isn’t just to cover up cuts. Yes, that’s where it can save your bacon when you found yourself forced to do an edit and have absolutely nothing relevant to cover it with. More importantly, B roll should help forward the story of the narrative. It should help forward the overall message of the video. It needs to be relevant to what’s being said.

Let’s say you’re planning a documentary which involves traveling across the United States and stopping in various cities to interview certain people on the same basic set of questions. One thing is for certain. You’re going to interview those people. That’s what you do know. But you’ve never been to these places and have no idea in advance what also should be shot in terms of B roll, so you’re going to ‘wing it’ and hope you’re covered.

Well actually, you already know more than that. You know you’re going to a particular city or town.

Most every city or town has a ‘Welcome to ___’ sign. Shoot it.

Mount your camera on the windshield or handhold it and shoot some traveling shots, highway signs, landmarks and anything of interest as you arrive. You may only wind up using one or two of those shots—or none—, but so what? You’re not doing anything but driving. (let me clarify that: If you’re alone, use a mount. If you’re with someone, have them hold the camera).

Shoot some establishment shots of the town and its landmarks. Long shots, medium shots, close shots, detail shots.

Shoot the bustling downtown–or the vacant one.

Shoot the exterior sign and buildings of the location you’re going to.

You know something else too. You know what the questions are and the subject material to be covered. So shoot stuff in that town that’s relevant to the subjects that will be discussed.

If relevant, shoot the interior of the establishment you’re going to. Not just long shots of interesting rooms, but detail shots of interesting details.

Yea, shoot atmospheric shots of flowers, trees, birds, lakes or whatever is there and may be relevant to what you’re doing.

And if it turns out you need a bird in flight for the Chicago edit and you didn’t shoot one, use the St. Louis bird you shot. Who will know?  It’s called ‘cheating’ and Hollywood has always gotten a kick out of cheating.

If you spend an hour or two on the interview, spend that much or much more on the B roll.

After the interview you may realize you need to shoot something else based on what was discussed.

So shoot that.

You get the idea.

For a video to be interesting, it should not only have interesting narrative or story content, it should also be rich in story-telling pictures (footage). You won’t use a fraction of what you shoot, but you will have so much latitude in editing for having shot it that you will be free to make a great video no matter the headaches you may have endured during the interview itself.

Chapter 12, Corporate Shootouts–Supplement

Chapter 12 covers various aspects of producing an interview-driven video. Specifically it covers a method of producing a corporate video without prior scripting.  For that to work, Chapter 12 puts a lot of attention on a workable approach to conducting interviews.

Good interview content, edited properly, can give you a much more sincere (and believable) narrative than any ‘scripted’ attempt because ‘scripts’ are best carried off by professional actors and presenters. The only problem there is that most people can spot them a mile away, so it’s a two-edged sword: Local talent are terrible at pulling off scripts and professional talent betray themselves as professional presenters reading script.

On this page I’m going to use an example of my latest corporate video done for a company called Axiom which was subcontracted by another company (Logistex) which was coordinating a number of construction activities on the Superdrug distribution site.

You will notice that no person from Axiom appears in the video. This is, in fact, a testimonial-type video whereby the end user (Superdrug) and the prime contractor (Logistex) are giving testimony on Axiom’s behalf.

For those interested, this is a training exercise, so is therefore presented as a case study for the corporate video model laid out in Chapter 12 of Run ‘n Gun Videography.

I will first link to the video itself so that you can watch it and then follow up with some screen shots of the editing timeline to illustrate a few points along with some additional comments.

Let me start off with a comment about something you will notice in the first 15 seconds of the video: You will notice that you’ll hear the narrative content (voice) a full 15 seconds before you ever hear the person talking. That’s longer than I would normally do, but in this case the specific reason you don’t see the person is that the first 15 seconds of narrative had 8 edits in it–mainly cutting out ‘ums’, ‘ahhs’ and longer than desirable pauses. That was necessary to build an introduction from the interview material available.

Here’s a screen shot of the beginning of the timeline showing the 8 edits in the first 15 seconds.

Narrative edit sample 1 (1)

The real key to seamless editing of interviews is to do it in such a way that no one would know there were any edits there by listening. It should sound to the listener as if that’s exactly how it was spoken live in the first place.

Here’s the video:

As described in the chapter, the first step after importing the full interview content is to mix the audio if necessary. That’s because you’re about to cut it up into a lot of pieces. If you mix the audio afterwards, you’re going to have to apply the mix to all the individual pieces, so this is an elementary time-saving step. In the case of this video, there were two interviews totaling about 25 minutes of material.

(for some reason I notice that the audio of the second person sounds very sibilant on this upload–something I’m going to have to look into when I get home)

After mixing the sound on the two interviews I then went through each and cut out all of the extraneous and unusable or irrelevant material (such as all my questions and bantering) leaving only the material that I would be drawing from to construct the narrative.

While I did each interview in separate projects, I’ve combined them here on the same timeline for ease of illustration. That’s the 17 minutes of material I would be working with to create the narrative.

interview bites

The next series of steps are a process of moving those pieces around into an order that will tell the story you want to tell. For example, you’ll find certain things suitable for possible beginning bits and others for possible ending bits. You just shuffle them around roughly in this fashion.

Once you’ve got things more or less in order, now start deciding what the best and most relevant material is and what the worst or least relevant material is. Once you’re certain you won’t be using certain bits, start deleting them.

Now you may find your 17 minutes of material is down to 9 minutes. But you’re going for a 3 minute video, so there’s lot’s more to do.

In this fashion, you start building your narrative out of the most relevant pieces. At this point you’re not concerned with any flaws in delivery such as um’s and ahs or other peculiarities such as always starting a sentence with ‘yeah…’ or any number of other quirks. You just was to see if you can put together a cohesive story that gets the point across. In other words, you’re concerned with message.

Sometimes you will find the same thing was said in two different ways at different times. Take the best of the two and delete the other.

Keep paring it down and clarifying the message.

Finally, when you’ve got your content down to the right length for your video (there’s no hard and fast rule except that the content should maintain interest for the prospective target viewer–in this case, someone looking for an industrial sortation system), now you can start fine-tuning it.

If you haven’t already, decide where you want to guy to absolutely appear on camera. The rest of the interview is fair game for editing. Now you can start slicing and dicing to remove the unwanted pauses, the self-corrections (happens a lot–they’ll start talking and then start over repeating what they just said, they’ll say a word wrong and correct themselves, etc.). You can freely trim all that stuff out and pace it all nicely together (even if you have to add small gaps in the narrative when things seem to run together too quickly).

By the time you’re done with that, you’ve probably cut out another 10, 15, or 20 seconds or more. You haven’t changed the content, but you’ve cleaned it up and clarified it.

Naturally when you get to the step of covering up all those edits with B roll, you will have to have some relevant B roll to cover those points so that it too is contributing to the flow of the story.

In this video you can probably tell that the first person who speaks was at ease with the process and interested in telling the story. The second person was more self-conscious and required more editing to smooth out his narrative. You can see it in the following screen grab from the timeline:

Narrative edit sample 2 (1)

Incidentally you can also see a place where I added in a small gap in the narrative for reasons I can’t remember, but which improved the overall narrative. You’ll also notice at the very bottom a piece of track in there which overlaps the gap. That’s room ambience that I recorded (which contained the factory noise in the background) so that the gap wouldn’t suddenly go ‘quiet’ and betray itself.

And while the subject of a different chapter in the book, you can also see where I extended the music track by finding a phrase of music that I could repeat at this point in the video so the music ending would correspond with the end of the video.

There’s nothing particularly novel or revolutionary about this approach. Editors have been doing this for a long time, but for those of you working on your chops, I hope you will have found this to be useful.

Chapter 16, Notes on Music–Supplement

This chapter in the book discusses how to choose and edit stock music so that it seems to have been scored for your video as opposed to just tacked on. The result can be a massive improvement in the overall quality of the production.

While I have been using the formula covered in this chapter for several years, it’s only been in the last year or so that I’ve been generally pleased with the results. Below you’ll find several different samples of the result with some notes on each.

As a note, one of the attributing factors to improved results came as a result of using the Audio Jungle website.  The reason for that is the ability to go to the portfolio of someone whose music you like. I might have found a piece that was a great piece, and the sort of thing I was looking for, but disappointingly it didn’t quite fit. So I’d go to the person’s portfolio on Audio Jungle and see what else he or she had. Not only would I find several more pieces that were generally suitable for the types of videos I do (and could ‘favourite’ them for future reference), I’d very often find the exact piece I needed. There might be several artists I generally like and I’ll check each one’s portfolio until I find what I’m looking for.

As I write this, the latest improvement on the Audio Jungle site is the ability to very specifically narrow down your search to the length you’re looking for (in addition to genre, instrumentation and tempo). So if my video was looking to come out at about 3:15, I’d set the parameters to 3:10-3:25 or something like that. This really speeds up the process of finding a suitable piece of music for what you’re working on.

Now let’s get to some finished examples. If you’re here to study the stock music model outlined in the book, do watch each video all the way through. Listen for the different cues of music (distinct phrases or melodies within the song), shifts in tempo, how certain edits align with the beat and even how often something the person is saying is complemented by the music.

Example 1:  PTA UK Case Study, Leyland Saint Mary’s High School

This is, so far, my favorite match of music to video.

It is also the first time I used two different pieces of music in a video, mainly due to it’s length.

The video is about how a PTA was formed in the aftermath of an arson fire that destroyed most of a school. The natural story-telling method is “Problem/Solution”. We start off with documenting the incidence of a devastating fire and it’s consequences, then end with how the formation of the PTA contributed to the successful continuance and improvement of the school. That required two different moods of music for each respective part of the video.

In the chapter I talked about how you can mark certain significant music cues to align with certain significant narrative or pictures. It happened several times in this video, but the most important one for me came near the end. The head teacher has just finished off her points about the past to the present and then shifts to what she hopes it will mean for the future. There was a point near the end of the music piece where there was a break and shift of mood to the final cue. This is a point I made sure to align to the narrative.  Other less important timings were accomplished simply by where and how I edited the B roll to the music. There were also some ‘happy accidents’ where the music seemed to echo the narrative without me doing a damn thing.

Example 2:  PTA UK, Annual Gold Star Awards

Another PTA UK video, this one much simpler. It’s essentially an assembly edit of an event driven by interview narrative. This was one of those cases where I found the music piece I knew it was perfect. It didn’t take me long to find it and I didn’t waste any time looking for anything else or anything ‘better’.

Confession: I just noticed in reviewing this that the volume shift of the music at the end was a bit too sudden. Whoops. That’s a mistake.

Example 3: Belvoir Castle Cricket Trust

This was a pure assembly edit with no narrative. Therefore I had total freedom to edit to the music without getting too robotic or crazy about it. There are perhaps a few dozen edits that are very specific to the music which will be very evident.

Again, for some reason, it didn’t take me very long to find this piece, but when I did I knew it was perfect. This video had different ‘chapters’ to it in a way. There was a section on working with special needs kids, and even a subdivision of dealing with wheelchair-bound kids, as well as a special needs tournament, but there was also a section on able-bodied kids as well as a subsection having to do with country activities outside cricket.

The piece of music I found had built-in ‘chapters’ to it and was of the right mood and tempo, so I knew it would work.

The video itself, was an unplanned one. The footage was part of a broader commission, but when I found out there was going to be a Marleybone Cricket Club fundraising dinner at Belvoir Castle with the Duchess of Rutland and various celebrities in attendance, I  offered to create a promo that could be shown at the dinner.

I took the best B roll that I had at that time (which totaled 8 minutes), then laid it all down on the editing timeline organized in blocks of shots that represented the various ‘chapters’ mentioned above.

I then put the music underneath, leaving room for the title sequence.

I put markers on the various significant ‘chapters’ in the music and then, block by block, whittled each chapter down to my best shots, and then edited them into the timeline to the music.

It was a very fast edit.

Example 4: Why Choose Redstones

This is a more typical example of fitting music to an edit. In this case because I was editing multiple interviews to construct the narrative, the narrative took more priority in the editing process. So in this case I completed the entire edit first and then selected a piece of music I felt would fit. This was also a case where, after finding the music, it was necessary to extend the music to fit the length of the edit. In some cases I timed certain action in the B roll to fit beats of the music, and in one or two cases I adjusted things so that the narrative would hit a certain music phrase or beat, but for the most part, this was a case of narrative determining the edit, finding music that would be suitable and then making a few adjustments to the edit in regards to the music.

%d bloggers like this: