Audio AI GUIDE
Filterbank and PLP Features
Mel filterbank and Perceptual Linear Prediction (PLP) features summarize speech spectra using perceptually spaced frequency bands.
Listen to this guideAI-narrated · ~2 min
On this page2 min read
Overview
Mel features pool short-time spectral energy into bands; PLP adds auditory weighting, loudness compression, and linear prediction. These are representations used by some speech systems, not guarantees of noise or speaker invariance.
Deep Dive
A typical filterbank pipeline divides speech into short frames, computes a spectrum, and applies overlapping filters spaced on a perceptual frequency scale such as mel. The weighted energy in each band forms filterbank features; taking a logarithm yields log-mel features. PLP, introduced by Hermansky, uses critical-band analysis, an equal-loudness curve, cube-root intensity compression, and all-pole linear prediction to produce a smooth spectral envelope. MFCCs are related features that apply a discrete cosine transform to log-mel energies. Modern speech systems use several input representations, including log-mel features and raw waveforms.
Real-World Implementation
Computing 40 log-mel filterbank features per frame as the input to a speech-to-text neural network
Using PLP features in noise-robust voice command systems for cars
Speaker recognition pipelines that rely on perceptually warped spectral features
Keyword spotting on low-power devices where compact filterbank features reduce computation
The Future of Filterbank and PLP Features
Learned front ends can replace or adapt parts of hand-designed feature pipelines, while mel filterbanks remain a common, interpretable baseline. Comparisons should distinguish representation quality from the model and training data using those features. Perceptual scales continue to inform audio-model design.
Putting it into practice
Why it matters
- Access and reach. It improves accessibility through transcription, narration, and voice interfaces.
- Cost and budget. Media teams can ship polished audio faster with smaller budgets.
- Speed and scale. Customer-facing systems can process spoken interactions at larger scale.
Watch out for
- Voice misuse and impersonation risks increase when consent is missing.
- Accuracy can drop across accents, dialects, or noisy environments.
- Synthetic audio can be mistaken for authentic speech without clear labeling.
Next steps
- Obtain explicit consent for voice capture, cloning, and reuse.
- Test quality across diverse speakers and background conditions.
- Define when a human must review or approve outputs.
- Label synthetic audio and keep provenance records for accountability.
Keep Exploring
Free newsletter
Keep up with AI in 3 minutes a day
One short email each weekday with the three AI stories that actually matter. Free forever, no ads.
One email each weekday. Unsubscribe in one click. We never sell or share your address.
Test yourself
Take the Filterbank and PLP Features quiz
Instant feedback on every answer, and a shareable certificate with a verifiable ID once you pass a course.
Support free AI education. AI Understanding is a 501(c)(3) nonprofit — no ads, no paywall, ever. Make a donation
Frequently asked questions
What is Filterbank and PLP Features?
Mel filterbank and Perceptual Linear Prediction (PLP) features summarize speech spectra using perceptually spaced frequency bands. Mel features pool short-time spectral energy into bands; PLP adds auditory weighting, loudness compression, and linear prediction. These are representations used by some speech systems, not guarantees of noise or speaker invariance.
Why do speech filterbanks use mel- or Bark-spaced bands instead of uniformly spaced hertz?
Mel and Bark scales model aspects of auditory frequency resolution; they do not guarantee noise removal or recognition accuracy.
What does the linear prediction step in PLP accomplish?
PLP fits an all-pole model to produce a smooth spectral envelope, capturing vocal-tract characteristics while suppressing speaker-dependent pitch harmonics.
Which feature type is the dominant input for modern deep speech recognition models?
Log-mel filterbank features are the standard, compact, perceptually meaningful input for today's deep speech models.
How do MFCCs differ from raw log-mel filterbank features?
MFCCs apply a discrete cosine transform on top of log-mel filterbank energies to decorrelate them into compact coefficients.
What perceptual element does PLP include that basic mel filterbanks do not?
PLP adds an equal-loudness preemphasis and cube-root intensity-to-loudness compression to better model human loudness perception.
Keep learning
Related guides
More guides picked for this topic