[MODEL] ddspXsynth

[dsuedholt]

A brief description of what your model does

A DDSP timbre transfer model trained on synth data that performs cross-synthesis between the input signal and the DDSP output, trained on the Helm synthesizer.

Checklist

• [x] I have checked the model works properly loaded locally in the Neutone plugin on my machine. MacOS Ventura 13.1, Reaper 6.69,
• [x] I have uploaded the .nm model file at a publicly available location: https://github.com/dsuedholt/ddsp_xsynth/blob/master/neutone_models/ddspXsynth.nm

Extra information

More info in the readme file at https://github.com/dsuedholt/ddsp_xsynth

Metadata

The model export function should dump a json file. Please paste the contents here for review and discussions.

{
    "model_name": "ddspXsynth",
    "model_authors": [
        "David S\u00fcdholt"
    ],
    "model_version": "0.1.0",
    "model_short_description": "DDSP cross-synthesis model trained on synth data",
    "model_long_description": "A DDSP timbre transfer model trained on synth data that performs cross-synthesis between the input signal and the DDSP output.",
    "technical_description": "A DDSP timbre transfer model trained on synth data that performs cross-synthesis between the input signal and the DDSP output.",
    "technical_links": {},
    "tags": [
        "timbre transfer",
        "DDSP",
        "cross synthesis",
        "synth"
    ],
    "citation": "\n    Engel, J., Hantrakul, L., Gu, C., & Roberts, A. (2020). DDSP: Differentiable Digital Signal Processing. ICLR.\n    ",
    "is_experimental": true,
    "neutone_parameters": {
        "p1": {
            "name": "Pitch Shift",
            "description": "Apply pitch shift (-24 to +24 semitones)",
            "type": "knob",
            "used": "True",
            "default_value": "0.5"
        },
        "p2": {
            "name": "Cross-Synth",
            "description": "How much the input signal is mixed into harmonics",
            "type": "knob",
            "used": "True",
            "default_value": "0.5"
        },
        "p3": {
            "name": "Noise",
            "description": "Volume of the filter noise",
            "type": "knob",
            "used": "True",
            "default_value": "0.5"
        },
        "p4": {
            "name": "Formant",
            "description": "Formant Shift",
            "type": "knob",
            "used": "True",
            "default_value": "0.5"
        }
    },
    "wet_default_value": 1.0,
    "dry_default_value": 0.0,
    "input_gain_default_value": 0.5,
    "output_gain_default_value": 0.5,
    "is_input_mono": true,
    "is_output_mono": true,
    "model_type": "mono-mono",
    "native_sample_rates": [
        48000
    ],
    "native_buffer_sizes": [
        2048
    ],
    "look_behind_samples": 0,
    "sdk_version": "1.1.2",
    "pytorch_version": "1.11.0",
    "date_created": 1678841309.141696
}

[bogdanteleaga]

Thank you David! We included your models with the new release, please check the website and the plugin and let me know if you notice any problems.

We can automatically update it on new releases of the SDK so you don't need to worry about that part, just let us know if you want to update the model itself sometime in the future.