Audio is an inherently complex signal.
Anything and everything we hear can be described by an audio signal.
Every sound has its own characteristics, and our ears can isolate and identify them.
High-resolution music has become quite the buzzword these days, but can we identify it simply by listening to it? The audio quality may also vary depending on factors like encoder settings, type of compression, speaker quality, etc. We will interpret this via an audio spectrogram.
What you will learn:
An audio spectrogram is a useful tool for analyzing digital audio, allowing you to visualize and understand how the audio signal evolves over time.
Features like frequency distribution, audio bandwidth, tone quality, etc. can be determined.
Observing an audio signal in the time domain gives us an idea of the overall volume of the track and how it varies with time. The loud, soft, and silent components of a track can be easily identified. This, however, does not tell us much about the quality of the instrument being played.
To identify sound quality, we need to look at it from the frequency domain.
Frequencies are the fundamental component of any sound.
High-quality audio contains a wide range of frequencies.
In any time interval, the resultant sound is due to the constructive and destructive interference of multiple frequencies.
Another metric of audio quality is timbre.
Timbre can be called audio flavor or tone. It allows a listener to distinguish between the musical tone of a violin or trumpet even if the tone is played at the same pitch with the same loudness.
The timbre of an instrument is determined by the overtones it emphasizes.
An overtone is any harmonic with a frequency greater than the fundamental.
The composition of a musical instrument’s sound in terms of its partials can be visualized by a spectrogram. Now, let's try to understand what a spectrogram really is.
A spectrogram is a heatmap type of visualization of all the frequencies in an audio track.
The higher the energy of a frequency, i.e., the louder it gets, the brighter it looks in the heatmap.
Looking at the patterns of the distribution of frequency energy, we gain valuable insights regarding an audio signal which cannot be identified in the time domain.
We see a constant energy level in the 2 kHz band.
Frequency patterns relate to the different components of a song. You can identify instruments, vocals, and tunes from lead instruments. The fundamental frequencies can be easily identified as they are the brightest in color in a given interval. The overtones are stacked above with decreasing volume.
Here, we see a single note played on an organ. The fundamental frequency is the brightest while the overtones are stacked above with decreasing volume.
A trained musician can control the loudness and sustain of an instrument, which is a marked element of performance. The bright colors in the spectrum are the loudest frequencies. This will relate to a lead tune that will dominate the sound.
This is a spectrogram of an arpeggio played on a piano.
Piano Slow Melody
A wide range of frequencies indicates better audio fidelity.
The sound’s complexity depends on the interference of the overtones.
More overtones symbolize a much richer sound created by musical instruments.
The volume modulation indicates the control and feel a musician is attempting to generate.
Example: spectrogram of a guitar
The guitar spectrogram contains frequencies up to 16 kHz.
We can see the notes being plucked in bright colors, with their fundamental frequencies in the range of 512 Hz.
The overtones are stacked in the higher frequency bands with decreasing volume.
A brief pause every couple of notes is also apparent.
We can see the sharpness and sustain of each note being played.
If we look at the fundamental frequency, we can guess the scale.
Example: spectrogram of a violin
This is a spectrogram of notes played on a violin.
It is softer in sound, smooth, and connected; in musical terms, it’s a clear example of legato.
Softer overtones are visible, which add to the complexity of the sound.
All notes have approximately the same volume and sustain.
A spectrogram is a great way to try and understand the sounds you hear in the world around you. With one, you will be able to analyze the characteristics of any sound source. Your entire music listening experience will become more intricate and fulfilling.
Do apply the above principles and remember to have fun while doing so. Stay tuned for similar content.
Did you like the blog? If yes, we're sure you'll also like to work with the people who write them - our best-in-class engineering team.
We're looking for talented developers who are passionate about new emerging technologies. If that's you, get in touch with us.
Audio is an inherently complex signal.
Anything and everything we hear can be described by an audio signal.
Every sound has its own characteristics, and our ears can isolate and identify them.
High-resolution music has become quite the buzzword these days, but can we identify it simply by listening to it? The audio quality may also vary depending on factors like encoder settings, type of compression, speaker quality, etc. We will interpret this via an audio spectrogram.
What you will learn:
An audio spectrogram is a useful tool for analyzing digital audio, allowing you to visualize and understand how the audio signal evolves over time.
Features like frequency distribution, audio bandwidth, tone quality, etc. can be determined.
Observing an audio signal in the time domain gives us an idea of the overall volume of the track and how it varies with time. The loud, soft, and silent components of a track can be easily identified. This, however, does not tell us much about the quality of the instrument being played.
To identify sound quality, we need to look at it from the frequency domain.
Frequencies are the fundamental component of any sound.
High-quality audio contains a wide range of frequencies.
In any time interval, the resultant sound is due to the constructive and destructive interference of multiple frequencies.
Another metric of audio quality is timbre.
Timbre can be called audio flavor or tone. It allows a listener to distinguish between the musical tone of a violin or trumpet even if the tone is played at the same pitch with the same loudness.
The timbre of an instrument is determined by the overtones it emphasizes.
An overtone is any harmonic with a frequency greater than the fundamental.
The composition of a musical instrument’s sound in terms of its partials can be visualized by a spectrogram. Now, let's try to understand what a spectrogram really is.
A spectrogram is a heatmap type of visualization of all the frequencies in an audio track.
The higher the energy of a frequency, i.e., the louder it gets, the brighter it looks in the heatmap.
Looking at the patterns of the distribution of frequency energy, we gain valuable insights regarding an audio signal which cannot be identified in the time domain.
We see a constant energy level in the 2 kHz band.
Frequency patterns relate to the different components of a song. You can identify instruments, vocals, and tunes from lead instruments. The fundamental frequencies can be easily identified as they are the brightest in color in a given interval. The overtones are stacked above with decreasing volume.
Here, we see a single note played on an organ. The fundamental frequency is the brightest while the overtones are stacked above with decreasing volume.
A trained musician can control the loudness and sustain of an instrument, which is a marked element of performance. The bright colors in the spectrum are the loudest frequencies. This will relate to a lead tune that will dominate the sound.
This is a spectrogram of an arpeggio played on a piano.
Piano Slow Melody
A wide range of frequencies indicates better audio fidelity.
The sound’s complexity depends on the interference of the overtones.
More overtones symbolize a much richer sound created by musical instruments.
The volume modulation indicates the control and feel a musician is attempting to generate.
Example: spectrogram of a guitar
The guitar spectrogram contains frequencies up to 16 kHz.
We can see the notes being plucked in bright colors, with their fundamental frequencies in the range of 512 Hz.
The overtones are stacked in the higher frequency bands with decreasing volume.
A brief pause every couple of notes is also apparent.
We can see the sharpness and sustain of each note being played.
If we look at the fundamental frequency, we can guess the scale.
Example: spectrogram of a violin
This is a spectrogram of notes played on a violin.
It is softer in sound, smooth, and connected; in musical terms, it’s a clear example of legato.
Softer overtones are visible, which add to the complexity of the sound.
All notes have approximately the same volume and sustain.
A spectrogram is a great way to try and understand the sounds you hear in the world around you. With one, you will be able to analyze the characteristics of any sound source. Your entire music listening experience will become more intricate and fulfilling.
Do apply the above principles and remember to have fun while doing so. Stay tuned for similar content.
.svg)
.svg)
.png)
