Realtime example of using SuperpoweredTimeStretching

[garynewby]

Any chance of posting a snippet outlining usage of SuperpoweredTimeStretching in the realtime audio callback?

I've tried utilizing the relevant parts of the SuperpoweredOfflineProcessingExample, but it looks like I need to incorporate a ring buffer as SuperpoweredTimeStretching produces a variable amount of samples and the audio buffer is fixed size. Using SuperpoweredAudiopointerList maybe the key but I don't understand it's usage from just the docs.

[superpoweredSDK]

Untested example code:

typedef struct liveTimeStretchingHandler {
    SuperpoweredTimeStretching *timeStretcher;
    SuperpoweredAudiopointerList *outputBuffers;
} liveTimeStretchingHandler;

static void setupLiveTimeStretching(liveTimeStretchingHandler *handler, int currentSamplerate) {
    handler->timeStretcher = new SuperpoweredTimeStretching(currentSamplerate);
    handler->outputBuffers = new SuperpoweredAudiopointerList(8, 16);
}

static void liveTimeStretchingProcess(float *outputBuffer, int numberOfSamples, int64_t *samplePositionOfTheFirstSampleInOutputBuffer, liveTimeStretchingHandler *handler) {
    while (numberOfSamples > 0) {
        // Try to fulfill the request from the output buffers first.
        int numberOfSamplesFromOutputBuffers = (numberOfSamples < handler->outputBuffers->sampleLength) ? numberOfSamples : handler->outputBuffers->sampleLength;

        if (handler->outputBuffers->makeSlice(0, numberOfSamplesFromOutputBuffers)) {
            *samplePositionOfTheFirstSampleInOutputBuffer = handler->outputBuffers->nextSamplePosition();
            numberOfSamplesFromOutputBuffers = 0;

            while (true) {
                int numSamples = 0;
                float *timeStretchedAudio = (float *)handler->outputBuffers->nextSliceItem(&numSamples);
                if (!timeStretchedAudio) break;

                // Incorporating this whole liveTimeStretchingProcess() method into your processing loop would be better,
                // as this memcpy could be done something useful instead of just copying.
                memcpy(outputBuffer, timeStretchedAudio, numSamples * 8);

                outputBuffer += numSamples * 2;
                numberOfSamplesFromOutputBuffers += numSamples;
            };

            handler->outputBuffers->truncate(numberOfSamplesFromOutputBuffers, true);
            numberOfSamples -= numberOfSamplesFromOutputBuffers;
            if (numberOfSamples < 1) break;
        }

        // There were not enough samples in the output buffers at this point. Let's feed the time stretcher.
        SuperpoweredAudiobufferlistElement inputBuffer;

        // This helps you tracking the current position in your music.
        // You can leave it at zero too, but samplePositionOfTheFirstSampleInOutputBuffer will not reflect the current position in this case.
        inputBuffer.samplePosition = 0;

        inputBuffer.startSample = 0;
        inputBuffer.samplesUsed = 0;
        inputBuffer.endSample = handler->timeStretcher->numberOfInputSamplesNeeded;
        inputBuffer.buffers[0] = SuperpoweredAudiobufferPool::getBuffer(handler->timeStretcher->numberOfInputSamplesNeeded * 8 + 64);
        inputBuffer.buffers[1] = inputBuffer.buffers[2] = inputBuffer.buffers[3] = NULL;

        // YOUR TODO HERE: Produce handler->timeStretcher->numberOfInputSamplesNeeded samples into inputBuffer.buffers[0], interleaved floating point audio.
        
        handler->timeStretcher->process(&inputBuffer, handler->outputBuffers);
    }
}

[garynewby]

That's awesome, thank you.

[AndrewsAlves]

What to do at
// YOUR TODO HERE: Produce handler->timeStretcher->numberOfInputSamplesNeeded samples into inputBuffer.buffers[0], interleaved floating point audio.

???

[LucasVanDongen]

I tried to adapt the example to make it work for me, but the only thing I hear is a static buzz like the 512 sample buffer loops over and over.

static bool audioProcessing(void *clientdata,
                            float **buffers,
                            unsigned int inputChannels,
                            unsigned int outputChannels,
                            unsigned int numberOfSamples,
                            unsigned int samplerate,
                            uint64_t hostTime) {
    __unsafe_unretained Superpowered *self = (__bridge Superpowered *)clientdata;
    //int64_t *samplePositionOfTheFirstSampleInOutputBuffer = 0; // Seems to be passed in
    //float outputBuffer[numberOfSamples * 2 + 16];
    float lalaBuffer[numberOfSamples * 2 + 16]; // Why do I need to do this intermediary step?
    SuperpoweredInterleave(buffers[0], buffers[1], lalaBuffer, numberOfSamples);
    float *outputBuffer = lalaBuffer;

    while (numberOfSamples > 0) {
        // Try to fulfill the request from the output buffers first.
        int numberOfSamplesFromOutputBuffers = (numberOfSamples < self->outputBuffers->sampleLength) ? numberOfSamples : self->outputBuffers->sampleLength;
        
        if (self->outputBuffers->makeSlice(0, numberOfSamplesFromOutputBuffers)) {
            //*samplePositionOfTheFirstSampleInOutputBuffer = self->outputBuffers->nextSamplePosition();
            numberOfSamplesFromOutputBuffers = 0;
            
            while (true) {
                int numSamples = 0;
                float *timeStretchedAudio = (float *)self->outputBuffers->nextSliceItem(&numSamples);
                if (!timeStretchedAudio) {
                    break;
                }
                
                // Incorporating this whole liveTimeStretchingProcess() method into your processing loop would be better,
                // as this memcpy could be done something useful instead of just copying.
                memcpy(outputBuffer, timeStretchedAudio, numSamples * 8);
                
                outputBuffer += numSamples * 2;
                numberOfSamplesFromOutputBuffers += numSamples;
            };
            
            self->outputBuffers->truncate(numberOfSamplesFromOutputBuffers, true);
            numberOfSamples -= numberOfSamplesFromOutputBuffers;
            if (numberOfSamples < 1) {
                break;
            }
        }
        
        // There were not enough samples in the output buffers at this point. Let's feed the time stretcher.
        SuperpoweredAudiobufferlistElement inputBuffer;
        
        // This helps you tracking the current position in your music.
        // You can leave it at zero too, but samplePositionOfTheFirstSampleInOutputBuffer will not reflect the current position in this case.
        inputBuffer.samplePosition = 0;
        
        inputBuffer.startSample = 0;
        inputBuffer.samplesUsed = 0;
        inputBuffer.endSample = self->timeStretcher->numberOfInputSamplesNeeded;
        inputBuffer.buffers[0] = SuperpoweredAudiobufferPool::getBuffer(self->timeStretcher->numberOfInputSamplesNeeded * 8 + 64);
        inputBuffer.buffers[1] = inputBuffer.buffers[2] = inputBuffer.buffers[3] = NULL;
        
        self->timeStretcher->process(&inputBuffer, self->outputBuffers);
        //self->timeStretcher->
        // YOUR TODO HERE: Produce handler->timeStretcher->numberOfInputSamplesNeeded samples into inputBuffer.buffers[0], interleaved floating point audio.
        //self->timeStretcher->numberOfInputSamplesNeeded
    }

    return true;
}

[temkich]

What to do at
// YOUR TODO HERE ?

if i do like this:

int samples = self->timeStretcher->numberOfInputSamplesNeeded;
        float *timeStretchedAudio = (float *)self->outputBuffers->nextSliceItem(&samples);
        if (timeStretchedAudio != 0) {
            SuperpoweredDeInterleave(timeStretchedAudio, buffers[0], buffers[1], numberOfSamples);
        }

in output no effect. what i do wrong?

[skopintsevDima]

Where I need to use methods setupLiveTimeStretching and liveTimeStretchingProcess?
Can you share example of implementaions of TODO mark?