How do I write RIFF chunk data to a WAV file that was made with NAudio?

2
votes

How do I write RIFF chunk data to the end of a WAV file? Yes, that's right, to the end of a WAV file. Why? Because that's the way the old program I'm replacing did it and the integrating programs are prepared for it that way and they cannot be modified.

2
c#.netaudiowavnaudio
possible duplicate of NTFS Alternate Data Streams - .NET – p.s.w.g
will this help in regards to at least reading from it msdn.microsoft.com/en-us/library/… – MethodMan
@DJKRAZE, in hind sight, I think that reads the comments from the AssemblyInfo class. Though it does state that it's for any file on disk. But I can't write with it. Man, this is crazy! – Mike Perrenoud
You could try this >link< - as I remember, that gets closest and the 'least' involved when dealing with this - see if the requirements match. – NSGaga-mostly-inactive
And since you mentioned .wav files - this involves something similar for mp3-s - codeproject.com/Articles/5036/… – NSGaga-mostly-inactive

2 Answers

1
votes

Personally I would use BASS.NET, it does handle the RIFF ICMT tag.

The EncoderWAV class provides an WAV_AddRiffInfo property that permits it.

// re-encodes a file and adds a comment to it.    
var init = Bass.BASS_Init(-1, 44100, BASSInit.BASS_DEVICE_DEFAULT, IntPtr.Zero);
if (init)
{
    int channel = Bass.BASS_StreamCreateFile("filename", 0, 0, BASSFlag.BASS_STREAM_DECODE);
    var tagInfo = new TAG_INFO { comment = "your comment" };
    var encoderWav = new EncoderWAV(channel);
    encoderWav.WAV_AddRiffInfo = true;
    encoderWav.TAGs = tagInfo;
    encoderWav.OutputFile = "output.wav";
    var start = encoderWav.Start(null, IntPtr.Zero, false);
    Bass.Free();
}

Note : do not forget to put bass.dll where your .EXE is

(at first BASS is a C library, BASS.NET provides a wrapper for it but also many cool stuff not present in the original library such as encoders, visualisers etc ...)

IMO BASS.NET is the best audio library for .NET, especially low-level stuff, actually that's the only one :-)

0
votes

Ok, so in the end it wasn't really a standard format. We found that out by leveraging the CueWaveFileWriter like Mark Heath stated. The format was actually something very custom and looked more like this:

LIST - this is a one-time blurb of text.
Tag Name (e.g. ICOP for copyright)
4 bytes indicating the length of the following value + 1. The 1 was a (byte)0 terminator.
Value

so an example might look like this:

ICOP<bh:15><bh:00><bh:00><bh:00>4/5/2013 10:48:40 AM<bh:00>

so, to get this done, we just created a BinaryWriter and wrote them out directly like this:

using (BinaryWriter bw = new BinaryWriter(writer))
{
    foreach (var item in RIFFChunks)
    {
        if (!(bw.BaseStream.Position % 2 == 0))
        {
            bw.Write((byte)0);
        }

        bw.Write(Encoding.Default.GetBytes(item.Key));
        bw.Write(item.Value.Length + 1);
        bw.Write(Encoding.Default.GetBytes(item.Value));
        bw.Write((byte)0);
    }
}

and you may want to take note to this line if (!(bw.BaseStream.Position % 2 == 0)), that needed to exist because each tag name needed to start on an even offset. I honestly don't know why, I just know it did.

I have to say though, there's a good chance I can now reverse engineer just about any file format. But really, what good is that?!?