SInES Tools: JavaSTRAIGHT Voice Morpher
(Auditory morphing, vocal tract scaling & F0 manipulation based on TANDEM-STRAIGHT according to Kawahara et al. 1999)
Ready — Load Audio A or generate Demo

1. Load Sounds & F0 Editor

Load a voice (Audio A) for manipulation, or add Audio B for voice morphing.

YIN F0 Range (30–1400 Hz):
Supports speech as well as bass/soprano singing. If F0 drops out or octave-jumps, adjust Min/Max F0 and click Re-analyze.

Spectrogram Tools:
• Time Anchors: Click in A & B to align time points.
• Draw F0: Paint pitch contours directly.
• Erase F0: Make regions unvoiced.
Audio A Source / Base Voice
No audio file loaded
Audio B Morph Target (optional)
Optional for voice morphing between A and B
Spectrogram Mode:
Click in A & B to set a time anchor. Click "Draw F0 Contour" to paint pitch melodies.
YIN F0 Analysis Settings (Speech & Singing):
Minimum F0 (Hz)65 Hz
Maximum F0 (Hz)500 Hz
Time Anchor Alignment (A ↔ B): Load Audio A and B for DTW or manual anchor alignment.
2. Voice Presets

Pre-calibrated profiles for age/gender, vocal tract length (VTL), and voice quality. Select a preset from any of the three dropdown menus.
Quick-Select Profiles
Auto-calibrates F0 & VTL from input F0
Fixed relative F0 / formant shift
Modifies AP, intonation & tremor
3. Auditory Morphing (A ↔ B)

Independent or linked morphing of timing, fundamental frequency, spectral envelope, and aperiodicity between Voice A and B.

Optionally draw a dynamic temporal morph trajectory on the right.
Morphing Weights (0% = A, 100% = B)
Master Morph (all 4 linked)0%
Time / Rhythm0%
Fundamental Frequency (Log-F0)0%
Spectral Envelope / Timbre0%
Aperiodicity / Breathiness (AP)0%
Curve controls:
4. Voice Manipulation

Independent control of pitch (F0), intonation range, vocal tract length (formants), speaking rate, breathiness, jitter, and phase dispersion.
STRAIGHT Manipulation Parameters
Pitch Shift (F0 in semitones)0.0 st (1.00×)
Intonation Range (F0 Modulation)100%
Formant / Vocal Tract Scaling1.00×
Time Stretch (Tempo / Duration)1.00×
Aperiodicity / Breathiness Offset0 dB
Voice Jitter / Aging Tremor0%
All-Pass Group Delay Dispersion15%
Tip: Formant scaling < 1.00× lengthens the vocal tract (larger speaker); > 1.00× shortens it (smaller speaker).
5. Result & Export

Synthesize via the TANDEM-STRAIGHT Mixed-Excitation Vocoder.

In addition to the synthesized audio (.wav), the resulting F0 contour can also be exported as WAV or MIDI.
Output Synthesized / Morphed Result
No result synthesized yet TANDEM-STRAIGHT Spectral Envelope (0–8 kHz) + F0 Contour (Blue)
6. Batch Continuum Generator

Automatically generates graded stimulus continua (for PsychoPy, jsPsych, R, SPSS) and packs all WAV files plus stimulus_metadata.csv into a ZIP file.
Continuum Configuration
Quick Steps:
Dimensions scaled in Morph Continuum: (unchecked dimensions keep their current slider value)
Load Audio A (and optionally Audio B for A↔B morph continua) to generate a continuum ZIP.
DSP Parameters: YIN + Viterbi F0 Tracking (30–1400 Hz) | TANDEM-STRAIGHT Spectral Envelope | Mixed-Excitation Minimum-Phase Synthesis | Frame Hop: 5 ms | Sample Rate: 24000 Hz
JavaSTRAIGHT Voice Morpher © 2026 by Christoph Reuter and SInES (last modified: 15.11.2026) is released under MIT License
Translation of the Chinese version of the website: Li Hao
How to cite: Reuter, C. (2026). SInES Tools: JavaSTRAIGHT Voice Morpher. Web Application. https://sinestools.univie.ac.at/jstraight.htm