Log in

View Full Version : DTS -> AC3 (and how to obtain a DTS)


Erupter
23rd June 2005, 22:49
Programs needed:
SmartRipper (2.41 used)
Hypercube Transcoder (3.05 used)
Besweet (1.5b29 used along with besweetgui 0.6b61)

Finally after days of wandering from one forum to another i found the way.
First: how to obtain a DTS from a dvd disc.
No need to decrypt to harddisk the dvd, just fire up SmartRipper.
Select the chain you want to rip the audio.
You can select a single cell: for example to obtain a single song from a dvd audio.
Now go to "stream processing" (second tab), enable it and select the audio track you want to rip.

Notice: smartripper, although it should, is not able to "batch process" a track.
I mean: even if you select in the options "every cell-id" or "every chapter" or whatever for file-splitting, should you select an entire chain do not expect SmartiRipper to automatically cut the audio stream in each single track.
You must do this manually, by manually selecting one at a time the cell-ids and start the ripping.

Aside of the audio track selection mark, mark also the "demux to extra file".
Select a directory and save.
SmartRipper will create many files along which you will find your DTS (or wav, or whatever).

Now it's the turn of Hypercube Transcoder.
It's the only program i have found to be able to decode a DTS.
Watch for it needs you to define the correct channel sequence.
By default it starts L-R-C-LFE-SL-SR.
For the dvd audio i had to rip, the correct sequence was L-C-R-SL-SR-LFE
So for example you can try to decode to single wavs (mono) and try to find the correct channel sequence.
Then once you have sorted this out, leave "directshow processing" alone as well as the frequency setting, select "Wav engine (single 6ch file)" in the output dropdown and lunch the process.

Once it has finished it's up to besweet.
Fire up besweet gui, select AC3 as output.
Deselect "compress dynamic range" and "downconvert sample rate".
Select your previously created wav, give the output a name and a folder, and start the conversion process.

Should you want to you can use this profile of mine

AC3@640=-ac3enc( -6ch -b 640))

to be added to the DDprofiles.ini file under the [Besweet] tag.
This will encode your file at a quality of 640.
Should this be too high for you, you just have to modify the number according to your preferences.


Hope this is of some help to anybody whishing to listen to their DTS tracks on the pc.
I find only one little thing i haven't been able to solve: the DTS sounds much louder then the AC3.
But i used no normalization or anything.
Any suggestion on why?
That means i would like the AC3 to sound as loud as the DTS, and not the contrary ;)

Erupter
30th June 2005, 14:39
I found and tried HeadAC3.
It produces much better AC3 then besweet.
And the global gain parameter works (unlike besweet).
I use a 10dB global gain, no DRC anywhere and 640 kbps.

scarf
16th July 2005, 17:14
I tried using this Hypercube Transcoder following your instructions, but the WAV file it creates just sounds horrible! very staticy. i also tried making 6 seperate WAV file and they are all staticy and horrible sounding. what could be the problem? i have program called mediainfo, http://mediainfo.sourceforge.net, that can show info about DTS file, including channel order, and it tells me my DTS has "L, R, C, SL, SR, Subwoofer". does this mean "front left, front right, center, rear left, rear right, lfe"? if i have this order wrong would it make the output WAV sound so horrible and staticy? please advise! thank you.

Erupter
17th July 2005, 23:35
and it tells me my DTS has "L, R, C, SL, SR, Subwoofer". does this mean "front left, front right, center, rear left, rear right, lfe"? if i have this order wrong would it make the output WAV sound so horrible and staticy? please advise! thank you.

1) yes that should be the order but it may need to be changed.
2) different channel order is not able to alter the sound.
Maybe you are trying to downsample from 48000 to 41000?
Also did you use "directshow processing"?
I always use DS as "native processing" doesn't work giving out only noise.

scarf
18th July 2005, 01:16
not trying to downsample, and "DirectShow Processing" is selected! if i use "native Processing" the WAV file is only comes out to be 1 or 2 minutes (while the DTS is ~50 minutes) and is completely silent! i see "Audio Unknown" in the "DirectShow Filter Graph" section, maybe this is an issue? this is the sequence shown:

DTSSource -> Input: None -> Output: Audio Unknown WaveFormatEx 6 ch 48000 hz 0 bits -> ffdshow Audio Decoder -> Input: Audio Unknown WaveFormatEx 6 ch 48000 hz 0 bits -> Output: Audio Unknown WaveFormatEx 6 ch 48000 hz 32 bits -> Hypercube Wrapper -> Input: Audio Unknown WaveFormatEx 6 ch 48000 hz 32 bits -> Output: None

scarf
21st July 2005, 07:24
i found this thread: http://www.videohelp.com/forum/viewtopic.php?t=254532. which instructs to use "foobar2000" (special version) and it makes a nice-sounding WAV which i can give to HeadAC3e