Are -3 dBFS and -3 dBTP the Same?
When we talk about audio, one important thing to know is how loud the sound is. This loudness is measured using decibels, which tell us the strength or intensity of the sound.
There are a few different ways to measure audio levels using decibels. Two common measurements are dBFS which means “decibels relative to full scale” and dBTP which is short for “decibels true peak.”
What is dBFS?
dBFS stands for “decibels relative to full scale.” It measures the amplitude of digital audio signals compared to the maximum possible digital level.
In digital audio, the highest possible level is 0 dBFS. All audio levels will be negative numbers below that. For example, -3 dBFS is a strong signal, while -20 dBFS is much quieter.
dBFS is used to measure the loudness of individual digital audio samples. It looks at the actual data values of each sample.
What is dBTP?
dBTP means “decibels true peak.” True peak measures the actual peak level of an audio waveform after the digital-to-analog conversion process.
Digital audio uses discrete samples to represent sound waves. When converted back to analog, the audio level can actually peak higher than the highest sample level. This is called inter-sample peaking.
For example, a digital audio signal might max out at -3 dBFS. But the true analog peak between those samples could actually hit -1 dBTP or higher.
True peak is an estimate of the real analog level, not just the raw digital values. It predicts how high the signal will actually peak in the real world.
Are -3 dBFS and -3 dBTP the Same?
Now that we understand dBFS and dBTP, the big question is: Are -3 dBFS true peak and -3 dBTP the same thing? The simple answer is no.
-3 dBFS measures the highest sample level in the digital signal itself. It’s the maximum value of the actual data.
-3 dBTP estimates what the reconstructed analog signal will peak at in the real world. It accounts for inter-sample peaking.
In most cases, the true peak level (dBTP) will actually be higher than the sample peak (dBFS). An audio signal at -3 dBFS will likely have true peaks above -3 dBTP when played back.
Why dBTP is Important
You might wonder, if our audio is digital, why do we care about this analog “true peak” level? Isn’t dBFS enough? Well, it’s important because all digital audio ultimately gets converted back to analog.
Whether it’s played through speakers, headphones, or broadcast on TV/radio, the digital signal is turned into a continuous analog waveform. And that’s where inter-sample peaks can cause problems.
If the true peak level is too high, it can cause clipping distortion when converted to analog. The audio can sound crunchy, distorted, or fuzzy because the signal is too hot.
To avoid this, audio should be mastered with true peak levels in mind, not just dBFS. Many guidelines recommend keeping true peaks below -1 dBTP to maintain headroom.
Measuring True Peak
Specialized audio meters can measure dBTP levels. They use oversampling and interpolation to predict what the signal will do between samples.
Oversampling reads the digital signal at a much higher sample rate. This fills in the gaps between the original samples. Interpolation then connects those points to model the analog waveform.
From this model, the meter can determine the actual peak amplitude in dBTP. This gives a more accurate assessment of the real-world loudness.
Loudness Normalization and True Peak
Another reason true peak is important is for loudness normalization. Many streaming platforms now adjust the loudness of audio content to a consistent level.
Loudness is typically measured in LUFS (Loudness Units relative to Full Scale). But the true peak level is also considered.
Content may be turned down if the true peak is too high. This is to avoid clipping when boosting quieter content up to the target loudness.
For example, a song that is -16 LUFS but has true peaks hitting +2 dBTP may be turned down more than a -16 LUFS song with -1 dBTP peaks. The first song has less headroom for normalization without clipping.
