Espressif DSP Library help

[mikemoy]

I apologize if that is how my statement came off. It's intention was just to explain I am under the gun on something I know very little about, and could use some help in a big way is all.

You'r expiation was very helpful. Embarrassing obvious now that it was pointed out. If I take the fft example as a starting point. https://github.com/espressif/esp-dsp/bl ... fft_main.c

My A/D data is in bufferA[]. Would I just copy my buffer into x1: memcpy(x1, bufferA, N);
Replacing that with dsps_tone_gen_f32(x1, N, 1.0, 0.16, 0);

FWIW, I realize later in the example it will mix my signal with x2, but just want to start simple with injecting a 1k tone.

Code: Select all

void app_main()
{
    esp_err_t ret;
    ESP_LOGI(TAG, "Start Example.");
    ret = dsps_fft2r_init_fc32(NULL, CONFIG_DSP_MAX_FFT_SIZE);
    if (ret  != ESP_OK)
    {
        ESP_LOGE(TAG, "Not possible to initialize FFT. Error = %i", ret);
        return;
    }

    // Generate hann window
    dsps_wind_hann_f32(wind, N);
    // Generate input signal for x1 A=1 , F=0.1
    dsps_tone_gen_f32(x1, N, 1.0, 0.16,  0);
    // Generate input signal for x2 A=0.1,F=0.2
    dsps_tone_gen_f32(x2, N, 0.1, 0.2, 0);

[mikemoy]

Seems pretty trivial to me: the functions take int16_t types

FWIW, in the example this function " dsps_tone_gen_f32(x1, N, 1.0, 0.16, 0); "
fills x1 with float values from -1.0 to 1.0. Why do you say it takes int16_t types ?

[ESP_Sprite]

That function (dsps_fft2r_fc32) does takes 32-bit floats as the name implies. The function that is used in the other example (dsps_fft2r_sc16_ae32) uses 16-bit signed numbers.

[mikemoy]

I got it working. Thank you guys for the help!

[ESP_Sprite]

You're welcome, good luck with the project!