Technique

Overmodulation: What Overmodulated Audio & Video Means

Overmodulated analog video with smeared chroma and streaking color bands

Overmodulation is what happens when a signal is driven harder than the system carrying it can handle — the level exceeds the maximum the recorder, transmitter, or file format was designed for, and the excess is cut off, squashed, or smeared into distortion. When audio or video is overmodulated, it means that signal was recorded or transmitted too “hot”: peaks have been clipped flat, harmonics that weren’t in the source have been added, and in video, brightness or color has spilled past the legal range into blown whites and bleeding chroma.

In conventional engineering, overmodulation is a mistake. In glitch art, it’s one of the most direct ways to make a medium show its edges. Every system has a ceiling, and what a system does when it hits that ceiling — the specific way it breaks — is a signature of that technology. Tape saturates warmly. Digital audio clips brutally. Analog video smears red across the frame. Learning to read those failures is how you learn to use them.

This guide covers what overmodulation means in audio and video, how to recognize it, how to prevent it when you don’t want it, and how to exploit it when you do.

What Overmodulation Means in Audio

Clipping and 0 dBFS

Digital audio stores each sample as a number within a fixed range. The largest value the format can represent is defined as 0 dBFS (decibels relative to full scale). Nothing can go above it. When a signal that “wants” to peak at +3 dBFS is written into a digital file, the samples above the ceiling are simply set to the maximum value. The top of the waveform is sliced off flat.

That flat-topped waveform is clipping, and it’s the most common form of overmodulation people encounter. It sounds like crackle and harsh fuzz on loud transients — a crunchy edge on vocals, a splattering snare, bass notes that turn buzzy. Clipping adds harmonics that were never in the original performance, and because they’re often high-order harmonics, they can alias and sound particularly harsh in digital systems.

Hard Clipping vs Soft Clipping

Not all clipping sounds the same:

  • Hard clipping cuts the waveform off abruptly at a fixed threshold. The corners where the wave meets the ceiling are sharp, which generates strong upper harmonics. Symmetrical hard clipping (same threshold for positive and negative peaks) produces predominantly odd-order harmonics — the square-wave-like buzz of digital overs and many fuzz pedals.
  • Soft clipping rounds the transition. As the signal approaches the ceiling, gain gradually decreases, so peaks are compressed into a curve rather than chopped. The result is smoother and often perceived as “warm” rather than broken. Most saturation plugins and overdrive circuits are soft clippers of one kind or another.

Asymmetrical clipping — where positive and negative halves of the wave are treated differently — adds even-order harmonics as well, which is part of why some tube stages and certain pedals have a distinctive character.

Analog Tape Saturation

Before digital recording, the ceiling was magnetic. As you drive more signal onto tape, the magnetic particles approach saturation and the tape responds less and less linearly. Instead of a brick wall, you get gradual compression of peaks, added harmonics, and a softening of transients.

Recording engineers call pushing tape this way “hitting it hot.” A little is a sought-after color; a lot becomes audible distortion, smeared high end, and a flattened, crushed dynamic. The important distinction for artists: analog overmodulation usually degrades gradually, while digital overmodulation is a hard wall with no warning.

AM and FM Broadcast Overmodulation

The term “overmodulation” actually comes from radio, where it has a precise technical meaning.

AM (amplitude modulation) varies the strength of a carrier wave to encode audio. The modulation index describes how deeply the audio varies that carrier. At 100% modulation, the carrier dips to zero on the loudest negative peaks. Push beyond 100% and the carrier gets cut off entirely for part of each cycle — the envelope no longer follows the audio, which causes severe distortion at the receiver. It also generates splatter: spurious sidebands that spread outside the station’s assigned bandwidth and interfere with neighboring frequencies. This is why broadcasters are legally required to control modulation.

FM (frequency modulation) encodes audio as variations in the carrier’s frequency. Overmodulation in FM means excessive deviation — the carrier swings further from its center frequency than the channel allows. The signal occupies more bandwidth than permitted, spills into adjacent channels, and can distort in receivers whose filters are designed for the legal deviation. See frequency modulation for the synthesis side of the same principle.

When older engineers say a recording is “overmodulated,” this broadcast meaning is where the word comes from — it has since spread to mean any signal pushed past its limit.

What Overmodulation Means in Video

Overdriven color and chroma bleed in analog glitch imagery

Video has the same concept with different units. Instead of loudness, the ceiling applies to brightness (luma) and color intensity (chroma).

Analog video levels are measured in IRE units. In NTSC, blanking sits at 0 IRE, black is typically placed at 7.5 IRE, and peak white is 100 IRE. Signals that exceed 100 IRE are overmodulated — “super-white” or “illegal” levels that broadcast equipment was designed to limit or reject.

Digital video inherited the idea. In standard 8-bit Rec. 709 video, the legal luma range is 16–235, not the full 0–255. Values above 235 are super-whites and values below 16 are super-blacks. They can exist in a file, but broadcast delivery specs and many displays clip them, so detail pushed into that zone is lost.

Overmodulated luma looks like:

  • Blown highlights — skies, faces, and light sources flatten into featureless white
  • Blooming — on analog gear and CRTs, overbright areas spread and glow into neighboring regions
  • Crushed shadows — the mirror problem at the bottom of the range, where dark detail collapses into flat black

In analog TV transmission, overmodulated video could cause problems beyond the picture. North American analog TV used negative modulation, where peak white corresponds to the lowest carrier level — so excessive white levels risked dropping the carrier out entirely, producing audible buzz in the sound channel.

Chroma Overmodulation and Color Bleed

Color is where analog overmodulation gets most visually interesting. Composite formats like NTSC, PAL, and especially VHS carry color information at far lower resolution than brightness. When chroma is driven too hard — oversaturated reds are the classic offender — the color signal can’t stay inside the edges of objects. It smears horizontally, bleeds past outlines, crawls, and vibrates.

That’s the “red smear” familiar from home video and the reason saturated reds were long avoided on broadcast sets. Oversaturated chroma can also push the combined composite signal beyond its legal limits even when the luma is fine, which is why broadcast-safe processing checks both.

Analog Video Overdrive

Analog video mixers, processors, and video synthesizers often have gain stages with more range than a legal signal needs. Crank them and the picture posterizes, colors invert or wrap, sync becomes unstable, and the image tears. In video feedback loops, gain accumulates with every pass through the camera-screen cycle — any brightness boost above unity quickly drives the loop into saturation, which is why controlling gain is the core skill of feedback work.

How to Tell If Something Is Overmodulated

In Audio

  • Waveform view — Zoom into the loudest sections in any editor. Clipped audio shows flat tops and bottoms where the wave should be rounded. In Audacity, enable View > Show Clipping to mark clipped samples in red, or use Analyze > Find Clipping.
  • Peak meters — A meter that hits 0 dBFS and lights its clip indicator is the obvious signal. Treat any clip light as a warning, not proof — and the absence of one isn’t a guarantee either.
  • True peak metering — Sample values can stay under 0 dBFS while the reconstructed analog waveform between samples exceeds it. These intersample peaks can clip at the converter or after lossy encoding. True-peak meters (reading in dBTP) catch them.
  • Your ears — Crackle on transients, fizz on sibilant vocals, and a flattened, fatiguing loudness are the audible tells.

In Video

  • Waveform monitor — Plots luma level across the frame. Anything piled up above 100 IRE (or above 235 in 8-bit digital) is overmodulated. A flat line pressed against the top of the scale means highlights have clipped and detail is gone.
  • Vectorscope — Plots hue and saturation. Traces that extend past the target boxes toward the edge of the scope indicate chroma that’s hotter than broadcast standards allow.
  • Histogram — A spike at the extreme right or left edge means pixels are stacked against the limit. The same reading applies to still images in Photoshop, Lightroom, or GIMP.

Every major editor includes these scopes: DaVinci Resolve, Premiere Pro, and Final Cut Pro all offer waveform, vectorscope, and histogram views.

How to Fix or Avoid Unwanted Overmodulation

The honest answer is that clipping is much easier to prevent than repair. Once samples or pixels are flattened against a ceiling, the original information is gone — any “fix” is a reconstruction.

Audio

  • Gain staging — Set input levels so that peaks land comfortably below 0 dBFS when recording. Recording at 24-bit gives you so much headroom that there’s no reason to record hot; leave generous margin and raise the level later.
  • Limiters — Put a true-peak limiter at the end of a mix or master chain with its ceiling slightly below 0 dBFS (around -1 dBTP is a common target for streaming delivery). A limiter catches peaks with controlled gain reduction instead of hard clipping.
  • Compression — Dynamic range compression tames peaks before they reach the ceiling, giving you a louder signal with less risk of overs.
  • Declipping — If clipping has already happened, tools like Audacity’s Clip Fix effect or dedicated restoration software can estimate and redraw the missing peaks. Results range from convincing on light clipping to unusable on heavy distortion.

Video

  • Expose correctly — Use zebras and camera waveforms to keep highlights from clipping at capture. Log profiles and raw formats offer more highlight headroom.
  • Broadcast-safe limiting — Resolve includes a broadcast-safe option in its project settings with selectable IRE limits; Premiere Pro offers a Video Limiter effect; Final Cut Pro has a Broadcast Safe effect. These clamp luma and chroma into legal range.
  • Grade with scopes, not just monitors — A bright display can hide overmodulated values. Scopes don’t lie.
  • Tame oversaturated reds — Pull saturation selectively on hot colors before delivery, especially for anything headed to analog formats or older playback chains.

Overmodulation as a Deliberate Glitch Technique

Everything above describes overmodulation as damage. For glitch artists, the ceiling is a tool. Driving a signal into its limit exposes the design decisions of a format — where the engineers drew the line, and what happens on the other side of it.

Audio: Clipping on Purpose

In Audacity, the simplest overmodulation experiment takes seconds:

  1. Select a section of audio
  2. Open Effect > Volume and Compression > Amplify (or Effect > Amplify in older versions)
  3. Check Allow clipping and set the gain far beyond the suggested value — +20 dB or more
  4. Apply, then repeat, or follow with Normalize to bring the crushed result back to a usable level

The result is square-edged, buzzing, heavily harmonic audio. For more control, Audacity’s Distortion effect offers hard clipping, soft clipping, several overdrive curves, and rectifier modes, each with a distinct flavor of overload. Combining overmodulation with bit crushing — reducing bit depth or sample rate — stacks two kinds of digital limit on top of each other.

Images: Overmodulating Data

When you databend an image by importing it as raw audio, the image bytes become sample values. Amplifying that “audio” past clipping forces large stretches of data to the maximum or minimum value. Reopened as an image, those regions become hard-edged streaks, blown-out color zones, and posterized bands — overmodulation translated from sound into pixels. Because the data in a raw image is laid out row by row, the damage often appears as horizontal smears that follow the scanlines. The Audacity databending guide covers the full workflow, including how to protect file headers.

Video: Feedback, Overdrive, and Bent Hardware

  • Feedback loops — In video feedback, pushing brightness or contrast just past unity drives the loop into saturation, where it blooms, posterizes, and locks into high-contrast fractal patterns. Riding the edge between stable and overmodulated is where most of the interesting imagery lives.
  • Overdriven analog gear — Cheap video mixers, processing amplifiers, and old camcorders can be fed hotter signals than they expect by chaining gain stages or daisy-chaining outputs into inputs. Expect crushed whites, chroma that bleeds and wraps, and unstable sync. See VHS effects for building analog chains.
  • Circuit-bent video devices — Circuit bending video mixers and signal processors lets you reroute gain and color circuitry directly, creating overmodulation that no front-panel knob can reach. Work carefully: never bend anything that plugs into mains power.

Digital Images: Pushing Channels to Extremes

You can overmodulate a still image without any audio tools:

  • Curves and Levels — Drag the curve into a near-vertical line or clamp the Levels input sliders together. Tones pile up against pure black and pure white, and the image collapses into hard graphic shapes — closely related to color banding and posterization.
  • Channel overdrive — Photoshop’s Channel Mixer allows values up to 200% per channel. Push one channel to the maximum and the others negative, and color clips unevenly across the image, creating saturated color fields that no camera would record.
  • Saturation stacking — Apply Hue/Saturation at full strength several times. Each pass pushes more pixels against the channel limits until gradients break into flat, electric color.
  • Pair it with offsets — Overmodulated channels combined with RGB splitting exaggerate the fringing, since the clipped channels have hard edges to separate.

The Photoshop glitch guide and color effects page cover these tools in more depth.

Overmodulation in Music

Some genres treat overmodulation as a foundation rather than an accident.

Noise music and especially harsh noise wall are built from signals driven permanently into saturation — dense, static walls of distortion where the clipping is the composition. Noise music practitioners commonly use no-input mixing boards, where a mixer’s output is routed back into its own input to create a feedback loop that self-oscillates and overloads.

Industrial music drew on overdriven tape, distorted vocals, and blown-out drum machines to evoke machinery and decay.

Glitch music works with the failures of digital systems — skipping CDs, buffer errors, and digital clipping among them. See what is glitch music for the genre’s background, and glitch hop for how those textures entered beat-driven electronic music.

The mainstream loudness war is overmodulation in a subtler form: masters limited and clipped to maximize perceived volume, trading dynamics for loudness. What glitch and noise artists do deliberately and audibly, commercial mastering often does quietly.

For live, patchable approaches to overload and distortion, Pure Data and Max/MSP let you build clipping, waveshaping, and feedback systems from scratch. Overmodulated audio also makes a strong driver for audio-reactive visuals, where clipped peaks translate into abrupt visual spikes.

Overmodulation vs Distortion vs Clipping

These terms overlap and are often used interchangeably. Here’s how they relate:

TermWhat It MeansWhere It AppliesTypical Result
OvermodulationA signal driven past the maximum a system is designed to carryBroadcast (AM/FM), recording, analog and digital videoClipping, splatter, bleed, blown highlights
ClippingThe specific effect of peaks being cut off at a ceilingDigital and analog audio, video levels, image channelsFlat-topped waveforms, blown whites, lost detail
DistortionAny change to a signal’s waveform that wasn’t in the sourceEvery mediumAdded harmonics, altered tone, noise

In short: overmodulation is the cause, clipping is the most common mechanism, and distortion is the broad category of result. All clipping is distortion, but not all distortion is clipping — phase shifts, filtering, and waveform distortion techniques alter a signal without hitting any ceiling.