Complete engine architectural refactoring: The monolithic IHFP module has been completely separated into logical, independent components: Pre-processing (source enhancement), Synthesis (separate modules for Harmonics, Textures, and Attacks), and Mixing (synthesis blending and post-effects).
New system: Flexible processing pipeline implemented. The execution order of DSP modules is no longer strictly fixed. Users can change the sequence of operations via drag-and-drop in the interface, enabling the creation of unique custom processing chains.
New DSP module: Harmonic Resonators. Added two modules ("General" and "Vocal") that analyze tonal peaks and generate musically related harmonics and subharmonics to enrich timbral texture.
New DSP module: Dynamic Boosters. Added "Clarity" and "Sibilants" boosters that function as dynamic equalizers, boosting target frequencies only when present in the source signal.
New DSP module: Contextual Tonal Warper. An experimental module that analyzes harmonic context and smoothly shifts the current frame's timbre to blend with surrounding frames, easing transitions.
New DSP module: HF Artifact Suppression ("Wart Remover"). A final cleaning stage that detects and suppresses isolated, unnatural HF anomalies ("warts") using median filtering based on spectral masks.
New DSP module: Upsampling Detector. Added a module that automatically scans files to determine their true source sample rate, even if they have been poorly upsampled.
Complete GUI redesign: The interface was rewritten from scratch. Five logically grouped columns are now used instead of four. Added file-specific preset management, a spectrogram preview widget, and playback controls.
New system: Localization system (i18n) added. The entire interface now supports translations, including English and Russian.
New system: Preset and app settings system introduced. Users can save, load, and delete presets. General application settings (language, thread count) are stored in "settings.json" and "presets.ini".
Version 0.8.0(13.08.2025)
Architectural refactoring: Automatic GUI generation implemented. All interface parameters and their metadata (ranges, descriptions, layout) have been moved to separate configuration files. The GUI is now built dynamically based on these files, making it much easier to add and modify settings.
New system: Dynamic HF synthesis threshold. The algorithm no longer uses fixed frequencies to start analysis and synthesis. Instead, it analyzes each time frame of the spectrogram, finds the point of natural HF roll-off in the source signal, and uses it as a dynamic boundary to start synthesis, making the processing significantly more adaptive.
Refactoring: Main processing loop simplified. The internal "hfp_core_synthesis_step" function was removed, and its logic was integrated directly into the main "iterative_hfp" loop to improve readability and eliminate redundant code.
Cleanup: Outdated parameters were removed in favor of more universal adaptive mechanisms.
Version 0.7.0(25.07.2025)
New module: MP3 artifact smoothing added. Implemented a new pre-processing stage that detects "gaps" and blocky anomalies in the HF spectrum typical of compressed files, and fills them with synthesized content based on surrounding audio.
New system: Synthesis density management introduced. The algorithm now analyzes the tonality of the LF signal and, based on the target "density" value, dynamically determines how many tonal "grains" (for harmonics) and textural "tiles" (for noise) need to be generated.
New system: HF-to-LF analysis feedback loop. If the algorithm detects an energy deficit in the upper HF spectrum during an iteration, it can adaptively boost the corresponding LF sources in the next iteration to compensate.
Refactoring: Centralized HF component placement logic. The function responsible for determining the target position for synthesized HF components has been moved to a shared module to eliminate code duplication across harmonics, textures, and attacks.
Version 0.6.0(27.06.2025)
New system: Adaptive processing system implemented. Parameters of certain post-effects (reverberation, spectral deconstruction, resonant filters) can now dynamically change based on the characteristics of the synthesized HF signal (such as density, peakiness, and brightness).
New system: Intelligent HF component placement. The system now avoids "collisions" (overlapping) of synthesized harmonics and textures, and takes the spectral centroid of the LF source into account for more musical HF placement.
New system: Weighted HF component mixing. Instead of simple linear addition, users can now control the weight (influence) of harmonics, textures, SBR, and attacks in the final HF signal.
New feature: Quiet LF source boost ("LF Source Boost") added. Allows raising quiet but important details in the LF range before they are used for HF synthesis, improving detail resolution.
New feature: Sub-bass contribution attenuation. Added attenuation of the contribution of very low (bass) frequencies to HF synthesis to prevent muddying the HF range with rumbling artifacts.
Refactoring: Major GUI code reorganization. The "MainWindow" logic was split into dedicated modules responsible for UI creation, state management, user action handling, and conversion process management to improve readability and maintenance.
Version 0.5.0(14.06.2025)
Major engine update: The analysis architecture has been fundamentally redesigned. Instead of a single low-pass filter, a more flexible system based on two parameters is now used: maximum frequency for LF analysis and minimum frequency for the start of HF synthesis.
New system: Multiband LF spectrum analysis. The algorithm can now divide the LF region into multiple sub-bands (with adjustable overlap), allowing for more precise and context-dependent source extraction for synthesis (harmonics, textures) from different frequency regions.
New system: Adaptive LF-to-HF mapping with adjustable aggressiveness. This allows controlling how high into the HF spectrum components from different parts of the LF range will be transposed.
Interface: The GUI was updated to support the new architecture. The old "lowpass_hz" slider was removed, and new controls for the number of LF sub-bands, their overlap, and mapping aggressiveness were added.
Version 0.4.0(04.06.2025)
Performance optimization: Key resource-intensive algorithms in the harmonics and textures synthesis modules have been rewritten and optimized using Numba for significantly faster processing.
Engine improvements: Deep recalibration of over 30 default parameters was conducted to improve "out-of-the-box" sound quality. Major changes include:
Significantly reduced the baseline attack synthesis strength for a more natural sound.
Modified spectral deconstruction balance (tonal component boost reduced from +6 dB to -3 dB).
Enabled mirroring of LF harmonics into the HF region by default for a denser sound.
Increased overall roll-off of the HF spectrum to reduce harshness.
Engine improvements: Refined the harmonic attenuation logic depending on the modulation mode ("cepstral" or "smoothed_lf"), making the modulation more predictable.
Version 0.3.0(23.05.2025)
Major engine update: The harmonic synthesis algorithm has been completely redesigned. Instead of simple overtone multiplication, a new "Timbral Tiling" method has been introduced. This method analyzes and transposes tonal grains from the LF region, preserving the original timbral characteristics more accurately.
Major engine update: The texture synthesis algorithm has been completely redesigned. The old noise modulation method has been replaced by "Spectral Tiling," which copies, transposes, and places noise fragments (tiles) from the upper LF spectrum into the HF region for more natural and context-dependent noise.
Improvement: Significantly improved the blending and transition algorithm between the original LF and synthesized HF. Added controllable crossfade based on the Hann window and global HF spectrum roll-off control for smoother and more natural integration.
Interface: Added new controls to the GUI for fine-tuning "timbral" and "spectral" tiling, as well as blending parameters.
Version 0.2.0(19.05.2025)
Refactoring: Conducted a major refactoring of the codebase. All DSP modules (Declipping, ATR, ASC, Remaster, IHFP) and configuration files have been split into separate, logically grouped files to improve readability, maintenance, and future extensibility.
New features: Added experimental parameters to the "Spectral Deconstruction" module, including adaptive noise reduction, ultrasonic attenuation, and transient temporal smoothing for finer post-processing.
Improvement: Adjusted default values for multiple IHFP parameters (e.g., number of iterations, feedback strength) to improve processing quality out-of-the-box.
Interface: Implemented "SettingsManager" for more reliable and centralized UI settings management, reducing potential errors when collecting and applying parameters.
Version 0.1.0(15.05.2025)
First alpha build of the program.
Basic functionality: Implemented loading and batch processing of ".wav" files in a separate thread.
Main processing engine: Introduced the Iterative High Frequency Reconstruction (IHFP) algorithm, which includes:
Harmonic synthesis based on overtones.
Texture synthesis based on modulated noise.
Attack synthesis and granular synthesis.
Waveshaping, FM synthesis, LPC analysis, and SBR modules.