The Drop · September 18, 2026
Reggaeton and dembow visualizer: the tempo the app is most likely to double
Reggaeton lives on one specific number, and it is the same number the tempo detector uses to decide whether to keep what it found or double it.
Photo via Unsplash
Drop a reggaeton track into the app and it prints a tempo next to the filename before you touch a single setting. On most genres that number is boring and correct. On this one it is a coin toss between the tempo you are feeling and exactly twice it, and the reason is not that detection is bad at reggaeton. It is that reggaeton parks itself on the one number the rule is built around.
Where reggaeton actually sits
Not a vibe, a measurement. Mixgraph publishes a tempo distribution taken from 296 analysed reggaeton tracks, measured 18 September 2026 with outliers stripped by an interquartile fence. The median is 95 BPM. Half the catalogue falls between 90 and 98. Nine in ten fall between 87 and 102.
Individual records land where you would expect from that. Gasolina reads 96 BPM and Despacito reads 89 BPM. Seven BPM apart, both unmistakably the same genre, and as far as the app is concerned they are on opposite sides of a cliff.
Why 95 is the number that matters
When a file loads, the app scans its first 60 seconds offline, autocorrelating the energy envelope against every tempo from 60 to 180 BPM in half step increments. The beat detection guide walks through that pipeline properly. What matters here is the single line that runs after the scan finishes: if the winning tempo came in under 95 BPM, and double time also scores above 40 percent of the winner, the number gets doubled.
That rule is there so a slow, sparse track still lands on a grid you can cut to. It fires or it does not, and the threshold is 95 exactly. Which is the median of the genre.
The second condition makes it likelier still, because of what dembow is. The pattern is not a bare kick sitting on each beat. It is a steady kick with snare and rim hits falling between the kicks, the syncopation the whole groove is named for. That puts real onset energy on the half beat positions, and half beat energy is precisely what a double time correlation rewards. A busy offbeat is the thing that makes the doubling rule say yes.
Two tracks, seven BPM apart, two different videos
Run the two examples through it. Gasolina at 96 is above the line, so the number is left alone and the grid is 96 BPM. Despacito at 89 is below the line, so if double time scores it becomes 178, and every content setting you pick is now measured against a grid running at twice the speed of the song you are hearing.
In milliseconds: a beat at 89 BPM is 674 ms of screen time and a beat at 178 BPM is 337 ms. Same menu option, half the dwell. It gets sharper at the default. The content change rate ships on Every 1/4 beat, which is 169 ms per change on an honest 89 grid and 84 ms on a doubled one. Flashes have a hard floor of 90 ms between them, so at 84 ms the picture is swapping faster than the app can physically flash and some of those changes happen with nothing lighting them up.
Reading which one you got
The tempo appears next to the file name and again in the status line as soon as decoding finishes, so the check costs nothing. Compare it against the song. If it printed something in the 90s, the grid matches the pulse. If it printed something in the 170s or 180s on a track you know is a normal reggaeton tempo, it doubled, and the fix is one step on the menu rather than anything technical.
- Printed near 90 to 100. Every beat gives you one content change per felt beat. Every 2 beats gives you one per half bar.
- Printed near 175 to 190. Every 2 beats is now the one change per felt beat setting, and Every 4 beats is the calm one.
Neither reading is a bug. A doubled grid is a genuine periodicity in the music, and on a busy dembow it can be the more useful of the two. You just want to know which one you were handed before you choose a rate.
Put your images on the snare, not the kick
Every uploaded image can take a role of Any, Kick, Snare or Hi-hat, and a hit on that drum flashes that image in as a temporary override. The instinct is to hang the cover art on the kick. On reggaeton that gives you the dullest possible result, because the kick is the steady part. The offbeat snare is what makes the groove identifiable, so that is where a cover or a logo earns its place. The trap post covers how the three drum triggers behave in general.
The triggers work on spectral flux, meaning the rise in a band since the previous frame rather than its level, so a busy percussion layer over sustained bass still reads cleanly. Put the shaker and the hat pattern on the Hi-hat role if you want something moving continuously, and leave one or two images on Any so the normal rotation still has something to show.
One setting to leave alone: the bass drop detector defaults to zero, and on a track with near constant low end there is not much for it to find. Turn it up only if the record has a real breakdown to come back from.
Dominican dembow is a different tempo entirely
Worth separating, because the two words get used interchangeably and the numbers are nowhere near each other. Dominican dembow runs much faster than reggaeton. Mixgraph's catalogue of 23 El Alfa tracks puts the median at 120 BPM. That is comfortably above 95, so the doubling rule never fires and the whole problem above disappears. You get the tempo you expect, and because the grid is faster to start with, Every 1/2 beat lands about where Every beat does on a reggaeton record.
A note on the fast grid
If your tempo printed in the 170s or 180s and you leave the content rate frantic, you are making something that flashes several times a second. Flash rates above roughly three per second are a genuine seizure risk for people with photosensitive epilepsy, and a doubled reggaeton grid sits right in that band before you add any persistent flicker on top. Keep the flicker opacity low or at zero for anything people will watch through to the end, and say so in the description when you post it.
The short version
- Reggaeton's measured median is 95 BPM, which is the exact threshold for doubling.
- Read the printed tempo before choosing anything. It is either the pulse or double it.
- Doubled grid, one step slower on the content rate. Every 2 beats instead of Every beat.
- Cover art goes on the Snare role. The kick is the steady part, the snare is the groove.
- Dominican dembow sits near 120 and is never doubled.
Load a track and check the number
Free, in the browser, no signup and no watermark. Your audio never leaves your machine and each render comes out as a 1080p WebM. Upload the record, read the tempo it prints, set the content rate from there, or take the tour from the home page.
Open the visualizer