change wav file ( to 16KHz and 8bit ) with using NAudio - c#

I want to change a WAV file to 8KHz and 8bit using NAudio.
WaveFormat format1 = new WaveFormat(8000, 8, 1);
byte[] waveByte = HelperClass.ReadFully(File.OpenRead(wavFile));
Wave
using (WaveFileWriter writer = new WaveFileWriter(outputFile, format1))
{
writer.WriteData(waveByte, 0, waveByte.Length);
}
but when I play the output file, the sound is only sizzle. Is my code is correct or what is wrong?
If I set WaveFormat to WaveFormat(44100, 16, 1), it works fine.
Thanks.

A few pointers:
You need to use a WaveFormatConversionStream to actually convert from one sample rate / bit depth to another - you are just putting the original audio into the new file with the wrong wave format.
You may also need to convert in two steps - first changing the sample rate, then changing the bit depth / channel count. This is because the underlying ACM codecs can't always do the conversion you want in a single step.
You should use WaveFileReader to read your input file - you only want the actual audio data part of the file to get converted, but you are currently copying everything including the RIFF chunks as though they were audio data into the new file.
8 bit PCM audio usually sounds horrible. Use 16 bit, or if you must have 8 bit, use G.711 u-law or a-law
Downsampling audio can result in aliasing. To do it well you need to implement a low-pass filter first. This unfortunately isn't easy, but there are sites that help you generate the coefficients for a Chebyshev low pass filter for the specific downsampling you are doing.
Here's some example code showing how to convert from one format to another. Remember that you might need to do the conversion in multiple steps depending on the format of your input file:
using (var reader = new WaveFileReader("input.wav"))
{
var newFormat = new WaveFormat(8000, 16, 1);
using (var conversionStream = new WaveFormatConversionStream(newFormat, reader))
{
WaveFileWriter.CreateWaveFile("output.wav", conversionStream);
}
}

The following code solved my problem dealing with G.711 Mu-Law with a vox file extension to wav file. I kept getting a "No RIFF Header" error with WaveFileReader otherwise.
FileStream fileStream = new FileStream(fileName, FileMode.Open);
var waveFormat = WaveFormat.CreateMuLawFormat(8000, 1);
var reader = new RawSourceWaveStream(fileStream, waveFormat);
using (WaveStream convertedStream = WaveFormatConversionStream.CreatePcmStream(reader))
{
WaveFileWriter.CreateWaveFile(fileName.Replace("vox", "wav"), convertedStream);
}
fileStream.Close();

openFileDialog openFileDialog = new openFileDialog();
openFileDialog.Filter = "Wave Files (*.wav)|*.wav|All Files (*.*)|*.*";
openFileDialog.FilterIndex = 1;
WaveFileReader reader = new NAudio.Wave.WaveFileReader(dpmFileDestPath);
WaveFormat newFormat = new WaveFormat(8000, 16, 1);
WaveFormatConversionStream str = new WaveFormatConversionStream(newFormat, reader);
try
{
WaveFileWriter.CreateWaveFile("C:\\Konvertierten_Dateien.wav", str);
}
catch (Exception ex)
{
MessageBox.Show(String.Format("{0}", ex.Message));
}
finally
{
str.Close();
}
MessageBox.Show("Konvertieren ist Fertig!");
}

Related

Converting .vox audio files to .wav or .mp3 with NAudio

I am trying to convert .vox to .mp3 or .wav with NAudio with the code below:-
var bytes = GetBytes(new FileInfo(#"D:\path\to\vox-file.vox"));
using (var writer = new WaveFileWriter(#"D:\path\to\wav-file.wav", new WaveFormat(8000, 8, 2)))
{
writer.Write(bytes, 0, bytes.Length);
}
However the converted .wav file I am getting has very distorted audio and its not listenable. Am I missing some configuration with NAudio?
I was able to convert .vox files to .wav with the following:-
var bytes = GetBytes(new FileInfo(#"path-to.vox"));
using (var writer = new WaveFileWriter(#"path-to.wav", Mp3WaveFormat.CreateMuLawFormat(8000, 1)))
{
writer.Write(bytes, 0, bytes.Length);
}
The important bit here is this part: Mp3WaveFormat.CreateMuLawFormat(8000, 1), this is basically some kind of format/configuration for the .wav file which you want to be created and these values may vary depending upon your .vox files

Naudio float Sample and Concentus.OggFile short Sample, convert mp3 to opus ogg

trying to convert mp3 file to opus ogg file by using
NAudio: https://github.com/naudio/NAudio
Concentus.OggFile https://github.com/lostromb/concentus.oggfile
using (var source = new MemoryStream(mp3File))
using (var mp3Reader = new MyAudioFileReader(source, FileReaderType.Mp3))
using (var memo = new MemoryStream())
{
var bufferFloat = new float[mp3Reader.Length / (mp3Reader.WaveFormat.BitsPerSample / 8)];
var count = mp3Reader.Read(bufferFloat, 0, bufferFloat.Length);
//convert float to short
var buffShort = new short[count];
var scale = (float)(short.MaxValue);
for (int i = 0; i < count; i++)
{
buffShort[i] = (short)(bufferFloat[i] * scale);
}
//encoder
var encoder = OpusEncoder.Create(48000,
mp3Reader.WaveFormat.Channels,
OpusApplication.OPUS_APPLICATION_AUDIO);
encoder.Bitrate = 65536;//64kbps
//tags
var tags = new OpusTags();
tags.Fields[OpusTagName.Title] = "Title";
tags.Fields[OpusTagName.Artist] = "Artist";
//
var oggOut = new OpusOggWriteStream(encoder, memo, tags);
oggOut.WriteSamples(buffShort, 0, buffShort.Length);
oggOut.Finish();
result = memo.ToArray();
}
I don't know the basics, did some GDD and here is result of what I get.
OpusOggWriteStream.WriteSamples()
requires short[] sample as input.
is it okay the way I convert the NAudio float[] sample provider to short[] ?
probably not cuz output file can't be played .
this code doesn't work and I have no idea why :"D
This is probably too little too late but whatever. As far as I can tell your code looks fine, so to debug I would try just a few things:
There is actually a WriteSamples() overload in OggOpusWriteStream that accepts float[]. Try using that first
I would make sure that mp3Reader.Read actually produces as much data as you believe it to be. I wonder if it might only be returning a single frame of decoded data or something like that. Try writing out the data as uncompressed pcm and sanity checking it
I checked to see if there was some bug in Concentus.Oggfile since you only ever WriteSamples() once - I thought maybe it wouldn't finalize the pages properly in that case, but I can't find anything.

Append WAV Header in NAudio

I am trying to convert audio MP3 files to WAV with a standard rate (48 KHz, 16 bits, 2 channels) by opening with "MediaFoundationReaderRT" and specifying the standard settings in it.
After the file is converted to PCM WAV, when I try to play the WAV file, it gives corrupt output:
Option 1 -
WaveStream activeStream = new MediaFoundationReaderRT([Open "MyFile.mp3"]);
WaveChannel32 waveformInputStream = new WaveChannel32(activeStream);
waveformInputStream.Sample += inputStream_Sample;
I noticed that if I read the audio data into a memory stream (wherein it appends the WAV header via "WaveFileWriter"), then things work fine:
Option 2 -
WaveStream activeStream = new MediaFoundationReaderRT([Open "MyFile.mp3"]);
MemoryStream memStr = new MemoryStream();
byte[] audioData = new byte[activeStream.Length];
int bytesRead = activeStream.Read(audioData, 0, audioData.Length);
memStr.Write(audioData, 0, bytesRead);
WaveFileWriter.CreateWaveFile(memStr, audioData);
RawSourceWaveStream rawSrcWavStr = new RawSourceWaveStream(activeStream,
new WaveFormat(48000, 16, 2));
WaveChannel32 waveformInputStream = new WaveChannel32(rawSrcWavStr);
waveformInputStream.Sample += inputStream_Sample;
However, reading the whole audio into memory is time-consuming. Hence I am looking at "Option 1" as noted above.
I am trying to figure out as to what exactly is the issue. Is it that the WAV header is missing which is causing the problem?
Is there a way in "Option 1" where I can append the WAV header to the "current playing" sample data, instead of converting the whole audio data into memory stream and then appending the header?
I'm not quite sure why you need either of those options. Converting an MP3 file to WAV is quite simple with NAudio:
using(var reader = new MediaFoundationReader("input.mp3"))
{
WaveFileWriter.CreateWaveFile("output.wav", reader);
}
And if you don't need to create a WAV file, then your job is already done - MediaFoundationReader already returns PCM from it's Read method so you can play it directly.

Playing a wav file after processing it with NAudio

I've converted a double array output[] to a wav file using NAudio. The file plays ok in VLC player and Windows Media Player, but when I try to open it in Winamp, or access it in Matlab using wavread() I fail.. (in Matlab I get the error: " Invalid Wave File. Reason: Incorrect chunk size information in WAV file." , which pretty obviously means something's wrong with the header). Any ideas on how to solve this? Here's the code for converting the array to a WAV:
float[] floatOutput = output.Select(s => (float)s).ToArray();
WaveFormat waveFormat = new WaveFormat(16000, 16, 1);
WaveFileWriter writer = new WaveFileWriter("C:\\track1.wav", waveFormat);
writer.WriteSamples(floatOutput, 0, floatOutput.Length);
You must dispose your WaveFileWriter so it can properly fix up the WAV file header. A using statement is the best way to do this:
float[] floatOutput = output.Select(s => (float)s).ToArray();
WaveFormat waveFormat = new WaveFormat(16000, 16, 1);
using (WaveFileWriter writer = new WaveFileWriter("C:\\track1.wav", waveFormat))
{
writer.WriteSamples(floatOutput, 0, floatOutput.Length);
}

change format from wav to mp3 in memory stream in NAudio

Hi there iam trying to convert text to speech (wav) in the memorystream convert it to mp3 and then play it on the users page.so need i help what to do next?
here is my asmx code :
[WebMethod]
public byte[] StartSpeak(string Word)
{
MemoryStream ms = new MemoryStream();
using (System.Speech.Synthesis.SpeechSynthesizer synhesizer = new System.Speech.Synthesis.SpeechSynthesizer())
{
synhesizer.SelectVoiceByHints(System.Speech.Synthesis.VoiceGender.NotSet, System.Speech.Synthesis.VoiceAge.NotSet, 0, new System.Globalization.CultureInfo("en-US"));
synhesizer.SetOutputToWaveStream(ms);
synhesizer.Speak(Word);
}
return ms.ToArray();
}
Thanks.
Just wanted to post my example too using NAudio.Lame:
NuGet:
Install-Package NAudio.Lame
Code Snip: Mine obviously returns a byte[] - I have a separate save to disk method b/c I think it makes unit testing easier.
public static byte[] ConvertWavToMp3(byte[] wavFile)
{
using(var retMs = new MemoryStream())
using (var ms = new MemoryStream(wavFile))
using(var rdr = new WaveFileReader(ms))
using (var wtr = new LameMP3FileWriter(retMs, rdr.WaveFormat, 128))
{
rdr.CopyTo(wtr);
return retMs.ToArray();
}
}
You need an MP3 compressor library. I use Lame via the Yeti Lame wrapper. You can find code and a sample project here.
Steps to get this working:
Copy the following files from MP3Compressor to your project:
AudioWriters.cs
Lame.cs
Lame_enc.dll
Mp3Writer.cs
Mp3WriterConfig.cs
WaveNative.cs
WriterConfig.cs
In the project properties for Lame_enc.dll set the Copy to Output property to Copy if newer or Copy always.
Edit Lame.cs and replace all instances of:
[DllImport("Lame_enc.dll")]
with:
[DllImport("Lame_enc.dll", CallingConvention = CallingConvention.Cdecl)]
Add the following code to your project:
public static Byte[] WavToMP3(byte[] wavFile)
{
using (MemoryStream source = new MemoryStream(wavFile))
using (NAudio.Wave.WaveFileReader rdr = new NAudio.Wave.WaveFileReader(source))
{
WaveLib.WaveFormat fmt = new WaveLib.WaveFormat(rdr.WaveFormat.SampleRate, rdr.WaveFormat.BitsPerSample, rdr.WaveFormat.Channels);
// convert to MP3 at 96kbit/sec...
Yeti.Lame.BE_CONFIG conf = new Yeti.Lame.BE_CONFIG(fmt, 96);
// Allocate a 1-second buffer
int blen = rdr.WaveFormat.AverageBytesPerSecond;
byte[] buffer = new byte[blen];
// Do conversion
using (MemoryStream output = new MemoryStream())
{
Yeti.MMedia.Mp3.Mp3Writer mp3 = new Yeti.MMedia.Mp3.Mp3Writer(output, fmt, conf);
int readCount;
while ((readCount = rdr.Read(buffer, 0, blen)) > 0)
mp3.Write(buffer, 0, readCount);
mp3.Close();
return output.ToArray();
}
}
}
Either add a reference to System.Windows.Forms to your project (if it's not there already), or edit AudioWriter.cs and WriterConfig.cs to remove the references. Both of these have a using System.Windows.Forms; that you can remove, and WriterConfig.cs has a ConfigControl declaration that needs to be removed/commented out.
Once all of that is done you should have a functional in-memory wave-file to MP3 converter that you can use to convert the WAV file that you are getting from the SpeechSynthesizer into an MP3.
This is a bit old now, but since you haven't accepted the answer I previously provided...
I have recently built an extension for NAudio that encapsulates the LAME library to provide simplified MP3 encoding.
Use the NuGet package manager to find NAudio.Lame. Basic example for using it available here.
Assuming you're trying to convert the output into MP3, you need something that can handle transcoding the audio. There are a number of tools available, but my personal preference is FFmpeg. It's a command line tool so you will need to take that into account, but otherwise it's very easy to use.
There's lots of information online, but you can start by checking out their documentation here.
I had a similar requirement in .net4.0 to convert 8bit 8Khz mono wav and used the following code
public void WavToMp3(string wavPath, string fileId)
{
var tempMp3Path = TempPath + "tempFiles\\" + fileId + ".mp3";
var mp3strm = new FileStream(tempMp3Path, FileMode.Create);
try
{
using (var reader = new WaveFileReader(wavPath))
{
var blen = 65536;
var buffer = new byte[blen];
int rc;
var bit16WaveFormat = new WaveFormat(16000, 16, 1);
using (var conversionStream = new WaveFormatConversionStream(bit16WaveFormat, reader))
{
var targetMp3Format = new WaveLib.WaveFormat(16000, 16, 1);
using (var mp3Wri = new Mp3Writer(mp3strm, new Mp3WriterConfig(targetMp3Format, new BE_CONFIG(targetMp3Format,64))))
{
while ((rc = conversionStream.Read(buffer, 0, blen)) > 0) mp3Wri.Write(buffer, 0, rc);
mp3strm.Flush();
conversionStream.Close();
}
}
reader.Close();
}
File.Move(tempMp3Path, TempPath + fileId + ".mp3");
}
finally
{
mp3strm.Close();
}
}
Prerequists:
.net 4 compiled yeti library (to obtain it download this older one (http://www.codeproject.com/KB/audio-video/MP3Compressor/MP3Compressor.zip) and convert it to .net4.0 then build the solution to obtain the new version dlls)
download the NAudio libraries (as Lame support 16bit wav sample only i had to first convert it from 8bit to 16bit wav)
I have used a buffer size of 64kpbs (my custom requirement)
have a try:
using (WaveStream waveStream = WaveFormatConversionStream.CreatePcmStream(new
Mp3FileReader(inputStream)))
using (WaveFileWriter waveFileWriter = new WaveFileWriter(outputStream, waveStream.WaveFormat))
{
byte[] bytes = new byte[waveStream.Length];
waveStream.Position = 0;
waveStream.Read(bytes, 0, waveStream.Length);
waveFileWriter.WriteData(bytes, 0, bytes.Length);
waveFileWriter.Flush();
}

Categories

Resources