I kept seeing news report headlines that a speech by a leading Pakistani politician Shaukat Yousafzai had been passed through a ‘cat filter’. I was not sure what that was and, because I am an old out-of-touch fogey, had the vague impression that he had used a physical filter like a surgical mask designed to catch cat dander to prevent his allergies triggering. It was only when I saw this video that I learned what this cat filter did.
The filter had apparently been turned on by accident by Yousafzai’s social media team when they live-streamed the event on social media. Yousafzai has apparently taken the mistake with good humor, which is nice to hear in these days when some prominent people are so full of their sense of self-importance and so sensitive to their image that they react furiously against any aide who even inadvertently does something that makes them look foolish.
Incidentally, did anyone else feel that the voiceover for that video sounded like it was computer-generated? It had a curiously flat affect.
Yes, I think the voice was one of the standard options available to everyone using an Apple product such as a Mac. My guess is that using a computer generated voice protects the organization from the controversy of recording it with a human with an accent that is not from one specific region or another of their country. Everyone who hears it isn’t jealous of some other region being favored. And nobody can criticize the campaign for pronouncing anything imperfectly, even though everything is slightly off to everyone. So it’s seen as a safe choice.
It was definitely a computer read out.
My guess is that the computer generated voice is just to make the videos quicker to produce. No need to record and edit a voice over.
I looked at a few other videos from that same Youtuber and they all feature that same voice, same “Breaking News” intro and the newscaster lady (at 0:30 in this video) that I think is supposed be the one doing the voiceover. That channel seems to produce close to 20 videos a day. If its just one person there’s no time for quality
The speech-synthesised voice over is frighteningly common with the myriad viral-video automation channels. Scrape some video from a website somewhere, scrape a paragraph or two of text (no matter how tangentially it relates to the content), automatically combine the two, automatically publish, automagically make money. By scraping, I mean stealing. It’s the shit on the shoe of the Internet.
It all goes a bit meta when a real person makes a “top ten” video, and an semi-automated scraper comes along and slices it, combines with some other top ten videos, and republishes. It shows when someone else’s transitions and edits are abandoned half way through the automated edit, pixel aspect ratios changing throughout, frame sizes all over the place.
Obligatory XKCD: Digital Data