Editing Extravaganza - Visual & Audio Fusion

Hey folks! I am back! But I wouldn’t be truly back if I didn't share the depths of my production process with you. 

Over the past couple of days, I have been struggling with the monotony of editing. However, I have made significant progress. All I have left to do is insert titles! As for this segment of editing, I have been working on audio and visuals. Take a look below to hear how I navigated these fields. 

Audio 

First up the audio. As I mentioned before I was planning on carving out 20-30 second chunks and then manipulated the visuals accordingly. To continue with this process I had to decide what parts of the interview would be split into a voiceover and what parts would remain in the interview format. 

The interview portion was a full 2 minutes and the ideal audio length would be around 1:30. The issue with snipping out the audio is that there is assumed to be a direct address — it should sound like the detective is narrating her story with this case directly with the audience— therefore it shouldn’t be choppy. 

I knew this would be an issue for me because my actor stumbled over a couple of words and there were external factors like movement and dogs barking that were hindering important case details. I was hoping that there would be some feature in CapCut to isolate the voice of only the main person talking in said clip. Now I did find that feature called “Isolate voice” but it didn't have the desired effect. It made it sound AI-generated and had a very weird effect on the audio. On the contrary, I found one that allowed me to set all the clips to be at the same volume level. This gave me somewhat of a head start to sorting out these clips and of course saved so much time. 

“Loudness adjustment” is the feature that I used. 

From here I finished with the audio, splitting, extracting, and deleting segments from the interview. To detect which clips would stay in the interview format or have the audio extracted I looked at how the interview looked and which was best suited visually as well as which parts of the script had Broll that was a graphic match. I followed the original storyboard and script loosely but inevitably had to make some changes to both. 

I don't want to spoil too much so I’ll keep it all a secret what I changed specifically. You’ll just have to wait for the official movie opening drop on March 15th, 2024 7:00 pm EST! 

Back to the sound editing, I had forgotten to add the TV reporter's voice to the first 30 seconds of the film. Therefore I had to add it in. The only issue: I had no idea how I was going to make it. News reporters have a specific sound to them due to their enhanced diction and authoritative reading. I recalled some of the voice training I did in my actor’s workshop earlier but knew it would sound odd with my voice. But then I recalled how I used royalty-free sound effects for the sirens, only if I could find an AI voice generator. Of course, a quick Google search, and I had access to so many. I sorted through them and I found a play. ht which had many voice actors to choose from. I ended up choosing to make the news headlines using “Larry (Narrative), Sarge, Ayla (Advertising), and Charlotte (Training)”. I wasn’t too sure how to navigate this website, but I saw there was an area to insert the script and it would create the audio. I tried it out for “Sarge” and it worked but you could tell it was AI-generated and it was also extremely slow. That’s when I noticed there was a feature to change the speed so I set the voice to be at 1.6x which created a faster voice that better fit the fast-paced narrative. I created the rest of the headlines and inserted them in the beginning. To be honest I wasn’t sure what I thought of these bits since it was so clearly “fake”. But, I was left without any other options so I just had to make due. 

All in all, after all the sound mixing I had a combination of synchronous and nonsynchronous sound that meshed well! Once I fix the volume of the song, River Waltz, to sound like a score, it should be all good to go!

Visuals 

Alright, onto the visuals. From here I had the audio set up and the matching B-roll in place which meant a large part of editing now was getting the time down to 1:57-2:00. I had to make a few editorial changes but throughout I was focused on lowering the time. 

I started with lowering the time down to around 3:00 and after some editing, I was drained but I got the time down to 2:41. This is when compromises had to be made. I had to cut out some B-rolls and A-rolls. Essentially, I deviated complexly from the storyboard and focused on the video in front of me from this point onwards. 

I knew I wanted to keep the detective getting ready since it was a prominent genre element. The issue with this clip is that it was extremely long. I decided that I could speed up the clip so it better fits the time allotted. I did the same thing in the music video and it worked perfectly so I implemented it here. However, it looked somewhat unnatural if all the attention was on this clip. For now, I am going to keep it but in an ideal situation, I would refilm or have another take to substitute in.  These weren’t options so I was left to just use the sped-up clip. 

Additionally, I filmed some segments in a governmental building in a wide medium shot and a slighter over-the-shoulder shot so then I could do a shot-reverse-shot later in editing. I did that here. 

After a lot of compromises like these, I got the time down to 2:05. This was perfect because I was yet to add in the “case files”. If you can recall, I had spoken about scanning in a couple of documents from the crime board to add to the film. 

Originally I was going to follow the technique used in claymation where they take photos of every slight move. In my case, that would be the addition of a paper from the case. After running through this process in my head I knew it was going to look unprofessional so I decided to do this in CapCut instead. After playing around I decided to start on a gray screen which then would have an overlay of the papers. The succession of papers would be delayed so it looked like it was piling up like in a real case. To get the desired effect I applied the animation “Zoom I” at 0.5 seconds. I like how this ended up looking. It made me feel quite proud of myself for figuring out this editing technique since it is somewhat unconventional in film as a whole. 

The image above shows the setting to have the animation set on and then image below shows how the overlay should be stratified. 

I went back and made a couple of small alterations and got the time down to official 2:00. Which means… the bulk of editing is done! As you can tell the past couple of days have been all about deviance while still adhering to general guidelines. The only real thing left for editing is the titles. This means I’ll have to go back to my typography and the editing quest continues!