MP3 is not the same as MP4 so the Mp3FileReader can't play it. The next NAudio includes MediaFoundationReader which will be able to decode M4A. There is no public release yet, but if you want to build the latest from Source Code (click the tab above) you can use it.
(by the way, there is no need for pcmStream or baStream above. For MP3s, just use mp3Reader)
Mark
Thank you for your reply. I am going through the article that you have suggested.
But actually I was creating WasapiCapture instance for my audio capture device and I faced above issues. But I tried creating WaveIn instance, configuring it with the required WaveFormat (16 kbps, 44 kHz, 2 channels), capturing with this instance and is working fine. I am also able to record received audio samples using above mentioned codec.
Thank you so much for your help.
yes, but you will still be recording in PCM. For WMA you'd need to encode afterwards
Thanks I'll try
I thought MediaFoundationReader was only working on Windows 8 ? Does it works with Windows 7 ?
No, it will work on Windows 7 too. The Media Foundation API was actually introduced with Windows Vista.
I am trying to build some audio analysis tool using naudio. I find it too much playback oriented and I am having quite a lot of troubles.
I also believe it should be refactored better so that all core classes (audio stream definition, providers) can be really portable and then have all other things (ASIO, WM and so on) as other nuget packages.
I do not understand why winforms controls are in the library itself. I find the source code confusing and a lot of dependencies that are not needed. Is htere any plan to move to something like most of the other nuget packaes are doing?
I would also like some help I have some very trivial code that should be computing the db peak and RMS of a track.
The results are really wrong.
I am using directly the iff file reader since i do not have access to the SampleCount property via the audiofilereader.
Any help .. pretty sure I am doing something wrong
void Compute() {var audioFile = @"C:\Users\colombod\Downloads\Trees V2.aif";double[] audioMono;double[] audioLeft;double[] audioRight;int window;using (var stream = new NAudio.Wave.AiffFileReader(audioFile)) { Console.WriteLine("Channels {0} Rate{1} Bit{2}",stream.WaveFormat.Channels, stream.WaveFormat.SampleRate, stream.WaveFormat.BitsPerSample);var bytesPerSample = (stream.WaveFormat.BitsPerSample / 8); window = (20 *stream.WaveFormat.SampleRate)/1000; Console.WriteLine("Buffer size required {0}",window);var bufferSize = (int)(window* bytesPerSample * stream.WaveFormat.Channels);var dstSize = stream.SampleCount; audioMono = newdouble[dstSize]; audioLeft = newdouble[dstSize]; audioRight = newdouble[dstSize];var buffer = newbyte[bufferSize];var dst = 0;var read = 0;while ( ( read = stream.Read(buffer, 0, bufferSize) )> 0) {var sampleCount = (read / bytesPerSample) / stream.WaveFormat.Channels;for (int i = 0; i < sampleCount; i++) {var sum = 0.0;for (int j = 0; j < stream.WaveFormat.Channels; j++) {var off = (j*bytesPerSample);var sSample = BitConverter.ToInt16(buffer, i + off);var sample = sSample / ( sSample >= 0 ? 32767.0 : 32768.0);if(j == 0) { audioLeft[i + dst] = sample; }else { audioRight[i + dst] = sample; } sum += sample; } audioMono[i + dst] = sum / stream.WaveFormat.Channels; } dst += sampleCount; } } var rms = audioMono.Aggregate(0.0, (current, f) => current + (f*f)); rms = rms / audioMono.Length; rms = Math.Sqrt(rms); var decibel = 20 * Math.Log10(rms);var peak = 20 * Math.Log10(audioMono.Max());var rmsl = audioLeft.Aggregate(0.0, (current, f) => current + (f * f)); rmsl = rmsl / audioLeft.Length; rmsl = Math.Sqrt(rmsl); var decibell = 20 * Math.Log10(rmsl);var peakl = 20 * Math.Log10(audioLeft.Max());var rmsr = audioRight.Aggregate(0.0, (current, f) => current + (f * f)); rmsr = rmsr / audioRight.Length; rmsr = Math.Sqrt(rmsr); var decibelr = 20 * Math.Log10(rmsr);var peakr = 20 * Math.Log10(audioRight.Max()); }
Ok get it alpha2 ! Works fine with wav output.
What is best practice to output into a MemoryStream ? The WaveFileWriter.CreateWaveFile only takes a string (filename) à as parameter. I see code with loop/read but there might be a more straigh forward function ?
using (var reader = new MediaFoundationReader(file))using (var resampler = new MediaFoundationResampler(reader, 16000))using (var memoryStream = new MemoryStream()) { WaveFileWriter.CreateWaveFile(new MemoryStream(), resampler); }
It looks like your offsets are wrong. Samples are interleaved left, right etc. Try incrementing off by bytesPerSample every time you read a sample (off needs to be declared outside the top loop);
var sSample = BitConverter.ToInt16(buffer, off);
off+= bytesPerSample;
Hello,
I use NAudio to capture the loopback sound from my computer and I receive 32 bit sound according to the WaveFormat. Is there a way I can convert the received bytes from 32 bit to 16 bit? Or maybe to make WasapiLoopbackCapture generate 16 bit sound?
Here is what I tried:
privatevoid btnStart_Click(object sender, EventArgs e) { _waveIn = new WasapiLoopbackCapture(); _waveIn.DataAvailable += new EventHandler<WaveInEventArgs>(_waveIn_DataAvailable); _waveIn.StartRecording(); } void _waveIn_DataAvailable(object sender, WaveInEventArgs e) { byte[] newArray16Bit = newbyte[e.BytesRecorded / 2]; short two; int value; for (int i = 0, j = 0; i < e.BytesRecorded; i = i + 4, j = j + 2) { value = (BitConverter.ToInt32(e.Buffer, i)); two = (short)(value % short.MaxValue); newArray16Bit[j] = (byte)(two & 0xFF); newArray16Bit[j + 1] = (byte)((two >> 8) & 0xFF); } //do something with the newArray16Bit }
I receive a sound which is similar with what my computer is actually playing but it sounds like a am listening to a radio station with a very poor signal (the sound is not clean - you can nearly understand what the sound is about).
Does anybody know how can I clean up the sound?
Thanks!
I just changed the type of the 'two' variable from int to float and calculated it by multiplying the 'value' with short.MaxValue.
So I replace:
two = (short)(value % short.MaxValue);
with:
two = (short)(value * short.MaxValue);
Thanks!
yes, it is likely that your 32 bit incoming data is float not int. Multiplying by short.MaxValue is correct. It may just be that your recorded signal level is low.
Hello community,
With the following piece of code:
LoopStream inputStream;
if (fileName.EndsWith(".mp3" ))
{
var mp3Reader = new Mp3FileReader(fileName);
inputStream = new LoopStream(mp3Reader);
}
A friend of mine is receiving the error " AcmNotPossible calling acmStreamOpen"
The stack trace looks something like this:
at NAudio.Wave.AcmMp3FrameDecompressor..ctor(WaveFormat sourceFormat) at NAudio.Wave.Mp3FileReader.CreateAcmFrameDecompressor(WaveFormat mp3Format) at NAudio.Wave.Mp3FileReader..ctor(Stream inputStream, FrameDecompressorBuilder
frameDecompressorBuilder) at NAudio.Wave.Mp3FileReader..ctor(Stream inputStream) at NAudio.Wave.Mp3FileReader..ctor(String mp3FileName) at Dream.World.AssetManager.CreateInputStream(String fileName)
Does anyone have any ideas?
It means you don't have an ACM codec installed that can decompress MP3. What operating system are you running on?