How Background Music Shapes What We See
INFO 303 . Major Project 2 . Joe Zhou
About
In Major Project 1 I collected other people’s examples of background music changing what a video seems to mean. For Major Project 2 I made four artifacts of my own. Each one changes a single thing and keeps everything else the same.
Artifact 1 keeps the picture and swaps the music.
Artifact 2 removes motion and leaves only a photo and a sound.
Artifact 3 keeps the music and asks what happens once you already know it.
Artifact 4 takes performances that were built around one song and puts a different song underneath.
Same Video, Different BGM
What changes when only the music changes?
- Form
- Video edit — the same clip re-scored. Two clips, total five versions.
- Audience
- People who scroll animal videos on Douyin or TikTok.
- Tone
- Cute → siren. Eerie → sunny → nostalgic.
Across the two groups the same footage moves from cute to evil, and from horror to sunny to nostalgic. The genre is supplied entirely by the music.
Group 1 — The Seal
One continuous 22-second take: a seal surfaces beside a dock, stares into the lens, and then turns away.
Nothing in the picture changes between the two versions below, only the soundtrack.
Version 1 — Original: a light, steady beat
The music is flat on purpose. It starts at full volume on the first frame and never builds or drops. Because nothing in the sound is marked as “the moment,” viewers watch the seal as a whole. The words people reach for are “curious,” or “a sunny afternoon.”
Version 2 — Siren: a whispering choir in waves
The replacement track breathes in waves, and the waves line up with the footage. The same stare now reads as a little bit horrible. Viewers start noticing the darkness of the water and the fact that there are two of them. And the most, the staring eyes!
The low, whispering sound of BGM is what makes it a siren rather than a monster.
The first track has no shape, so the clip’s atmosphere is continuous. The second track has a shape (which is the woman’s sound), so the clip has a plot. The music decided the vibe of the video.
Group 2 — The Empty Swing
Content: A tire swing on a metal frame rocks back and forth with nobody on it. Bright daylight, wood chips, autumn trees, suburban houses. Only an empty swing moving by itself. What sign it might include? A horror cliché, a childhood memory, or just a windy afternoon? The three BGM below each pick one. The original video has no background music, so I added three different BGMs.
Version 3 — Eerie
Imagining a horror movie, the daylight stops being reassuring. The horror is in the gap between the picture and the sound. Something invisible is playing with this BGM.
Version 4 — Sunny
The music is about sunny days, clear sky, touching memories. This could be a swing in your childhood.
Version 5 — Nostalgic
The swing starts to look like it is keeping time with the music. The muffled, old-record tone turns the scene into something already in the past
Three moods from one shot. Horror, youth and nostalgia are not in the footage. Each soundtrack tells the viewer who was on the swing, and the picture goes along with it.
Why I made it this way
- A single clip isolates one variable, which is the music. Fair comparison.
- Animal-video viewers already have strong expectations about sound: it might be cute. But it’s not.
- Every version is the same length and cut at the same frame, making sure only the audio differs.
- Original audios come first, so viewers have a baseline before the swaps.
Photo + BGM: What Did You Notice?
Photos of my own cats, just image first and then with a soundtrack.
- Form
- Non-video visual — a still photo, then the same photo with BGM.
- Audience
- Anyone who love pets, even someone just loves funny content.
- Tone
- Comedy (Pair 1)
Anime-energy (Pair 2).
About
Artifact 2 is the “static” experiment, so each piece is a still photo, shown twice: first silent, then with a BGM. Technically it plays as a video file, but I count it as a non-video visual. The picture never changes, so whatever changes in your head has to be coming from the sound.
The question I was chasing: when you add music to a photo, does it reinforce something that was already in the frame, or does it redirect you to something you wouldn’t have noticed on your own? Pair 1 turned out to be a redirect. Pair 2 turned out to be a reinforce.
Origin
The idea started with my habit: taking a photo of cat sending it with a song/BGM attached. And most frequently, I love match interesting audio with pictures. So it’s time to think about why the song makes the photo funnier. Then I noticed that a photo of two kittens with an anime opening behind it stopped looking like two kittens and started looking like a cast, and that was the whole artifact.
The feasibility
One artifact has to be a non-video visual, so here I did not add any “video element”, and music’s “on/off” is the only variable, which is exactly what the requirement is about. It also covers the “different tone” requirement from a side angle.
Pair 1 — Eggy in the pot
My cat (I called him Eggy in English), sitting in a white pot, caught mid-effort.
first see the photo, then with sound.
Silent: a grumpy-looking cat with narrowed eyes.
With sound: the same photo under a slowed-down hum.
If you notice: the BGM is actually the intro of “Please Don’t Go” (by Joel Adams), slowed down until the “Umm Umm” becomes a low, continuous groan. BGM gives the picture a voice, and the voice tells you what the audience should focus: the face expression, showing this cat is taking a huge effort in toilet!
In this case, BGM did not create something new, it just amplified the element in the photo
Pair 2 — Two overlapped kittens
Two kittens were afraid, and they just overlapped! Must be something startled them. Three photos at all, shown silent first, then dynamic with sound.
Two frightened kittens. The arched backs and raised tails read as fear, and the first thing we might think is “what scared them?”
The soundtrack opens with an upbeat anime-girl voice line, and then an exciting electronic beat drops in. At that moment, two scared kittens become a duo.
This is the other outcome: Redirection to something new. The music picks one feature that in audience’s mind (no matter what it is). Thought matched the picture, prompting the whole meaning.
Reminds you of…
Tom & Jerry
Rem & Ram (Re:Zero)
Asuka & Rei (Evangelion)
Trademark / logo
Why I made it this way
- A still photo, not a clip, so any thought the audience feels can only because of BGM.
- Pair 1 uses a sound with no dynamics (continuous “Umm”), so the effect comes from timbre and duration rather than a drop. With no BGM, audience may only see the whole picture; but with BGM, they will focus on the cat’s face expression.
- Pair 2 uses a sound with a clear two-part structure, so the audience can feel the exact moment of the expansion of image.
- Shown static photo first, so viewers commit to a first reading before the music arrives and can feel it change.
Expectation essay: The “Indian Cough Syrup”
When we are already familiar with the BGM, what expectations do we have for it? What if it changes?
The “Indian cough syrup” (#obhcombisachet) is a short clip of a song that nobody can name, but probably everybody can recognise: the coughs are the core. An intro phrase ends on a drumbeat, then the vocals come in, then the cough. The original trend was a hand dance: you perform the gestures and bring a fist to your mouth on each cough. I choose six clips to show where the trend went. video 1-2 are the standard version. 3-4 keep the sound and change the subject. 5-6 change the sound itself. The artifact is a whole essay of how a person that never touch this trend before recognise this BGM, and how the expectation of this BGM built and changed.
- Form
- Informal personal narrative essay, with six reference clips.
- Audience
- People who might ever know this trend.
- Tone
- First person and reflective.
Reference clips
Six short clips the essay refers to. Covers link to the videos.
Clip 1 — The standard hand dance
One form of the dance
Clip 2 — Same dance
Another form of the dance
Clip 3 — Same sound, now it is a cat
Same dance, but a cat
Clip 4 — Same sound, now it is a cat
same dance, but a cat (I didn’t find a dog)
Clip 5 — The intro on loop
One variant: The intro repeats
Clip 6 — intro variant
Another variant: endless beat
The “Cough”s are Coming!
I am not a “hand-dance (✋💃️✋)️ person”. I know them, but usually scroll past, but not this one.
The first time I stopped for this trend was because of the music. I was not interested in the dance. The sound was better than I expected, so I stayed. The cough is the core of it. I did not understand a single word of the song, but I still knew right away that this was a video about coughing. That is the strange part, because the BGM itself does not need to be understood, as long as you could recognise it.
The trend did what trends do, like every hand dance looked the same. Perhaps the intro part changed a little from one person to the next, but the core part (which is the cough part) never changed. After a few videos I knew the whole clip before it started. So once I could predict it, I stopped caring about that, and suddenly lost interest in about a day.😭
Then the videos changed and the music did not, person was not in the video. In their place was a cat. When I first scrolled to it, I thought it was an ordinary pet video. However I knew what happen when the intro started, and still, I knew before anything happened: someone was about to do the hand dance on their cat. The fist would come up to the cat’s mouth on the cough, and It did.😂
What had changed was who the video was for. The first wave was people filming themselves, but this wave was people filming their pets. The subject could not have been more different, and my guess was still right. I noticed that the music had become the content. The image was only the place where the music happened. As long as I knew the sound, I could predict the video no matter what was in it.
The last change was the one I did not see coming. The music itself changed. One video loops the intro again and again: the vocals never arrive, the cough never arrives either. Another video keeps the first intro and then repeats its last two beats “Doon Doon” until the end. On screen a cat gets a soft fist on the head with every thump. ✊️🐱
This was quite new! The cough was supposed to be the point of this song, but these videos cut the cough out. The thing people remember is now the “Boon” at the end of the intro. I knew this sound as well as anyone, but I was still surprised. Both of them are funny, and I believe the rule breaking of the original music is the core.
Looking back the order makes sense: I stopped because the music was good; I left because I could predict it: I came back because it became unpredictable again. What the image of the video might be a reason, but far less important. Knowing the BGM meant I was not really watching the videos, but checking them against the song. As a result, the videos that got a reaction out of me were those failed the check on purpose.
The brief
I chose to write an essay for this artifact because what I wanted to explore takes a bunch of time. One clip can create a mood, but it’s difficult to show how an expectation builds and what happens when that expectation is broken. The six clips are examples, and the essay is the final artifact. This one looks at how Background Music shapes the viewer’s experience (expectation).
BGM Deconstructs the Contents
Rustic performances, with sound removed and different song added.
All of the material here comes from Douyin. The originals are in Chinese and the songs in the edits are in three different languages. That is on purpose: If the effect still works for someone who understands none of the words, then the meaning is coming from the BGM’s melody rather than lyrics.
I will introduce some members that you might see on the edits. These are all original video.
The edits are made by strangers who took the originals, cut the sound off, and put the performers next to each other under a completely different song.
- Form
- Remix: original performance videos with their sound replaced.
- Audience
- Any Viewers, without language basics.
- Tone
- Fierce melodrama → comedy.
The originals
Original 1 — Grandma and the red guitar
A grandmother in an old house. The word she said at very beginning is “Rock you”, it is the loudest moment of the clip. Then she plays along with a BGM. Plus, the guitar is not plugged into anything you can hear. The attitude is the essence!
Original 2 — The bucket drummer
An aunty in a red apron on a red plastic chair. There is a BGM, but the hit sounds are hers. She holds the beat for the whole clip without dropping it. Of the three originals this is the only one that is actually good at the thing it is doing.
Original 3 — A bowl of forgetting
A woman lip-syncs a heartbreak ballad with a bottle in her hand as a prop. Red face, wet hair, mouth wide open on every line (That looks very heart-broken!😭). The subtitles are lyrics, it doesn’t matter.
( But it’s about ask the shopkeeper for a bowl of soup that makes you forget if you are curious!)
What they have in common
All three are one person, performing for a phone at their home. All three were posted for fun and became popular anyway. And in all three the atmosphere is shaped by the music they chose.
From strangers to a band!🎸
- First, they removed the original sound.
- Second, they put two or three performers side by side in a split screen. The people never met, they live in different cities and filmed in different years.
- Third, they gave each person a “job”, and the “job” was decided only by what their body was doing. Someone hitting things is the drummer, someone holding a guitar is the guitarist, and someone with their mouth open is the singer. That is the entire casting process.👏👏👏
- Fourth, they put a new song underneath. It could be a song from a different language, a different speed and even a different mood.
- Fifth, in the longer edits, they cut between members when the song changes section.
The result reads as a band (Although it should not). The drummer is keeping a beat that does not exist in the new song; the singer is mouthing words from a different language. But the real thing you notice is: when the mouth opens on a lyric, you hear her singing it, and when a stick comes down on a beat, you hear her playing it. The timing does the work and the mismatch is what makes it funny. The originals were rough and sincere, and the new song is polished. Put together they look like a group that has been rehearsing for years, and viewers gave the group a name.
That is the deconstruction I wanted to show. The crying in the original only made sense because of the sad song under it. Remove the song and the crying becomes a free-floating “intense singing” that fits anything, and I really love that feeling! On a calm song it is absurd, but on a sad song it almost fits again.
The edits
Edit 1 — A calm song
We have drum, guitar and vocal! hitting bowls and bottles on a table on the left, crying and singing in the middle, play an electric guitar on the right. The song is light and city-sounding, and the lyrics on screen say that “nothing happened in today’s news and that the sunlight is warming breakfast”. The most peaceful lyrics in the project are being “performed” by the most distressed (or touched?) face. This one is the clearest case of the mismatch doing the comedy.
Edit 2 — A sad song
The bucket drummer on both sides, and a crying man in the middle. Because the song is sad, the sobbing face almost fits here. When there are too many mismatches, an unexpected fit can also be very interesting.
Edit 3 — A fast animate song
A very funny one! Using the genre of ボカロ(bokaro). This song is a sad but fierce song, and the crying woman a good choice of vocal. And this one has a very amazing edit: The intro is properly well-fitted. By speeding the elder’s playing, they really match the powerful prelude. And Bokaro is famous in teenagers, but these “band members” are old! Unlike before, this time the contrast in the visuals is largely reflected in "age" or "era". And it is the deconstruction of the BGM that gives this visual contrast its effect.
Why I made it this way
- I used Douyin originals, so we can see what strangers did to them.
- These edits are made in three languages, so the lyrics will not be the explanation.
- Three originals before three edits, it’s enough to clearly feel how the original videos perform.
A little difference
In artifact 1, the video itself was neutral and the music gave it a mood.
By contrast, the content here was highly impacted by its own music.
Conclusion
Artifact 1: same image → new sound → new meaning. The music decided where the event was.
Artifact 2: no motion, so the sound supplied the whole meaning.
Artifact 3: a sound I knew well predicted the video, but variant BGM could break the expectation.
Artifact 4: the new song overrode performances and meanings.
To sum up, all the four artifacts discuss the same thing from four angles. Background music is part of how the meaning gets made.
For viewers: notice when a sound is doing your interpreting for you, especially on platforms where the same sound is reused thousands of times.
And for creators: the music choice is an editorial choice, a wise choice can really contribute to the achievement.
Of course, these are just personal experiments, so it only offers a way of noticing, and you can try it on your own video edit.
References
(Artifact 1)
[1] 馨. (2026, October 14). 配上幽灵🎵 就不那么可爱了哈 #海豹 #幽灵音乐 [Video]. Douyin. https://v.douyin.com/oJq_yUjDF1U/
[2] JMT 35. (2025, October 30). Empty playground swing in suburban park [Video]. Pexels. https://www.pexels.com/video/empty-playground-swing-in-suburban-park-34507817/
(Artifact 2)
[3] @601. (2026, June 29). 只能使出那一招了吗 [Photo slideshow]. Douyin. https://v.douyin.com/zW79nh94h3A/
(Artifact 3)
[4] 정빈 [@binibeen_]. (2026, September 13). 아 벌레 [Video]. TikTok. https://www.tiktok.com/@binibeen_/video/7684958640423456007
[5] Kimnana [@angelinawj]. (2026, June 28). obh combi sachet lagi aaah🤷🏻♀️ #fyp #tutorialdance #dancetutorial #obhcombisachet #obhcombi [Video]. TikTok. https://www.tiktok.com/@angelinawj/video/7656393629912567048
[6] 招财喵🐱. (2026, September 25). 呆呆的财 #印度咳嗽 #指猫挑战 [Video]. Douyin. https://v.douyin.com/Nx5hdCua2Ds/
[7] 哪儿拿. (2026, September 27). 做最冷静的一只咪 #印尼咳嗽药 #正经小猫 [Video]. Douyin. https://v.douyin.com/u25YOO1w5tM/
[8] 瑜伽鱼（见过鹿晗版）. (2026, September 26). 排队领锤 人人都有，一人两锤，重在参与哈哈哈哈哈 [Video]. Douyin. https://v.douyin.com/gR_NqDciQjg/
[9] 爆炸的橘子🍊. (2026, October 2). 停停停停停停停🤚 #捶小猫手势舞 [Video]. Douyin. https://v.douyin.com/AxxuVuLfbMA/
(Artifact 4)
[10] 发神经理. (2025, June 30). 奶奶说的是rock U哦！ #摇滚奶奶申请出战 [Video]. Douyin. https://v.douyin.com/9WaQxgKt1Cw/
[11] 贪玩的小芳. (2021, January 25). 这个音乐让我过百万粉丝，你们还在等什么？ [Video]. Douyin. https://v.douyin.com/-Q-cIP_bW2g/
[12] 甜甜妹. (2026, May 3). 老板给我来碗面～什么面？《忘情牛肉面……！ [Video]. Douyin. https://v.douyin.com/91T9k7_3t_s/
[13] 灰曲1. (2026, August 2). 《晨间新闻》 有的时候 没有新闻 就是好新闻 [Video]. Douyin. https://v.douyin.com/ipjS3hQnSa4/
[14] popo搞泥. (2026, August 14). 老牧师乐队现场唱《Unhappy》，氛围真的好伤感 [Video]. Douyin. https://v.douyin.com/JQPRIULlF64/
[15] 喧嚣. (2026, August 1). #术力口 #老牧师乐队 #少女A [Video]. Douyin. https://v.douyin.com/zFKXMbwJl9E/