2 April 2020 by Lars

Post-editing of a Zoom recording

Warning: academic / geeky / Cornell-only sections ahead

I recently recorded a Zoom meeting/ webinar. How well that went is a different story. Here’s a commented list of tasks I did for post-processing it into a video to be posted online.

Tools

Zoom 4.6.7 (Windows) for recording
ScreenFlow 7.3.1 on MacOS 10.14.6 for editing
Cornell’s Video-on-demand service (Kaltura)

Comments

ScreenFlow is not per-se a video-editing tool, but meant as a screen recording tool. The version I used is not the latest. I just happen to have had it. It had some issues (black screen). Mostly, it had the feature I needed: adding overlays and copying/ cutting in video and stills.

Steps taken

Recording

The video was recorded with Zoom’s cloud recording. Once the video was recorded (I stopped before the Q&A), it got posted to the (private) cloud. Cornell’s subscription has a 120 day deletion policy, but also automatically moves the video into Kaltura (as a private video).

Kaltura does not allow to edit the video other than to cut out pieces at the start and the end (as far as I can tell)
Kaltura does do automatic transcription
I have not figured out how to download a video from Kaltura…

Downloading

I downloaded the video from Zoom’s cloud storage, and stored a local copy.
(Cornell-only) I also downloaded the “Video Bug” from Cornell’s brand page.
(Me-only) I created a visual “CV” page, in Powerpoint, and exported it as a PNG. (note to self: that picture was taken 6 years ago…)

Editing

I edited out the start (recording of me setting up).
I also overlaid a clean PNG of my slide, rather than the recording of the first slide. That means that the picture-in-picture (PIP) of me fades in about 45s into the talk. Your preferences may vary.
Within the first 2-3 minutes, I edited out the “ums” and “ahs”. I need to a while to warm up. But I do speak slowly at first, with an attempt to have distinct pauses – makes it easy to edit this out. YMMV.
- ScreenFlow tip: i starts marking a section, o ends the section, ⌘-delete cuts that section out, and moves the rest up.
I then ignored the rest, up through the end of the talk.
I did a bit of editing at the end, taking out the verbal setup of the Q&A (not relevant if viewing the recording), and most everything after I show the final slide.
I then added back in a PNG of the “Thank you slide” (with the DOI of the slides), and leave that on screen for 2s.
I then added the CV PNG, and leave that on-screen for 2s.
I then added the Cornell Video Bug (logo, fade out)
- Based on the counsel of an in-house Youtube consultant with vast experience, the Cornell Video Bug is too long (“boring, Dad!”). I therefore shortened the time it stays on screen substantially.
- i , o, ⌘-delete
Finally, I added a text field for copyright, and added the license logo for the CC-BY-NC license that I want to apply to this video.
Rendered the whole thing for 1920×1080 video.

Here’s the end sequence (as a GIF):

Uploading

I uploaded to both Cornell’s VOD and Youtube

Kaltura (VOD) only has a free-form field for license. I entered “CC-BY-NC”
Youtube only offers Youtube license and base CC-BY. I chose “Youtube”.
In both cases, I added the abstract and as much metadata as each platform allowed.