I have heard it said many times that bad video footage can be ignored better than bad sound can.

When watching film I have experienced this to be an accurate statement.

Does anyone know of any science, resources or theories on why this might be so?

Thank you

  • There can't possibly be a rigid scientific underpinning to this, because there's no meaningful metric to compare the badness of audio with the badness of video. Jul 20 '15 at 6:30
  • Shouldn't the title be 'bad sound vs bad video' ?
    – audionuma
    Jul 20 '15 at 6:44
  • OK so no rigid science, how about subjective ideas on the subject?
    – Filmduck
    Jul 20 '15 at 7:19
  • this question is a bit broad, can you make it a bit more specific with examples? this will narrow down the scope of the question and make it easier to answer it. Jul 20 '15 at 8:13
  • 1
    @Filmduck - may I suggest you clarify what you mean by "bad audio". It seems there is some confusion as to if you are referring to low fidelity or actual problems with the audio. The context of the phrase you are referring to generally refers to actual problems rather than fidelity, but I don't want to alter your meaning if your impression was that it was more fidelity focused.
    – AJ Henderson
    Jul 20 '15 at 17:28

I refer you to this study by MIT and this study by Cleveland State University.


It's pretty simple actually, though it depends on the specific issue. The main thing is a combination of naturally having skills to deal with bad video and the extra information video provides.

If video is a little bit shaky or the color is a bit off or the angle is bad, these are all things that we deal with all the time in real life. Your eyes are always dealing with your head moving around. You are always adjusting for changes in lighting color, often even within one environment. You don't always see things from an ideal angle. You have a lot of practice naturally dealing with many of the things that make video "bad".

Similarly, if detail is missing from video, it's pretty easy for our brains to adjust for detail loss from compression artifacts and such because we are good at matching patterns of what we see and filling in the details. Really, this is what persistance of vision and our ability to move our eyes around without it being a hugely jarring experience does all the time (not to mention hiding the giant blind spots where our nerves enter the back of our eyes).

On the other hand, for audio, we have drastically better ability to perceive sound than we have to play it back. We have quite good positional noise filtering capability, so we can normally make decent sense of a noisy situation, but when it is recorded and then played back through a few speakers, the noise and the audio we are interested in are coming from the same place. Our normal tricks for filtering don't work and so we struggle.

Additionally, our senses don't work completely independently of each other. We are very good at combining what we see and what we hear and thus it is jarring if sound doesn't match up with what we are seeing (such as with out of sync audio or bad foley).

Many of the issues that can plague audio simply are things we are either not well adapted for or are actually outright adapted against. For video, it is the other way around. Our natural adaptations aid us rather than work against us.


I am certain this claim stems from the early days of Youtube.

First, for video; Considering video clips with the quality of a 90s "America's Funniest Home Videos"-clip do sometimes go viral, clearly high video quality is not that important.

However, for audio, this used to be true in the early days of Youtube because many PC and Mac microphones and headsets created a lot of background noise. They still do, but the problem is not that bad anymore.

The problems with bad audio are specific; Noise, clicking, echoing, buzzing and the likes is what makes people want to stop watching a video with bad audio. Background noise was a huge problem in the early days of Youtube.

Edit: I want to add a bit more about why I think bad video does not bother people as much as bad audio.

Bad video simply is not as jarring as bad audio can be. While bad microphones can capture noise, bad video cameras do not record non-existing flashes of light in the background. Bad audio recording can lead to crackling and popping, but bad cameras do not randomly change the color of the face you are watching.

If cheap video cameras did that, it would turn people off as well.

  • I don't get your reference to MP3 as evidence of bad audio being an issue. MP3s are not "bad audio". You can have very clean, well timed and accurate audio in an MP3. It may be missing some subtler details from compression, but that doesn't make it "bad" audio in this kind of context, just lower fidelity. (And even that is arguable since many people can't tell the difference reliably.)
    – AJ Henderson
    Jul 20 '15 at 17:22
  • The point is that audio does not have to be of high quality for it to be bearable, but yeah, I will edit my answer to something that makes my point clearer.
    – Avatrin
    Jul 20 '15 at 17:24
  • ah, ok, I see where you are coming from. I've asked the OP to clarify their meaning in "high quality". Generally speaking, the saying to which he is refering does not mean fidelity, but rather presence of issues in the audio, such as poor capture, poor dynamics, poor timing, etc.
    – AJ Henderson
    Jul 20 '15 at 17:30

For visuals the brain needs 24 frames per second to believe that it is seeing moving image. For audio the needs 44,100 "frames" per second to be believe the audio.

  • 2
    Neither of those statements are true, and even if they were the sampling rate (as it were) is only one piece of our perception of each medium. Jul 20 '15 at 12:30
  • 1
    It's also worth pointing out, it doesn't take 24 samples per second for video, it takes almost 50 million if we assume your 24fps 1080p concept is correct (which it is not). Frames are not the analog of a sample, pixels are about the closest analog you can get and there are a little over 2 million of them per frame (for 1080p, even more if you move to UHD). Pixels and samples still aren't even close to a perfect analog, but it's far closer than comparing samples to frames.
    – AJ Henderson
    Jul 20 '15 at 17:33

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.