ari_paper

The goal of ari_paper is to showcase the ari package. We show how to use each function and the output videos. This README uses the video package to render videos inside the markdown. The paper can be viewed at https://johnmuschelli.com/ari_paper/.

Making videos with ari: ari_stitch

library(tuneR)
library(ari)
result = ari_stitch(
  ari_example(c("mab1.png", "mab2.png")),
  list(noise(), noise()),
  output = "noise.mp4")
isTRUE(result)
attributes(result)$outfile

Creating Speech from Text: ari_spin

Let us take a part of Mercutio’s speech from Shakespeare’s Romeo and Juliet [@shakespeare2003romeo] and overlay it on 2 images from the Wikipedia page about Mercutio (https://en.wikipedia.org/wiki/Mercutio):

res = ffmpeg_audio_codecs()
fdk_enabled = grepl("fdk", res[ res$codec == "aac", "codec_name"])
if (fdk_enabled) {
  audio_codec = "libfdk_aac"
} else {
  audio_codec = "aac"
}
speech =  c(
  "I will now perform part of Mercutio's speech from Shakespeare's Romeo and Juliet.", 
  "O, then, I see Queen Mab hath been with you.
   She is the fairies' midwife, and she comes
   In shape no bigger than an agate-stone
   On the fore-finger of an alderman,
   Drawn with a team of little atomies
   Athwart men's noses as they lies asleep;")
mercutio_file = "death_of_mercutio.png"
mercutio_file2 = "mercutio_actor.png"

output = "romeo.mp4"
if (!file.exists(output)) {
  run_voice = "Joanna"
  ari_spin(
    c(mercutio_file, mercutio_file2),
    speech, output = output, voice = run_voice,
    service = "amazon",
    audio_codec = audio_codec)
}

We can also generate the video using the voice Brian, which is an British English male voice: