By SongTools.ai
Visual frequency analysis catches problems that listening fatigue hides. Here's how to read a spectrogram and what to do with what you see.
There's a well-documented phenomenon in mixing: the longer you work on a track, the less accurately you hear it. After several hours of listening, your auditory system adapts to the sound you've been making decisions around, and subtle frequency imbalances become invisible to your ears. You stop hearing that 300 Hz buildup because it's been there so long your brain has normalised it.
This is why professional engineers have always used visual tools — meters, spectrum analysers, spectrograms — as a reference check alongside their ears. Visual analysis doesn't replace listening. But it gives you a second channel of information that doesn't suffer from listening fatigue.
A spectrogram displays three dimensions of audio simultaneously: time on the horizontal axis, frequency on the vertical axis, and amplitude as colour intensity. Bright areas represent high energy at a given frequency and moment in time. Dark areas represent silence or very low energy.
This three-dimensional view reveals things a standard frequency spectrum can't: how frequency content evolves over the duration of a track, where transients occur, how different elements move in and out of frequency bands, and where the mix is consistently dense or consistently thin.
Sub-bass (20–60 Hz): Should have energy when the kick drum hits and when a bass note sustains, but shouldn't show continuous high-energy presence between notes. Constant brightness in this region usually indicates rumble or room resonance bleeding into the mix.
Bass (60–250 Hz): The fundamental frequencies of most bass instruments and the kick drum body. In a well-balanced mix, you'll see alternating patterns between kick drum hits and bass notes — they shouldn't both be at maximum energy simultaneously in the same frequency band if you've done your low-frequency arrangement work correctly.
Low mids (250–800 Hz): This is where muddiness lives. A spectrogram that shows persistently bright energy across this range in busy sections often indicates too many instruments competing without sufficient EQ separation.
Upper mids (800 Hz–4 kHz): The presence region — where the human voice is most intelligible and where instruments cut through a mix. If your vocal appears dim in this range while the instruments around it are bright, you'll have an intelligibility problem.
Highs (4–20 kHz): Air, sparkle, and harshness all live here. A consistently bright upper region can indicate an over-bright mix that will fatigue listeners. Very dark highs can indicate an over-filtered or dull mix that will sound muffled on consumer speakers.
Horizontal bands of constant brightness: Usually indicate noise — room noise, electrical hum, or a plugin that's generating continuous tone. These are nearly impossible to hear when buried in a mix but immediately visible in a spectrogram.
Vertical bright lines: Impulse events — usually transients from drums, clicks, or other percussive elements. If you see unexpected vertical lines in between intentional transient events, there may be clicks or edits that didn't get properly crossfaded.
Smearing across time in the high frequencies: Often indicates excessive reverb or delay on high-frequency elements, or phase artefacts from processing. High frequencies should have relatively sharp attack and faster decay than low frequencies.
Asymmetry between the left and right channels: If you're viewing a stereo spectrogram and one channel is consistently brighter than the other in certain frequency bands, you have a stereo balance or panning problem that may not be immediately obvious on casual listening.
One of the most practically useful applications of spectrogram analysis is identifying sub-bass content that shouldn't be there. Most microphones and instrument pickups capture some content below 80 Hz even when the source has nothing going on in that range. This content adds up in a mix and creates unwanted low-end weight.
A spectrogram showing persistent bright content below 60 Hz on tracks like acoustic guitar, piano, or vocals is almost always something to high-pass filter out. The filtered material wasn't contributing to the sound you wanted, but it was contributing to the buildup of energy your subwoofer or bass speakers are trying to reproduce.
The Spectral Visualizer in SongTools.ai generates a high-resolution spectrogram from your uploaded audio. What makes it useful as part of a mix workflow:
Pre-master check. Before mastering, run your final mix through the spectrogram. Look for the patterns described above — particularly anything in the sub-bass, consistent mid-range muddiness, or frequency ranges where elements aren't clearly separated.
Comparing mix versions. If you're iterating between mix revisions, compare spectrograms side-by-side. Sometimes what felt like a significant change in the studio looks nearly identical spectrally — and sometimes a subtle processing decision creates a dramatically visible difference in the frequency balance.
Identifying masking issues. If you can see that two important elements — say, the lead vocal and a prominent synth pad — are occupying almost identical frequency space with similar energy levels, you've visually confirmed the masking problem that your ears might have been telling you about less clearly.
Spectrogram analysis is a diagnostic tool, not a prescriptive one. It can tell you that there's a frequency buildup in the low mids. It can't tell you which specific element is causing it or how it should be addressed — that's an artistic judgment that requires listening.
Similarly, a technically clean spectrogram doesn't guarantee a great-sounding mix. Music isn't engineering: the goal is emotional impact, not spectral tidiness. Some great recordings have all kinds of messy frequency content that would look "wrong" on a spectrogram but sounds exactly right in context.
Use spectrograms to identify and confirm problems you're already suspecting, not as a checklist to optimise toward.
The most practical approach is to use spectral analysis at specific decision points in your workflow rather than as a constant monitoring presence. Run it when:
- You've finished a rough mix and want to check your low-end balance before moving to the arrangement phase.
- You've received a reference track from a client and want to understand its frequency character before you start working toward it.
- You're about to submit a mix for mastering and want a final sanity check.
- Something doesn't sound right but you can't identify what — the spectrogram often makes the culprit immediately visible.
Visual analysis combined with careful listening is more powerful than either alone. Your ears tell you how something feels. The spectrogram tells you what's actually there.
All the tools discussed in this article are available on SongTools.ai — free to explore.