mulaw to wav headers linaer 16 working

.version

@@ -1 +1 @@
-v0.1.5
+v0.2.1

audio/audio.go

@@ -5,9 +5,10 @@ type AudioInfo struct {
 	Channels    int
 	BitDepth    int
 	FloatFormat bool
+	Verbose     bool
 }
 
 type AudioFormat interface {
-	Decode(data []byte, info AudioInfo) []float64
-	Encode(audioData []float64, info AudioInfo) []byte
+	Decode(data []byte, info AudioInfo) []byte
+	Encode(audioData []byte, info AudioInfo) []byte
 }

audio/formats/pcm/pcm.go

@@ -9,13 +9,12 @@ import (
 
 type PCMFormat struct{}
 
-func (f *PCMFormat) Decode(data []byte, info audio.AudioInfo) []float64 {
+func (f *PCMFormat) Decode(data []byte, info audio.AudioInfo) []byte {
 	log.Printf("Not implemented")
 	return nil
 }
 
-func (f *PCMFormat) Encode(audioData []float64, info audio.AudioInfo) []byte {
-
+func (f *PCMFormat) Encode(audioData []byte, info audio.AudioInfo) []byte {
 	// convert float64 to byte
 	data := make([]byte, len(audioData))
 	for i := 0; i < len(audioData); i++ {

audio/formats/ulaw/mulaw.go

@@ -9,20 +9,26 @@ import (
 
 type MuLawFormat struct{}
 
-func (f *MuLawFormat) Decode(data []byte, info audio.AudioInfo) []float64 {
-
-	pcmData := make([]float64, len(data))
+func (f *MuLawFormat) Decode(data []byte, info audio.AudioInfo) []byte {
+	pcmData := make([]byte, len(data)*2)
+	// remember that each byte need to be converted to two bytes
 	for i := 0; i < len(data); i++ {
-		pcmData[i] = float64(utils.DecodeFromULaw(data[i]))
-	}
 
-	log.Println("[Bytes][Org]", data[0:100])
-	log.Println("[Bytes][Pcm]", pcmData[0:100])
+		newFrame := utils.DecodeULawToPCM(data[i : i+1])
+		pcmData[i*2] = newFrame[0]
+		pcmData[i*2+1] = newFrame[1]
+
+	}
+	if info.Verbose {
+		log.Println("[MuLaw][Decode] Decoded to PCM")
+		log.Println("[MuLaw][Decode] PCM Data [0  :100]", pcmData[0:100])
+		log.Println("[MuLaw][Decode] PCM Data [100:200]", pcmData[100:200])
+	}
 
 	return pcmData
 }
 
-func (f *MuLawFormat) Encode(audioData []float64, info audio.AudioInfo) []byte {
+func (f *MuLawFormat) Encode(audioData []byte, info audio.AudioInfo) []byte {
 	log.Printf("Not implemented")
 	return nil
 }

audio/utils/endianness.go

@@ -0,0 +1,19 @@
+package utils
+
+import "encoding/binary"
+
+type ENDIANNESS int
+
+const (
+	LITTLE_ENDIAN ENDIANNESS = iota
+	BIG_ENDIAN
+)
+
+// GetEndianess returns the endianness of the CPU safety
+func GetEndianess() ENDIANNESS {
+	if binary.LittleEndian.Uint16([]byte{0x01, 0x00}) == 1 {
+		return LITTLE_ENDIAN
+	} else {
+		return BIG_ENDIAN
+	}
+}

audio/utils/endianness_test.go

@@ -0,0 +1,10 @@
+package utils
+
+import "testing"
+
+func TestGetEndianess(t *testing.T) {
+	e := GetEndianess()
+	if e != LITTLE_ENDIAN && e != BIG_ENDIAN {
+		t.Errorf("unexpected endianess: %d", e)
+	}
+}

audio/utils/tables.go

@@ -1,7 +1,7 @@
 package utils
 
-// ulawDecode is a lookup table for u-law to LPCM
-var ulawDecode = [256]int16{
+// mulawToPcmTable is a lookup table for converting u-law to 16bit LPCM
+var mulawToPcmTable = [256]int16{
 	-32124, -31100, -30076, -29052, -28028, -27004, -25980, -24956,
 	-23932, -22908, -21884, -20860, -19836, -18812, -17788, -16764,
 	-15996, -15484, -14972, -14460, -13948, -13436, -12924, -12412,

audio/utils/ulaw_utils.go

@@ -1,6 +1,15 @@
 package utils
 
-func DecodeFromULaw(uLaw byte) int16 {
-	pcm := ulawDecode[uLaw]
-	return int16(pcm)
+// DecodeULawToPCM decodes raw ULaw-encoded audio data to linear PCM.
+// Each ULaw frame is 8-bit logarithmic PCM, which is converted to 16-bit linear PCM.
+func DecodeULawToPCM(ulaw []byte) []byte {
+	if len(ulaw) == 0 {
+		return nil
+	}
+	pcm := make([]byte, len(ulaw)*2)
+	for i, ulawFrame := range ulaw {
+		pcmFrame := mulawToPcmTable[ulawFrame]
+		copy(pcm[i*2:], []byte{byte(pcmFrame), byte(pcmFrame >> 8)})
+	}
+	return pcm
 }

audio/utils/wav_headers.go

@@ -3,6 +3,8 @@ package utils
 import (
 	"encoding/json"
 	"fmt"
+	"log"
+	"strings"
 	"sync"
 )
 
@@ -36,28 +38,28 @@ type WaveFormat uint16
 
 // WAV FORMAT CODES
 const (
-	WAVE_FORMAT_UNKNOWN            = 0x0000 // Microsoft Unknown Wave Format
-	WAVE_FORMAT_PCM                = 0x0001 // Microsoft PCM Format
-	WAVE_FORMAT_ADPCM              = 0x0002 // Microsoft ADPCM Format
-	WAVE_FORMAT_IEEE_FLOAT         = 0x0003 // IEEE float
-	WAVE_FORMAT_VSELP              = 0x0004 // Compaq Computer's VSELP
-	WAVE_FORMAT_IBM_CVSD           = 0x0005 // IBM CVSD
-	WAVE_FORMAT_ALAW               = 0x0006 // ALAW
-	WAVE_FORMAT_MULAW              = 0x0007 // MULAW
-	WAVE_FORMAT_DTS                = 0x0008 // Digital Theater Systems DTS
-	WAVE_FORMAT_DRM                = 0x0009 // Microsoft Corporation
-	WAVE_FORMAT_WMAVOICE9          = 0x000A // Microsoft Corporation
-	WAVE_FORMAT_WMAVOICE10         = 0x000B // Microsoft Corporation
-	WAVE_FORMAT_OKI_ADPCM          = 0x0010 // OKI ADPCM
-	WAVE_FORMAT_DVI_ADPCM          = 0x0011 // Intel's DVI ADPCM
-	WAVE_FORMAT_IMA_ADPCM          = WAVE_FORMAT_DVI_ADPCM
-	WAVE_FORMAT_MEDIASPACE_ADPCM   = 0x0012 // Videologic's MediaSpace ADPCM
-	WAVE_FORMAT_SIERRA_ADPCM       = 0x0013 // Sierra ADPCM
-	WAVE_FORMAT_G723_ADPCM         = 0x0014 // G.723 ADPCM
-	WAVE_FORMAT_DIGISTD            = 0x0015 // DSP Solution's DIGISTD
-	WAVE_FORMAT_DIGIFIX            = 0x0016 // DSP Solution's DIGIFIX
-	WAVE_FORMAT_DIALOGIC_OKI_ADPCM = 0x0017 // Dialogic Corporation
-	WAVE_FORMAT_MEDIAVISION_ADPCM  = 0x0018 // Media Vision ADPCM
+	WAVE_FORMAT_UNKNOWN            = 0x0000                // Microsoft Unknown Wave Format
+	WAVE_FORMAT_PCM                = 0x0001                // Microsoft PCM Format
+	WAVE_FORMAT_ADPCM              = 0x0002                // Microsoft ADPCM Format
+	WAVE_FORMAT_IEEE_FLOAT         = 0x0003                // IEEE float
+	WAVE_FORMAT_VSELP              = 0x0004                // Compaq Computer's VSELP
+	WAVE_FORMAT_IBM_CVSD           = 0x0005                // IBM CVSD
+	WAVE_FORMAT_ALAW               = 0x0006                // ALAW
+	WAVE_FORMAT_MULAW              = 0x0007                // MULAW
+	WAVE_FORMAT_DTS                = 0x0008                // Digital Theater Systems DTS
+	WAVE_FORMAT_DRM                = 0x0009                // Microsoft Corporation
+	WAVE_FORMAT_WMAVOICE9          = 0x000A                // Microsoft Corporation
+	WAVE_FORMAT_WMAVOICE10         = 0x000B                // Microsoft Corporation
+	WAVE_FORMAT_OKI_ADPCM          = 0x0010                // OKI ADPCM
+	WAVE_FORMAT_DVI_ADPCM          = 0x0011                // Intel's DVI ADPCM
+	WAVE_FORMAT_IMA_ADPCM          = WAVE_FORMAT_DVI_ADPCM // Intel's DVI ADPCM
+	WAVE_FORMAT_MEDIASPACE_ADPCM   = 0x0012                // Videologic's MediaSpace ADPCM
+	WAVE_FORMAT_SIERRA_ADPCM       = 0x0013                // Sierra ADPCM
+	WAVE_FORMAT_G723_ADPCM         = 0x0014                // G.723 ADPCM
+	WAVE_FORMAT_DIGISTD            = 0x0015                // DSP Solution's DIGISTD
+	WAVE_FORMAT_DIGIFIX            = 0x0016                // DSP Solution's DIGIFIX
+	WAVE_FORMAT_DIALOGIC_OKI_ADPCM = 0x0017                // Dialogic Corporation
+	WAVE_FORMAT_MEDIAVISION_ADPCM  = 0x0018                // Media Vision ADPCM
 )
 
 // GenerateWavHeadersWithConfig generates the headers for a WAV file without the size fields
@@ -69,6 +71,7 @@ func GenerateWavHeadersWithConfig(head *WavHeader) []byte {
 
 	totalLength := head.Length - 8
 	if totalLength < 0 {
+		log.Printf("WARNING: Length is less than 8, defaulting to 0\n")
 		totalLength = 0
 	}
 
@@ -134,3 +137,34 @@ func GenerateWavHeadersWithConfig(head *WavHeader) []byte {
 		IntToBytes(subChunk2Size),
 	)
 }
+
+// PrintWavHeaders prints the headers of a WAV file
+// first 44 bytes of a WAV file
+func PrintWavHeaders(headersWav []byte) {
+	if len(headersWav) != 44 {
+		log.Println("[ERROR] Headers are not 44 bytes long")
+		return
+	}
+	fmt.Println("Headers (WAV):")
+	comments := []string{
+		"(4) Chunk ID [RIFF]",
+		"(4) Chunk size",
+		"(4) Format [WAVE]",
+		"(4) Sub-chunk 1 ID [fmt ]",
+		"(4) Sub-chunk 1 size",
+		"(2) Audio format (PCM) & (2) Number of channels",
+		"(4) Sample rate",
+		"(4) Byte rate",
+		"(2) Block align & (2) Bits per sample",
+		"(4) Sub-chunk 2 ID [data]",
+		"(4) Sub-chunk 2 size",
+	}
+	for i := 0; i < 44; i += 4 {
+		fmt.Println(
+			fmt.Sprintf("[%2d,%2d]", i, i+4),
+			fmt.Sprintf("% 2x", headersWav[i:i+4]),
+			"\t<"+strings.ToUpper(string(headersWav[i:i+4]))+">\t",
+			comments[i/4],
+		)
+	}
+}

examples/transcoder/mulaw_2_wav.go

@@ -2,7 +2,6 @@ package main
 
 import (
 	"io"
-	"log"
 	"os"
 
 	"github.com/sopro-dev/sopro-core/audio"
@@ -30,6 +29,7 @@ func main() {
 		Channels:    1,
 		BitDepth:    8,
 		FloatFormat: false,
+		Verbose:     false,
 	}
 
 	transcoder := audio.NewTranscoder(&mulaw.MuLawFormat{}, &pcm.PCMFormat{})
@@ -52,21 +52,10 @@ func main() {
 		Channels:   1,
 		SampleRate: 8000,
 		BitDepth:   16,
-		Verbose:    true,
+		Verbose:    audioInfo.Verbose,
 	})
 
 	f.Write(headers)
 	f.Seek(44, 0)
 	f.Write(outputData)
-
-	f.Seek(0, 0)
-	// print first 100 bytes of the output file
-	out, err := io.ReadAll(f)
-	if err != nil {
-		panic(err)
-	}
-
-	// get data from :100
-	log.Println("[Bytes][out]", out[0:100])
-
 }