Variable convolveConst

convolve: MultipassFilter<
    {
        bias: {
            default: 0;
            description: "Added to the result, useful for kernels whose weights add up to zero (edge detectors).";
            label: "Bias";
            max: 1;
            min: -1;
            step: 0.01;
            type: "float";
        };
        kernel: {
            default: (0 | 1)[];
            description: "The weights. The centre cell is the pixel itself.";
            label: "Kernel";
            size: 5;
            type: "kernel";
        };
        normalize: {
            default: true;
            description: "Divide by the sum of the weights, so a blur keeps the overall brightness.";
            label: "Normalise by sum";
            type: "bool";
        };
    },
> = ...