Whenever a youtuber captures slow motion video, the audio is usually faked using something that sounds like it fits the scene, even if its unrelated. (e.g. crushing tomatoes into a microphone for a squish sound for a slow motion of water balloons)

Is there any way to capture the true audio from the moment and maybe tweak the frequencies into the audible spectrum? Is there any example of someone doing this?

  • the slowmotion videos I've seen are all accompanied by the original audio, slowed down to the same extent as the video, not by sound effects. – Hobbes Jan 3 at 8:48
  • @Hobbes I'm referring to the extreme slow motion videos, 1000fps or higher. SmarterEveryDay did a video about how they fake sound effects that sound realistic. – Daffy Jan 3 at 11:11

Videographers never actually capture video in slow-motion. All video is captured in real time. What differs with "slow-motion" is that a higher frame rate is used to capture the vision so that more frames are available in post to generate the final vision.

The reason for this is that for smooth motion, you need a 50%/180 degree shutter angle at 25 frames per second (PAL) - and to get that slow-motion, you need to play out fewer frames per second than you actually captured.

For example, for 1/4x speed, you will capture (in real time) at 100FPS so that when playing out at 25FPS you only see 1/4 of the action for every second. Also, for 4 seconds of video (after processing) you will only need to capture 1 second of real-time vision.

A similar technique can be applied to audio. If you want to hear full-range audio from a 'slow-mo' capture, you need to capture at a higher sampling rate. Full-range audio will require an output sampling rate of 48kHz (to allow 20Hz-20KHz spectral range).

Therefore, if you are capturing vision for playback at 1/4 speed, you should be looking to have equipment that will sample at 192kHz. This will allow full-range playback at 48kHz when the vision has been slowed.

But, when you have slowed down that audio from 192kHz it will be 1/4 of the original pitch.

Now, there will be a reasonable argument made at this point that nothing you will capture above 24KHz will be useful, however this isn't entirely correct as possibly inaudible sonic components can be brought into the audible range by this process. Depends entirely on what you are working with.

It is perfectly possible to re-pitch this slowed down audio by increasing the relative pitch by two octaves (in this example) but you will definitely notice pitching artifacts as a result of this. It is do-able, but it won't sound great. The higher the sampling-rate the better your result will be.

The reason why a lot of these 'youtubers' will design the sound directly around a 'slow-mo' video is to achieve greater control over the story and to reduce the possibility of artifacts.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.