0

我已经实现了许多 SuperpoweredFX 类并且没有遇到任何问题，但是，在使用 SuperpoweredTimeStretching 实现音高变换器时，我遇到了一些麻烦。

    static bool audioProcessing(void * __unused clientdata, short int *audioInputOutput, int numberOfSamples, int __unused samplerate) {

    float *floatBuffer = (float *)malloc(numberOfSamples * 2 * sizeof(float) + 1024);

    SuperpoweredShortIntToFloat(audioInputOutput, floatBuffer, numberOfSamples);

    if(isFX_PitchShifter_1){
        __android_log_print(ANDROID_LOG_ERROR, "FrequencyDomain", "audioProcessing - isFX_PitchShifter_1 - fx_PitchShift_1");    
        SuperpoweredAudiopointerList *outputBuffers = new SuperpoweredAudiopointerList( 4 , 0 );//POSSIBLE MEMORY PROBLEM   
        SuperpoweredAudiobufferlistElement inputBuffer;
        inputBuffer.startSample = 0;
        inputBuffer.samplesUsed = 0;
        inputBuffer.endSample = numberOfSamples;
        inputBuffer.buffers[0] = SuperpoweredAudiobufferPool::getBuffer((unsigned int) (numberOfSamples * 8 + 64));           
        inputBuffer.buffers[1] = inputBuffer.buffers[2] = inputBuffer.buffers[3] = NULL;            
        SuperpoweredShortIntToFloat(audioInputOutput, (float *)inputBuffer.buffers[0], (unsigned int) numberOfSamples);

        if(fx_PitchShift_1!=NULL){
            fx_PitchShift_1->process(&inputBuffer, outputBuffers);

            if (outputBuffers->makeSlice(0, outputBuffers->sampleLength)) {
                while (true) { // Iterate on every output slice.                      
                    int numSamples = 0;
                    float *temp_floatBuffer = (float *)outputBuffers->nextSliceItem(&numSamples);
                    if (!temp_floatBuffer) {
                        break;
                    }

                    SuperpoweredFloatToShortInt(temp_floatBuffer, audioInputOutput, numSamples);

                }                   
                outputBuffers->clear();
            }

        }

        SuperpoweredShortIntToFloat(audioInputOutput, floatBuffer, numberOfSamples);    

    }           
    SuperpoweredFloatToShortInt(floatBuffer, audioInputOutput, numberOfSamples);      
    return true;
}

以上是我的处理函数，我使用的是FrequencyDomain示例中的SuperpoweredAndroidAudioIO。

我像这样声明 SuperpoweredTimeStretching 类：

fx_PitchShift_1 = new SuperpoweredTimeStretching((unsigned int)thisSampleRate);
fx_PitchShift_1->setRateAndPitchShift(1.0f, (int)pitchShiftOctaves);

同一个应用程序使用 SuperpoweredFX 类就好了，但是，当音高变化被投入到混音中时，输出变得一团糟。可以在此处找到示例文件：https ://drive.google.com/file/d/1Hy52SsgOYxlkxFq5CCsK5sCKX3Y_9G8J/view?usp=sharing

我这样计算采样率：

samplerateString = null;
    buffersizeString = null;
    if (Build.VERSION.SDK_INT >= 17) {
        AudioManager audioManager = (AudioManager) context.getSystemService(Context.AUDIO_SERVICE);
        samplerateString = audioManager.getProperty(AudioManager.PROPERTY_OUTPUT_SAMPLE_RATE);
        buffersizeString = audioManager.getProperty(AudioManager.PROPERTY_OUTPUT_FRAMES_PER_BUFFER);
    }
    if (samplerateString == null) samplerateString = "44100";
    if (buffersizeString == null) buffersizeString = "512";
4

1 回答 1

0

音高移位器在 FFT 大小为 2048 的频域中工作。这与您当前的缓冲区大小不一致。

您还可以在音频处理回调的每一轮创建一个新的输出缓冲区列表，然后丢弃音高转换器提供的所有内容。

您正确地将所有输入样本提交到音高移位器，但您需要全局处理输出列表。它将存储音高移位器的输出，如果输出列表中有可用的样本数量，您可以删除输出所需的样本数量。

于 2018-02-27T13:16:09.913 回答