How I Edited My NAB Show New York 2024 Interviews

I’m Kayle asked: “Love to hear more about your post production process. FCP or Resolve? Did you record to camera SSD or to the MacBook or both? Software used during the recording and processing. Noise reduction? How did you name and manage your files for all the interviews? Geeky stuff like that.

All good questions. Here’s what I did.

NOTE: This article details the gear I used, what worked and what didn’t.

PRODUCTION

I recorded video on each camera using these SDXC cards. While their capacity was 256 GB, I only use about 90 GB for the two days of recording.

NOTE: I’m still leery of using SanDisk due to Western Digital’s manufacturing problems.

I recorded at the highest image quality supported by the camera: AVCHD, 1920 x 1080 at 59.94 FPS. (I didn’t need a frame that that high, but that was the only choice for this format provided by the cameras.)

AVCHD stores files in multiple folders during recording, so I didn’t change any file names on the card itself. Since both days shooting were stored in a single folder, I copied the “private” folder from the card to my editing system. Camera 1 went in one folder, Camera 2 went in another.

Because the Canon XA35 cameras could not record external audio, all audio was recorded on my MacBook Pro using Adobe Audition. The screen shot, above, was the original recording of the Blackmagic Design interview.

I used Audition’s Template function to create a recording template so that all interviews were formatted the same:

• Larry on Track 1 and labeled
• Guest on Track 2 and labeled
• All record inputs set
• All track effects added

Given the extremely short time between the end of one interview and the start of the next, using this template I could save and close the old interview, create a new interview and start recording in less than a minute.

All interviews were recorded in a single take. This became important during editing because I only needed to sync one clip for each speaker.

Because the cameras could not record audio, Bill Rabkin, our resourceful production factotum, had the brilliant idea of hand-slating each interview. (I was too frazzled to think clearly.) Without these hand-slates syncing audio to video would have taken forever. Maybe longer. THANK YOU BILL!

Fill lighting was provided by the two Lume Cube units. Trade show lighting is generically ugly, but these units did a nice job smoothing things out.

POST-PRODUCTION

All editing was done in Adobe Premiere Pro. There were several reasons for this – all situation specific.

1. I knew that my audio would be tinny, so I needed significant audio processing. Final Cut’s audio is clip-based. With dozens of edits in each interview, making changes clip by clip was a non-starter. While I could apply Roles to create the essence of a track, I decided to use actual tracks and save time. I seriously considered editing this in Resolve, but there’s a bug in Fairlight which makes detailed audio editing impossible. I’m hoping Blackmagic fixes this problem soon.
2. To speed finding highlights to quote, I wanted to use speech-to-text.
3. In case I needed more audio editing power, I needed a video editor that could easily move files into Audition. While, in fact, I didn’t need Audition, it was one of my considerations.
4. The only trade-off is that Premier is not as fast, or as stable, as either Final Cut or Resolve. For example, during every edit, the audio would just disappear for a few seconds. And exports took a bit longer. But I could live with these issues.

Here’s what Premiere looked like when all editing was complete.

NOTE: All sequences were shot and edited at 1920 x 1080 / 59.94 fps. However, the finished videos were exported at 29.97 fps to reduce file size. Cutting the frame rate in half made no difference to the image quality of the interview.

Premiere handles AVCHD files with ease. I imported each camera into its own bin. Clips appeared automatically. Since both cameras recorded the same interview at the same time, the clip numbers matched between cameras. So, I only need to label the guest. Finding my shot was simple – find the same clip number.

Sequences were numbered in the order in which I created them. The trailing number refers to the source clip of the interview guest.

I had two close-up angles to work with: me and the guest. Budget constraints prevented me from having a camera operator, so the framing of these shots was not as good as I would have liked.

However, in only one case was I forced to lower the sequence frame size from 1920 x 1080 to 1600 x 900 in order to frame a shot acceptably.

Here’s a portion of the Colin Rocker interview:

• V3: Titles and opening animation
• V2: Larry
• V1: Guest (Colin)
• A1: Guest audio
• A2: Larry audio
• A3: Opening animation music

Because I needed access to both tracks of audio and lots of cutaways, I decided against editing this using multicam. Instead, I kept the guest video enabled all the time, but disabled my video whenever I wanted to see the guest.

NOTE: I created two custom keyboard shortcuts: one to cut the selected track at the position of the playhead and the other to toggle between enabling and disabling a clip. I used both constantly.

Clips were synced on Bill’s hand-clap. While time-consuming, the process worked great. Editing was the same as any other two-camera shoot. The real problem was audio.

AUDIO EDIT & MIX

Unlike Audition, Premiere does not support audio buses, which would have helped a LOT in this mix. Consequently, I needed to add more audio effects than I would have preferred.

This mix had a lot of challenges:

• To avoid audio pops, we needed to position the mics under the guest’s chin. This created a very tinny sound.
• Trade shows are noisy places. I needed to reduce noise to make sure guests were intelligible.
• Audio levels, though loud enough, were inconsistent and needed smoothing.
• My mic was much softer than the guest; which surprised me. Most of the time, it’s the other way round.
• And, due to time pressures, I didn’t have a whole lot of time to polish each interview.

In the edit, I didn’t use any of the audio filters applied during recording. I simply copied the recorded clip into the new sequence and started the audio mix fresh with each edit. Because it was a single clip, once I synced it, all the audio from that speaker was synced for the entire interview.

Here’s what the track mixer looked like for every interview.

• DeNoise – to reduce, but not remove, expo floor noise
• 10-band Graphic EQ – to add back the missing bass. I didn’t need a lot of bands to do this
• Multiband Compressor – to boost and even out levels
• Hard limiter – to prevent any peaks causing distortion
• Loudness meter – to monitor average audio levels.

NOTE: Social media likes audio levels around – 15 LKFS. This is easy to monitor in Premiere and Audition.

All I did with DeNoise was select the Light Noise Reduction preset. I left all other settings unchanged.

I added a 10-band Graphic EQ, then significantly boosted the low end (red bracket).

NOTE: For women, I used the same settings, just one slider up (250, 500, 1K)

Multiband Compressor had four changes:

• Preset = Broadcast
• Gain = +2 dB
• Margin = -4 dB
• Brickwall Limiter = On

For general dialog, I disable the Brickwall Limiter. But, in this noisy environment, adding this feature made sure that all voices could be heard easily.

Hard limiter was placed on the main output and guarded against excessive audio levels. I changed the default from -0.1 dB to -2 dB, because that extra headroom wouldn’t hurt anything and provided additional protection.

For the Loudness meter, I simply watched the Integrated level. As long as it was within shouting distance of -15 LUFS, I was happy. In the case of all interviews, I used this for reassurance.

SUMMARY

Every edit, every mix, is different. The audio EQ settings in this project are larger than I would typically use, but they solved the problem without causing distortion.

Syncing clips was not hard, but took time. Without the hand-claps, though, I’d still be editing.

While nothing about this project was “perfect,” perfection wasn’t my goal. I had guests with interesting and important things to say that I thought you would like to hear.

As always, we do the best we can with what we’ve got. And now you know how I did it.

---

Bookmark the permalink.