Log in

View Full Version : tranzcode DTS-HD to WAV results in 3iVX?


HD-DVD±RW
26th September 2007, 16:10
I'm hoping someone can help me with a problem here. I'm using tranzcode v0.40 (beta) (http://tranzcode.byethost22.com/) to convert a DTS-HD file to 6 mono wave files. That process works but the end wav file shows this in media info:


General #0
Complete name : C:\MOVIE-C.wav
Format : Wave
Format/Family : RIFF
File size : 1.22 GiB
PlayTime : 1h 53mn
Bit rate : 1536 Kbps

Audio #0
Codec : 3ivX
Codec/Family : MPEG-4
Codec/Info : IEEE FLOAT
Bit rate : 1536 Kbps
Channel(s) : 1 channel
Sampling rate : 48 KHz
Resolution : 32 bits


When comparing against another feature while converting DD+ to mono wav mediainfo shows this:

General #0
Complete name : C:\Movie.C.wav
Format : Wave
Format/Family : RIFF
File size : 1.05 GiB
PlayTime : 2h 10mn
Bit rate : 1152 Kbps

Audio #0
Codec : PCM
Codec/Family : PCM
Codec/Info : Microsoft PCM
Bit rate : 1152 Kbps
Channel(s) : 1 channel
Sampling rate : 48 KHz
Resolution : 24 bits


How can I use tranzcode to convert DTS-HD audio to PCM? I've reviewed the tranzcode readme but nothing jumps out at me. TIA.

tebasuna51
26th September 2007, 16:44
I don't know what is:
Codec : 3ivX
Codec/Family : MPEG-4
for Media Info, but the difference between your two wave samples are the bitdepth:

- first is 32 bit float
- second is 24 bit int

both are valid wav files and can be selected (the more appropriate format for the next step conversion) with the Tranzcode parameters:

/16 outputs 16 bit int
/24 to 24 bit int
/32f to 32 bit float
/64f to 64 bit float

HD-DVD±RW
26th September 2007, 23:37
@tebasuna51 thanks for the tip on the bit rate. What is the difference between floating and integer bit rates? and does a higher number really make a difference (sorry I'm a noob)? Thanks.

I was able to get the PCM output by first installing the 3ivX package (http://www.3ivx.com/download/) and then removing it. I must of had a conflicting directshow filter or something.

The new output using the options of /24 /48000 outputs this:


General #0
Complete name : movie(24-bit,48.0kHz)-C.wav
Format : Wave
Format/Family : RIFF
File size : 936 MiB
PlayTime : 1h 53mn
Bit rate : 1152 Kbps

Audio #0
Codec : PCM
Codec/Family : PCM
Codec/Info : Microsoft PCM
Bit rate : 1152 Kbps
Channel(s) : 1 channel
Sampling rate : 48 KHz
Resolution : 24 bits


I hope this helps someone who had the same issue as me.

tebasuna51
27th September 2007, 02:00
@tebasuna51 thanks for the tip on the bit rate. What is the difference between floating and integer bit rates? and does a higher number really make a difference (sorry I'm a noob)?

Not bitrate (number of bits per second) but bitdepth (number of bits per sample).

More bitdepth means more precission, more dynamic range, less digital noise.

The difference between integer and float is how is stored the value of a sample: using integers or real (float) numbers (exponent and mantissa). Float is preferred when we need operate (timestretch,...) or the encoder use this format like Aften for instance. Integer is preferred to lossless codecs like flac, ...