This song I based on a sample created by my dad in his early 20’s. He sent me this sample recently while he was archiving his old cassettes. I immediately had an emotional response to the sample, and I felt compelled to use it for a song, attempting to preserve the sample as much as possible in the production. 

My dad had started archiving old cassettes he had recorded on as a young adult, sending me these recordings. One night he sent me a “sound-letter”, as he called it, which he had made on his cassette loop machine. When listening to this, I was immediately struck with inspiration. I put the sample into Ableton and listened to it. Feeling the sample was to slow, I changed its warp function to Re-pitch, meaning that when changing the tempo of the project, the sample would pitch itself accordingly. I prefer doing this to ensure no stretching or warping occurs to the sample, leaving the sound file unaltered. 

I fine-tuned the tempo so that the sample got pitched up exactly two semitones. I did this by duplicating the sample and pitching it up manually, then tuning the sample to this duplicated sample. I then placed a clap on the second and fourth beat and a hi-hat, however, quite quickly I figured that I didn’t want a hi-hat, or any standard trap drums for that matter, so I removed the hi-hats and utilized convolution reverb with a snippet of the original sample in order to disguise the clap within the sample. 

I then played a bass line. I wanted the bass to be hard-hitting, as the sample hit me very hard emotionally. I utilized an 808 with quite a heavy impact and turned it very loud in the mix. I then started freestyling. Since the sample made me feel sad, I freestyled in a verry high pitch, almost out of my register, leading to my voice cracking a lot. During the freestyle, my emotions started shifting from sad to angry, and I started reflecting this in my flow and tone, adopting a much more aggressive, raspy and deep tone. I did this by lowering my larynx and changing how I used my nasal cavity, however, I am not quite entirely confident in describing how I am affecting my physical instrument, as I often just “do what sounds right”. I see this as an aspect of my functional tacit knowledge, which Borgdorff mentions as a key part of artistic research, investigating what one does not know (Borgdorff, 2012, p. 163). 

This affected the timbre of my voice, showing my current state of mind, also affecting the message conveyed through my voice. This is in line with Kari Iveland’s view on vocal timbre (Iveland, 2024, p. 104). The meat of the intro and hook came to me instantly when freestyling, quite quickly finding the theme “try to make me do something, I'm never gonna do that”. I took a break from writing and focused on mixing. I added heavy distortion and chorus to the initial freestyle and placed it far back in the mix, almost blurring the words, and I found myself hearing the words I wanted to hear. I then started dubbing a new cleaner vocal on top, one octave down in the register, switching between recording on the clean and the wet chain. I found this process enjoyable, and quite quickly I had written most of the song. This experimentation aligns with what Bordgorff states as an integral part of artistic research (Borgdorff, 2012, p. 165).

I had been thinking about this song for a few days before I started rewriting it. I was quite scared as the flow from the English version came from the words “try to make me do something”, pronounced “try da may me do something”, fitting very nicely into a triplet rhythm. The only line I had been thinking of was “prøv og få meg til å gjøre noe”, and I was afraid it would feel forced when rapped. However, after rapping it a few times and blending the words “gjøre noe” together, I felt it worked quite nicely. I started recording from the hook. The rewriting of the hook went quite smoothly as I rhymed with the same word “det”, and I bended the pronunciation of “deg” to sound more like “det”, allowing me to rhyme the last word with “yeah”. During this recording, I made sure to change my delivery to match the intensity that English version has in the hook. 

I then went back to the intro and started rewriting. I started with the line “he wanted to show me”. I chose to write this as “han vile vise meg”, allowing me to rhyme the next lines with “meg” also, like the English version. However, when I got to the line “you gotta live life”, I got stuck, finding my only option being “du burde leve livet”. This changed the rhythm of the flow a bit, adding one more syllable. I once again struggled on the next line “keep it all so tight”, deciding rather to refer to one’s waist, which in Norwegian is also spelled and pronounced as “livet”, landing on the line “hold det fast, tett til livet”, translating into “hold it, close to your waist”. I find this quite interesting as it adds another layer off imagery to the line that previously wasn’t there.

During the verse in the aggressively autotuned falsetto part, the intent was to sound unintelligible. This allowed me to be a bit freer with my writing. For example, the line “I ain’t never done nothing” got rewritten to “men jeg har ikke ikke tatt noe”, contain two negatives in order to convey the correct message. I managed to write a really interesting bar by accident in the last lines of the lyrics, when I rewrote the line “I mean no offense… fans” to “mente ikke å trakasere… seere”, only realising afterwards that “seere” could be interpreted as viewers, aligning with the use of the word “fans”. 

I came back to the song and listened to what I had previously done. I felt like there wasn’t enough contrast between the wet and clean vocal layers, and felt like I wanted to have a less aggressive autotune on the clean version. Since I record the autotune baked in, I had to rerecord the clean track with a lower retune speed. I also felt I wanted the intro to feel more sincere, so I tried to have a more talkative tone during the intro of the song. I think working with the autotune in relation with my vocal delivery is something that is quite interesting and elevates the perceived message of the lyrics, where autotune itself can effect the timbre and tone of the voice, something Duinker also stated (Duinker, 2025, p. 20).

I decided to wait with introducing the clean vocal until a bit into the intro, and I felt this anticipation added a lot to the feeling of the song. I continued playing with this effect, removing and adding the clean version where necessary. I continued writing the verse, rapping with a falsetto voice, and after a few bars I found myself wanting the aggressive autotune again. I then decided that the next part of the verse needed to be unintelligible. This might be due to the themes I was discussing in this part, but also as it fit the aggressive autotune. To further this distinction, I added even more delay and heavy chorus only to this part of the verse. 

After two bars, I changed my flow back to the talkative flow, noting in the song how the previous part was unintelligible writing “it don’t make no sense”. I intuitively added an adlib after the line saying “sense” almost like a manual delay. However, I wanted this adlib to feel unique, so I used the button in Antares autotune called “Ignores vibrato”, which made the pitch correction miss the notes slightly, giving this adlib an otherworldly feeling. 

5.3. Do Something

Freestyle:

Norwegian without drums:

English without drums:

Norwegian with drums:

English with drums:

5.3.3. Rewriting phase

5.3.2. Writing phase