pyCryptoChart2Wav — Chart-to-Audio Tools

Convert cryptocurrency price charts into audio waveforms and synthesizer wavetables.

This project includes two main tools:

1. chart2wav - Command-line chart-to-WAV converter (works on desktop and Android)
2. Chart FM Synth Pro - Full-featured FM synthesizer with chart-to-wavetable import (desktop only)

---

📊 chart2wav — Chart-to-Waveform Converter

chart2wav is a lightweight Python tool that converts PNG price charts into audio waveforms, ready for samplers, synths, DAWs and wavetable engines.

It extracts the blue chart line from your screenshot, converts it into amplitude data, shapes it into a clean audio waveform, and exports a 16-bit mono WAV.

No external dependencies beyond Pillow.

Supports three synthesis modes:

---

🎛 Modes

1. cycle (default)

Turns the chart into a single-cycle oscillator (like Serum/PPG/DW style wavetable waves).
Useful for synth engines and looped tones.

2. sweep

Stretches the entire chart across the output duration.
A one-shot gesture of the full price motion.

3. multi

Carves the chart into N sequential cycles, scanning from left → right.
Produces evolving multi-cycle patterns rather than a single repeated shape.

---

✨ Features

• PNG → waveform extraction based only on the DEXSCREENER chart’s blue color (#2962FF by default)[NOTE: to work on another chart with a different color line, you must update this hex color value]
• Baseline (dashed center line) auto-detected
• Missing data automatically interpolated
• Loop-safe output
  • endpoint ramp removal
  • symmetric cosine fade
• Three waveform modes (cycle / sweep / multi)
• User-defined number of cycles in multi-mode
• Fixed-duration output for sampler compatibility (Digitakt, MPC, Octatrack, etc.)
• Python 2.7 compatible
• No dependencies except Pillow

---

📦 Requirements

Python 3 (recommended) - Works on desktop and Android:

• Python 3.6+
• Pillow (pip install pillow)

Python 2.7 (legacy) - Desktop only:

• Python 2.7
• Pillow (pip install pillow)

---

🚀 Usage

Python 3 (recommended):

python3 chart2wav.py chart.png chart.wav

Python 2.7 (legacy):

python chart_png_to_wave.py chart.png chart.wav

Advanced Options

python3 chart2wav.py chart.png output.wav \
  --mode cycle \
  --points 256 \
  --samplerate 48000 \
  --seconds 1.0 \
  --fade 0.05 \
  --blue 2962FF \
  --thresh 35

Options:

• --mode : cycle, sweep, or multi
• --points : samples per cycle (default: 64)
• --cycles : number of cycles for multi mode (default: 8)
• --samplerate : output sample rate (default: 44100)
• --seconds : duration in seconds (default: 1.0)
• --fade : fade amount 0.0-0.5 (default: 0.05)
• --blue : hex color to detect (default: 2962FF)
• --thresh : color matching threshold (default: 35)

---

🎹 Chart FM Synth Pro — Full-Featured FM Synthesizer

A complete FM synthesizer application that imports cryptocurrency chart images as custom wavetables for sound design.

Platform: Desktop only (Linux, Mac, Windows) - Requires PyQt5 which is not available on Android/Pydroid3.

✨ Features

Synthesis Engine

• 6 FM operators with chart-derived wavetables
• 32 classic FM algorithms (DX7-inspired) + custom matrix routing
• ADSR envelopes per operator + master amplitude envelope
• 3 LFOs with 6 waveform shapes (sine, triangle, saw up/down, square, sample & hold)
• Multi-mode filter (lowpass, highpass, bandpass) with resonance
• Polyphonic voice management (up to 8 voices)
• Real-time audio with Numba JIT optimization (~50x faster DSP)

Chart Integration

• Import PNG charts directly from the UI
• 3 conversion modes (cycle, sweep, multi)
• Automatic wavetable generation from price action
• Preview and adjust conversion parameters
• Preset configurations for common use cases

User Interface

• 49-key on-screen keyboard with velocity sensitivity
• Real-time oscilloscope and spectrum analyzer
• Hardware-inspired dark UI with rotary knobs
• Patch save/load system (JSON format)
• MIDI support (note on/off, mod wheel, pitch bend, aftertouch)
• Mod matrix for flexible modulation routing

📦 Requirements

Required:

pip install sounddevice numpy scipy pyqt5 pillow

Optional (for MIDI support):

pip install mido python-rtmidi

Optional (for performance boost):

pip install numba

🚀 Usage

python3 chart_fm_synth_pro.py

First Steps

1. Import chart WAVs or create them using chart2wav
2. Load wavetables via "Chart Import" button
3. Select algorithm (1-32) or use custom matrix
4. Assign wavetables to operators
5. Play notes on the on-screen keyboard or via MIDI
6. Adjust envelopes, LFOs, and filter to shape the sound
7. Save patches for later use

Chart Import Workflow

1. Click "Chart Import" button
2. Browse and select a chart PNG
3. Choose conversion mode:
   • Cycle - Single-cycle wavetable (recommended for most synth work)
   • Sweep - Full chart as one-shot gesture
   • Multi - Multiple evolving cycles
4. Adjust parameters (points, cycles, duration)
5. Click "Import"
6. The wavetable becomes available in operator dropdowns

Keyboard Shortcuts

• Z-M keys - Bottom octave (C-B)
• Q-U keys - Middle octave (C-B)
• Arrow keys - Change octave
• Space - Panic (all notes off)

MIDI Support

If mido is installed, the synth automatically connects to the first available MIDI input port and responds to:

• Note on/off messages
• Control Change 1 (Mod Wheel)
• Pitch bend
• Channel aftertouch

🎛 FM Algorithms

The synth includes 32 DX7-style FM algorithms. Each algorithm defines how the 6 operators modulate each other:

• Algorithm 1-5: Classic stacks (6→5→4→3→2→1 to various splits)
• Algorithm 6-11: Parallel carriers with modulators
• Algorithm 12: All 6 operators in parallel (additive)
• Algorithm 13-20: Mixed configurations
• Algorithm 21-32: Complex routings for evolving timbres

You can also create custom matrix routing by enabling "Custom Matrix" mode and adjusting the modulation amounts directly.

🔊 Audio Backend

• Low-latency real-time audio via sounddevice (PortAudio)
• Callback-based processing (1024 sample blocks)
• Configurable latency settings for stability

📄 Patch Format

Patches are saved as JSON files containing:

• Operator parameters (level, ratio, detune, feedback, envelope)
• Algorithm or custom matrix configuration
• Master settings (level, FM depth, portamento, tuning)
• LFO settings (rate, depth, shape, delay)
• Filter settings (type, cutoff, resonance)
• Modulation matrix entries

Example patch structure:

{
  "name": "My Patch",
  "algorithm": 1,
  "operators": [...],
  "master_level": 0.7,
  "fm_depth": 0.5,
  "lfos": [...],
  "filter": {...}
}

🎨 Sound Design Tips

Creating Evolving Pads

• Use multi-mode chart import with 8-16 cycles
• Slow attack/release envelopes on operators
• LFO modulation on operator levels
• Low FM depth (0.3-0.5)

Creating Aggressive Basses

• Use cycle mode with sharp chart movements
• High FM depth (0.7-1.0)
• Fast attack, medium release
• Filter with resonance

Creating Bell Tones

• Use Algorithm 12 (all parallel)
• Enable 2-3 operators only
• Different ratios (1.0, 2.0, 3.5)
• Fast attack, long release

Creating Plucks

• Use cycle mode with volatile chart patterns
• Very fast attack (0.001s)
• Short decay/release
• Higher operator ratios (2.0, 4.0, 8.0)

---

🛠 Technical Details

DSP Optimization

• Numba JIT compilation for critical DSP kernels (~50x speedup)
• Block processing (1024 samples) for efficiency
• Pre-allocated buffers to minimize memory allocation
• Vectorized operations using NumPy

Chart-to-Wavetable Pipeline

1. Load PNG image
2. Color-based pixel detection (Euclidean RGB distance)
3. Baseline detection (row with most chart pixels)
4. Per-column median Y extraction
5. Linear interpolation for missing data
6. DC offset removal (linear ramp subtraction)
7. Cosine window fade at endpoints
8. Normalization to [-1, 1]
9. Resampling to target points/duration

Voice Architecture

• Polyphonic voice allocator with voice stealing
• Per-voice operator states (phase, envelope, feedback)
• Per-voice filter and LFO instances
• Portamento with linear frequency interpolation

---

📜 License

See LICENSE file for details.

🙏 Credits

• FM algorithms inspired by Yamaha DX7
• Chart2wav conversion based on cryptocurrency chart analysis
• Built with PyQt5, NumPy, SciPy, and Numba

---

🐛 Troubleshooting

"No audio output"

• Check sounddevice installation: pip install sounddevice
• Verify audio device: python3 -c "import sounddevice; print(sounddevice.query_devices())"
• Try increasing buffer size or latency in the code

"Numba warnings"

• Numba is optional - synth works without it (just slower)
• Install for better performance: pip install numba

"MIDI not working"

• Install mido: pip install mido python-rtmidi
• Check available MIDI devices
• On Linux, may need to install libasound2-dev

"Chart import fails"

• Ensure Pillow is installed: pip install pillow
• Check that chart has visible blue line (#2962FF)
• Try adjusting threshold (25-50 range)
• Different chart colors need --blue RRGGBB parameter

"Can I run the FM synth on Android?"

• No - PyQt5 is not available on Android/Pydroid3
• Use the command-line chart2wav tool instead to create wavetables on Android
• Then load those WAVs into desktop synthesizers or DAWs