J vs Sparse Neural FX Modeler

J-Chorus is a vintage BBD-style chorus for iPhone and iPad. Use it as a standalone audio app or insert the included AUv3 effect in a compatible host such as GarageBand. Its circuit-informed processing models the character of a 256-stage BBD path, op-amp modulation source, and compander. Three classic modes provide bright movement, deeper richness, or a combined wide stereo sound. Aging adds adjustable clock instability and noise for a more weathered response. FEATURES - Mode I, Mode II, and combined I+II - Adjustable Aging for BBD jitter and noise character - Equal-power Depth dry/wet blending - Adjustable input gain - Instrument and line input modes - Optional 2x and 4x oversampling - Standalone iOS/iPadOS app - AUv3 audio effect for compatible hosts TRY BEFORE UNLOCKING The app provides 10 minutes of full-featured active use, shared between the standalone app and AUv3 extension. Demo time advances only while audio is actively processing. A one-time, non-consumable In-App Purchase unlocks unlimited use in both formats. Restore Purchases is available in the standalone app. J-Chorus adds rich stereo movement to synthesizers, guitar arpeggios, electric piano, vocals, pads, and other melodic sources.

Plug in and play through real amplifiers, captured with neural networks and reproduced live on your iPhone. Amber Neural FX Modeler models the sound of tube amps sample by sample. Pick from 25 captures — chiming Vox, clean Fender, cranked Deluxe Reverb, vintage fuzz — and play them in real time with latency low enough to feel like the amp is in the room. THE PEDALBOARD Twenty blocks you arrange in any order: up to four neural amps in one chain, drive, tape echo with wow and flutter, shimmer reverb, compressor, gate, wah, phaser, flanger, chorus, tremolo, EQ, octave, delay, and a convolution reverb. Drag blocks to reorder the signal path. Save and load your rigs as presets. RECORD YOUR TAKES One tap records both a dry and a processed track straight into the Files app. BUILT ON RESEARCH The engine comes from a published paper on efficient real-time neural amp modeling — the same models, running entirely on your device. No ads. No accounts. No data collected. Plug in a guitar through any audio interface, or just use the mic. Capture names identify the amplifiers the models were measured from; no affiliation or endorsement is implied.