Logo

See the Frequency Content of Any Audio, Over Time

Render a whole audio file as a spectrogram, or watch one scroll live from your microphone. Choose the FFT size, a linear or log frequency axis, and a viridis or grey colour map, then hover for the exact time, frequency and level in dB.

The FFT runs on the file or your live microphone entirely in this tab. Files are never uploaded, and microphone audio is never recorded or sent anywhere.

免费音频工具

音频格式转换

在MP3、WAV、OGG、AAC、FLAC、M4A之间互转

音频剪辑

在浏览器里剪 MP3、WAV、M4A，顺手做个手机铃声

铃声制作

把任意歌曲剪成 iPhone .m4r 或安卓 MP3 铃声

MP3 标签编辑器

编辑 ID3 标签和封面，无需重新编码

BPM 检测器

精确到小数一位读出歌曲速度，不用手动打拍子

打拍测速

跟着歌敲几下,立刻读出它的 BPM

WAV to MP3

在MP3、WAV、OGG、AAC、FLAC、M4A之间互转

M4A to MP3

在MP3、WAV、OGG、AAC、FLAC、M4A之间互转

在线节拍器

使用我们的免费节拍器保持完美节奏。

在线乐器调音器

精准调音您的乐器。

带音效的录音机

录制并用专业音效转换你的声音。

音乐可视化

观看令人惊叹的视觉效果随音乐起舞。

在线 MIDI 编辑器

在浏览器中逐音符作曲。

五度圈

通过交互式五度圈掌握音乐理论。

音符频率计算器

计算任何音符的精确频率。

节奏游戏

用你自己的音乐玩节奏游戏。

MusicXML播放器

加载MusicXML乐谱，搭配逼真的Salamander三角钢琴音色播放

How the spectrogram is built

The tool slices the audio into short, overlapping windows, applies a Hann window to each one to reduce artifacts at the edges, and runs a Fast Fourier Transform on it to find how much energy is present at each frequency during that instant. Stitching thousands of these slices side by side produces the image: time runs along the x-axis, frequency runs up the y-axis, and brightness shows the level in decibels at that point.

File mode renders a whole recording, up to 100 MB and 10 minutes, in one pass so you can hover anywhere for an exact reading and download the result as a PNG. Live mode skips the file entirely and scrolls a spectrogram straight from your microphone in real time, better suited to watching a room, an instrument or a feedback loop as it happens rather than inspecting a finished recording.

Reading the picture, with an example

A file that was really encoded from a lossy source, then converted back to a lossless format and passed off as one, usually shows a hard horizontal ceiling with nothing above it, commonly right around 16 kHz or 19 kHz depending on the original bitrate. A genuinely lossless file typically shows content reaching much closer to half the sample rate, with no sharp cut-off line. Steady hum from mains electricity shows up as a thin constant line, usually at 50 or 60 Hz.

The log frequency axis stretches the low end and compresses the high end to match how the ear actually perceives pitch, each octave taking up roughly the same height, which is usually the better default for music. The linear axis spaces frequencies evenly instead, which makes it easier to read off an exact cut-off frequency or compare two close high-frequency features precisely.

Limits and what to keep in mind

FFT size is a trade-off, not a quality setting: 4096 separates close frequencies more sharply but smears fast transients across a wider time window, while 1024 tracks quick changes better but blurs frequencies that sit close together. The colour map, viridis or grey, only changes how easy the same data is to read, it doesn't change what was measured.

A spectrogram shows frequency content, not loudness; it won't tell you how loud a track is or whether it will clip on playback, that's a separate measurement handled by the LUFS meter. Live mode's low-frequency accuracy depends on your device's microphone, a laptop's built-in mic in particular tends to roll off well before 60 Hz.

Spectrogram blueprint

What the FFT settings and axes mean in practice.

InputAn audio file (up to 100 MB, up to 10 minutes) or a live microphone stream.
OutputA spectrogram image with a PNG export, plus a hover readout of time, frequency and level in dB.
MethodShort-time Fourier transform with a Hann window; FFT size of 1024, 2048 or 4096.
Range/LimitsLinear or log frequency axis; viridis or grey colour map; live mode has no file size limit but no PNG export.
Best forSpotting a lossy-upsampled file, finding hum or a resonance, checking a mix's low end, seeing formants.
RelatedLUFS meter for loudness, not frequency.

How to Read a Spectrogram Online

Three steps, all inside your browser

1

Choose a file or your microphone

Drop an audio file up to 100 MB and 10 minutes long, or switch to live mode to scroll a spectrogram straight from your microphone.

Tips: File mode is better for close inspection with hover readouts; live mode is better for watching a room, an instrument or feedback in real time.

2

Set the FFT size and axis

Pick an FFT size of 1024, 2048 or 4096, then switch between a linear or log frequency axis and a viridis or grey colour map.

Tips: A bigger FFT size gives sharper frequency detail but blurs fast changes in time; a smaller one does the opposite.

3

Read it and save a PNG

Hover anywhere on the image for the time, frequency and level in dB at that point, then download the spectrogram as a PNG.

Tips: The log frequency axis matches how the ear hears pitch, so it's usually the better default for checking music rather than test tones.

Spectrogram Questions

How to read the image and what it can and can't tell you

A file really encoded from a lossy source, like a 128 or 192 kbps MP3, and then converted to a lossless format usually shows a hard horizontal cut-off in the spectrogram, commonly around 16 kHz or 19 kHz, above which there's no content at all. A genuine lossless file normally shows content reaching much closer to half the sample rate, with no sharp ceiling.
It trades frequency resolution against time resolution. A larger FFT size like 4096 separates close frequencies more clearly but smears fast transients across a wider time window; a smaller size like 1024 tracks quick changes better but blurs frequencies that sit close together. Neither setting is more correct, they just suit different material.
No. Both modes run the analysis, an FFT with a Hann window applied to the file or the live microphone signal, entirely inside this browser tab. Nothing is uploaded, and microphone audio is not recorded or stored anywhere.
Brightness represents level in decibels at that time and frequency, brighter for louder. The viridis colour map runs from dark purple at low levels to bright yellow at high ones; the grey map uses black to white for the same thing, which some people find easier to read fine detail in.
Steady electrical hum sits as a thin horizontal line, usually at 50 or 60 Hz and its multiples. A resonance or room mode shows up as a persistent bright band at one frequency, and vowel formants in a vocal appear as a stack of moving bright bands. It's also a fast way to confirm how much high-frequency content a sample actually has.
Yes, the LUFS meter measures integrated loudness, true peak and loudness range for a whole file against streaming and broadcast targets, which is a different measurement from what a spectrogram shows.

More ways to inspect your audio

Measure loudness or true peak on the same file with the LUFS meter.

Want a single loudness number instead of a frequency picture? AI 音乐生成器

价格

选择适合您的计划。没有隐藏费用，没有意外。

入门版

开始您的音乐之旅

14.99
1 月
USD
600点1 月
200 首1 月
100 次1 月
2 个(并行任务数)
AI音乐生成器
AI歌词生成器
AI人声消除
AI音轨分离
AI母带处理
音乐分析器
AI翻唱生成器
1000 次(音乐下载)
商业许可证
混音与编辑
AI歌声模型训练
邮件支持
热门

爱好版

释放您的创意潜能

29.99
1 月
USD
1500点1 月
500 首1 月
250 次1 月
5 个(并行任务数)
AI音乐生成器
AI歌词生成器
AI人声消除
AI音轨分离
AI母带处理
音乐分析器
AI翻唱生成器
无限(音乐下载)
商业许可证
混音与编辑
AI歌声模型训练
邮件支持

专业版

专业级音乐制作工具

49.99
1 月
USD
3600点1 月
1200 首1 月
600 次1 月
10 个(并行任务数)
AI音乐生成器
AI歌词生成器
AI人声消除
AI音轨分离
AI母带处理
音乐分析器
AI翻唱生成器
无限(音乐下载)
商业许可证
混音与编辑
AI歌声模型训练
邮件支持