What This Actually Is
Floor Therapy Video refers to a specific approach in audio engineering where floor-level ambient noise recordings are used as the sonic foundation for a piece. You capture the room tone, the low rumble of HVAC, the subtle vibrations from traffic or footsteps, and then process that raw material into something intentional. It is not ASMR. It is not white noise. The distinction matters because people consistently conflate the three and then wonder why their mix sounds muddy instead of grounded. I have spent the better part of a decade working on immersive audio projects that rely heavily on this technique. The first time I went full-in on it, I assumed I would just record quietly and layer it under everything. That did not work at all. The low end from a real floor has harmonic content that is unpredictable. If you do not know what you are doing with it, you get phase cancellation and a mix that collapses on any system other than the one you built it on.
How to Create a Floor Therapy Video
Start with the recording chain. Use a contact microphone or an accelerometer mounted to the floor surface, not an air microphone. Contact mics pick up the actual vibration of the structure. Air mics pick up the room, which is two different things. I use a Bruel & Kjaer 4375 accelerometer paired with a preamp that has a flat response down to 1 Hz. That allows me to capture the sub-bass rumble that actually drives this kind of work. Record for at least twenty minutes. The longer the sample, the more you capture recurring mechanical cycles, like a refrigerator compressor kicking on every twelve minutes or a bus that passes at predictable intervals. Those patterns are the material. Short recordings give you nothing to work with. A twenty-minute clip gave me roughly four full cycles of the building's mechanical system, which was enough to build a looped bed without obvious repetition. From there, band-pass filter the recording to isolate the frequency range you want. Most of the useful energy sits between 20 Hz and 200 Hz. Everything above 300 Hz is usually just noise that will clash with your actual instruments. I then sidechain the filtered bed to the kick or bass element of the mix so it ducks automatically. This prevents the floor material from masking the low end of the actual track. Without sidechain compression here, the mix becomes a sludge within forty seconds.
Process the bed with convolution reverb using an impulse response from a similar physical space. This makes the floor recording sound like it belongs in the imagined environment rather than sounding like a separate layer sitting on top. Mix the processed bed at roughly -18 LUFS relative to your main mix bus. Anything louder and the listener will consciously register it as a separate element. Anything quieter and it vanishes entirely underneath the other instrumentation. Export at 96 kHz sample rate if you can. The sub-bass transients in these recordings require headroom during processing. If you are working at 44.1 kHz and applying heavy EQ or distortion, you will introduce aliasing artifacts that become very audible in the 20 to 50 Hz region. I learned that the hard way on a project that had to ship to a Dolby Atmos facility.
Get the Full Details
Common Mistakes and What to Do Instead
The biggest error I see is treating floor therapy as a master bus effect. It is not. It needs to sit in its own bus with its own processing chain, EQ'd independently, and panned or spatialized separately. When you slap it on the master, you lose control and the low end fights with every other element that occupies that space. Build it into the arrangement from the ground up, literally. Another issue is using the raw recording without cleaning it first. HVAC rumble often contains a 60 Hz hum from faulty grounding. If you leave that in, it will dominate your low end and make the entire mix sound boomy no matter how much you EQ elsewhere. I run the recording through a notch filter at 60 Hz and its harmonics before doing anything else. That single step saves me hours of mixing. Here is a specific edge case that cost me two days on a recent project. The floor material I recorded had a subtle periodic pulse at 0.8 Hz, caused by a nearby elevator motor cycling. At first I could not identify it. It only revealed itself when I soloed the low end and visualized the waveform in the time domain. The pulse was so slow that it escaped my attention during normal monitoring. I tried EQ'ing it out and could not find a frequency to cut. The fix was to manually automate a volume dip every 1.25 seconds on the bed track, syncing to the pulse rate. Once I matched the automation to the cycle, the artifact disappeared completely and the mix locked together.
Where This Approach Fails
Floor therapy does not work for fast-paced music with heavy drum programming. The low-frequency energy from floor recordings tends to smear transients and reduce punch. If your track relies on tight, defined kicks and snares, this technique will actively work against you. In those cases, generating synthetic sub-bass instead gives you far more control over the transient response. The tradeoff is that you lose the organic textural quality, but your mix will be cleaner. It also requires a quiet source environment. If your building has inconsistent noise floors, like random construction or heavy foot traffic, the recordings become unusable for anything other than abstract experimental work. I once spent three weeks trying to salvage a set of recordings from a building next to a demolition site. The concrete dust vibrations were embedded in every take. I had to abandon the project and re-record at a different location. Budget extra time for environmental uncontrollability. If you want to explore this further, search for a Floor Therapy Video tutorial on YouTube that walks through the signal chain from accelerometer to DAW. Most of the written guides skip the preprocessing step, which is where most beginners lose the signal before it even reaches their mixer.
