DTS to multi-channel FLAC. Advice needed.

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I wish there were a straightforward DTS file container that could be used for DTS at its native file size. Would be even better if it were possible to decode DTS-CD directly to such a file format. Alas...

Can't understand what do you mean. Could you explain it better?
 
Can't understand what do you mean. Could you explain it better?
Uncompressed PCM audio can be stored in AIFF and WAV file containers.

FLAC losslessly compressed audio is stored in FLAC files.

MP3 lossily compressed audio is stored in MP3 files.

DTS lossily compressed audio has no widely supported equivalent file type. Decompressing to PCM and then recompressing to FLAC means using more file space than necessary. The same goes for storing DTS-CD as stereo FLAC.
 
Matroska Audio (MKA) can hold a DTS stream. DTS-CD to DTS:X, same with Dolby, everything including Atmos.

EDIT: Playback via HDMI and your AVR will decode the contained stream: DTS or Atmos etc
 
I always recommend to rip and convert a dts cd as a entire single file, then when it's converted to flac use the cuesheeet to split into tracks...

Fascinating and thank you!

Crazy format. I have tested out ripping as one file and it ends up in the same strange situation with what appears to be minor clipping. Yet the process seems good so perhaps it is not clipping after all and there was something weird with the earlier DTS rips. Or an issue with the foobar dts plugin (v0.6.8 - the latest) AND the one DVD Audio Extractor uses, which seems doubtful. I'll include the foobar Dynamic Range Meter results for context.

Edit: I suspect I MIGHT have found the issue. A ReplayGain analysis if the one large dts wav file states the peak at 1.021664. My guess is that everything above 1 is being chopped in the conversion to FLAC.

Ripping as individual tracks (EAC -> DVD Audio Extractor to convert to FLAC):

DR Peak RMS Duration Track
--------------------------------------------------------------------------------
DR12 0.00 dB -17.19 dB 4:08 01
DR11 -1.30 dB -17.44 dB 2:57 02
DR11 0.00 dB -15.23 dB 5:11 03
DR12 -0.04 dB -17.06 dB 4:08 04
DR12 0.00 dB -17.26 dB 4:21 05
DR13 0.00 dB -18.12 dB 3:12 06
DR12 0.00 dB -16.54 dB 4:25 07
DR12 -1.66 dB -18.62 dB 4:00 08
DR10 -1.75 dB -19.47 dB 3:25 09
DR12 -2.28 dB -20.06 dB 3:51 10
--------------------------------------------------------------------------------

Number of tracks: 10
Official DR value: DR12

Samplerate: 44100 Hz
Channels: 6
Bits per sample: 24
Bitrate: 3738 kbps
Codec: FLAC

Ripping as one track (EAC -> Foobar2000 to convert to FLAC -> CueSheet in Foobar to make individual files):

DR Peak RMS Duration Track
--------------------------------------------------------------------------------
DR12 0.00 dB -17.19 dB 4:09 01
DR11 -1.30 dB -17.44 dB 2:57 02
DR11 0.00 dB -15.23 dB 5:11 03
DR12 -0.04 dB -17.07 dB 4:08 04
DR12 0.00 dB -17.26 dB 4:21 05
DR13 0.00 dB -18.12 dB 3:12 06
DR12 0.00 dB -16.54 dB 4:25 07
DR12 -1.66 dB -18.62 dB 4:00 08
DR10 -1.75 dB -19.47 dB 3:25 09
DR12 -2.28 dB -20.06 dB 3:50 10
--------------------------------------------------------------------------------

Number of tracks: 10
Official DR value: DR12

Samplerate: 44100 Hz
Channels: 6
Bits per sample: 24
Bitrate: 3744 kbps
Codec: FLAC
 
Last edited:
This may not be the best solution; but it looks like the DTS WAV having a peak above 1 was the issue. This forces more editting of the original files than I would like, but appears to solve the clipping.

1. Open Exact Audio Copy.
2. Action -> Copy Image and Create CUE Sheet. Gives one big dts wav file.
3. Change resulting file extension from wav to dts.
4. Open dts in foobar (which has dts plugin).
5. Right Click -> Replay Gain -> Scan as a single album; save the result to the file.
6. Right click -> Convert -> FLAC (24 bit). Gives one big flac file.
Note: Processing -> ReplayGain -> 'Album';
'prevent clipping according to peak'
7. Edit cue sheet from Exact Audio Copy to refer to flac file
From *.wav WAVE to *.flac FLAC
8. Open cue sheet in Foobar.
9. Right click -> Convert -> FLAC (24 bit). Gives individual tracks.
Note: Processing -> ReplayGain -> 'None'

DR Peak RMS Duration Track
--------------------------------------------------------------------------------
DR12 -0.18 dB -17.38 dB 4:09 01
DR11 -1.49 dB -17.62 dB 2:57 02
DR11 -0.11 dB -15.42 dB 5:11 03
DR12 -0.23 dB -17.25 dB 4:08 04
DR12 0.00 dB -17.45 dB 4:21 05
DR13 -0.16 dB -18.31 dB 3:12 06
DR12 -0.15 dB -16.73 dB 4:25 07
DR12 -1.85 dB -18.81 dB 4:00 08
DR10 -1.94 dB -19.65 dB 3:25 09
DR12 -2.47 dB -20.24 dB 3:50 10
--------------------------------------------------------------------------------

Number of tracks: 10
Official DR value: DR12

Samplerate: 44100 Hz
Channels: 6
Bits per sample: 24
Bitrate: 3737 kbps
Codec: FLAC
 
DTS lossily compressed audio has no widely supported equivalent file type. Decompressing to PCM and then recompressing to FLAC means using more file space than necessary. The same goes for storing DTS-CD as stereo FLAC.

Now i got it... because DTS wasn't intended as a consumer file format but as a inserted element in Video ones, VOB AVI or whatever.

The same goes for storing DTS-CD as stereo FLAC.

The storing in the DTS-CD is in WAV format 16/2/44100, Flac compression can't do anything useful to a dts-cd file in order to save space, but it would not be bigger than the wav one.
 
The Nick Cave & the Bad Seeds 5.1 mixes on DVD are 48kHz.

This ... has ended up also being strange.

For each album, I have used DVD Audio Extractor - set it to extract the DTS 5.1, which is always labelled (but is not necessarily) 48kHz 6ch, then on the next form tell it to use the sample rate "same as input". A fascinating discussion on how DVDAE behaves in this situation is in this thread -> https://www.quadraphonicquad.com/fo...vd-audio-extractor-question-or-2.20591/page-2

Anyway, I have three to test:
The Good Son = 96kHz
Murder Ballads = 96kHz
Dig Lazarus Dig = 48kHz

(!)
 
This ... has ended up also being strange.

For each album, I have used DVD Audio Extractor - set it to extract the DTS 5.1, which is always labelled (but is not necessarily) 48kHz 6ch, then on the next form tell it to use the sample rate "same as input". A fascinating discussion on how DVDAE behaves in this situation is in this thread -> https://www.quadraphonicquad.com/fo...vd-audio-extractor-question-or-2.20591/page-2

Anyway, I have three to test:
The Good Son = 96kHz
Murder Ballads = 96kHz
Dig Lazarus Dig = 48kHz

(!)

Oh, weird, I could have sworn they weren't 24/96 DTS!
 
Audacity's 'DTS/AC3 to S/PDIF' converter puts DD and DTS lossy bitstreams inside a wrapper that can be converted to FLAC and then tagged. File size barely changes (because the audio data remains undecoded lossy compressed). All my AVRs so far have been able to decode such files. Pipeline is WASAPI out from foobar2k-->hdmi or coax in on AVR)
 
I wish there were a straightforward DTS file container that could be used for DTS at its native file size. Would be even better if it were possible to decode DTS-CD directly to such a file format. Alas...
DTS on CD encoded to FLAC, decodes back to perfect FLAC with all 4.0 or 5.1 discreet channels playing like they should. It encodes to smaller size, and then springs back to full normal NATIVE size with no loss of anything. That is as easy as it ever got.

So where is the problem here? Oh the only thing needed with these DTS CDs is dBpoweram music ripper / converter
 
DTS on CD encoded to FLAC, decodes back to perfect FLAC with all 4.0 or 5.1 discreet channels playing like they should. It encodes to smaller size, and then springs back to full normal NATIVE size with no loss of anything. That is as easy as it ever got.

So where is the problem here? Oh the only thing needed with these DTS CDs is dBpoweram music ripper / converter
The problem is that it's still two steps removed from the original multichannel DTS stream.
 
The problem is that it's still two steps removed from the original multichannel DTS stream.
In lossless encoding it’s not removed from the original even one step - that’s why they call it lossless.

FLAC was or is a wonderful gift to the music enthusiast.
 
In lossless encoding it’s not removed from the original even one step - that’s why they call it lossless.
But the original DTS stream is four to six discrete, compressed channels, which is not two-channel uncompressed PCM. It has to be shoehorned into this format to be put on what is, ostensibly, an audio CD. If one were to rip a DTS-CD and then losslessly compress it, while the original information is left intact, it is still two formats removed from the original DTS stream. I'd prefer a way to get back to that stream and store it in a DTS-specifically multichannel audio file.
 
But the original DTS stream is four to six discrete, compressed channels, which is not two-channel uncompressed PCM. It has to be shoehorned into this format to be put on what is, ostensibly, an audio CD. If one were to rip a DTS-CD and then losslessly compress it, while the original information is left intact, it is still two formats removed from the original DTS stream. I'd prefer a way to get back to that stream and store it in a DTS-specifically multichannel audio file.
The fact that you have four to six channels is irrelevant to the issue correct?

DTS is not shoehorned because it’s going to a CD. DTS and it’s varieties are compressed because they were designed to be oif smaller file sizes, and there’re was the trade-off.

FLAC encoding offers you a way to “get back” to that original DTS stream exactly and perfectly.

If you rip those files and save as wav files you’ll have DTS that can be played by any player that recognizes a DTS encoded wav. That is what is in the DTS CDs, I don’t see the benefit here, but if you sleep better knowing it is in its native format then rip as wav and save that. Tagging is much more limited this way - another drawback there.

But most people have moved onto encoding DTS into FLAC because it works so well and is a smaller file size, and is recognized as a DTS file when decoded. Real honest and true DTS nothing but.
 
...it is still two formats removed from the original DTS stream. I'd prefer a way to get back to that stream...

Get back to the original 6-channel uncompressed audio, before the encoding to DTS?
 
DTS is not shoehorned because it’s going to a CD. DTS and it’s varieties are compressed because they were designed to be oif smaller file sizes, and there’re was the trade-off.
You're not getting it.

DTS—a lossily compressed multichannel format—is shoehorned into uncompressed stereo PCM (which is not WAV, by the way) so that it can be written to an audio CD. Contrast this with native DTS as written to DVD-Video muxed into MPEG2 video, with discrete channels. When played back through a DTS decoder, the same stream as written to either format will sound the same, but the data are written and stored quite differently.

It is possible to demux DTS from MPEG2, but the resulting file isn't a consumer standard. Furthermore, there is currently no way to convert DTS-CD to native DTS. So, I propose that: 1) a DTS filetype standard should exist, and 2) there should be a method to arrive at this filetype regardless of the storage medium from which the stream originated.
 
You're not getting it.

DTS—a lossily compressed multichannel format—is shoehorned into uncompressed stereo PCM (which is not WAV, by the way) so that it can be written to an audio CD. Contrast this with native DTS as written to DVD-Video muxed into MPEG2 video, with discrete channels. When played back through a DTS decoder, the same stream as written to either format will sound the same, but the data are written and stored quite differently.

It is possible to demux DTS from MPEG2, but the resulting file isn't a consumer standard. Furthermore, there is currently no way to convert DTS-CD to native DTS. So, I propose that: 1) a DTS filetype standard should exist, and 2) there should be a method to arrive at this filetype regardless of the storage medium from which the stream originated.
I completely get it. I know DTS has had various types of flavors depending on what year or era you refer to. They had to get it on a CD disc using that disc playback format and wrote it like it’s a wav file. It’s not a 16/44 wav, but it’s written like one with encoded channels in that static noise.

But I simply send these (ripped) files to a player that is known to play DTS files and bingo they play decoded to surround.

I see it like Dolby Digital which also has evolved along the way. Dolby noise reduction has also evolved and has various types. Dolby SR for recording studio professionals was one hot system in the analogue world, just before digital recording took off, Dolby finally got it right. “Native” Dolby is not a term I ever heard used in the studios. Dolby A, B, C, and SR is what I heard about.

Ok so data is written differently depending on the playback media which you said in so many words. This data (DTS), has various flavors as improvements in formats and encoding have come into play.

DVD-A discs were hot for a NY minute, SACDs came along, BluRay, and DTS CDs before them. I never expected the encoding nor the data to be written the same way, decode the same, play with the same software, rip to the same file types, nothing. Nothing being the same or a set permanent standard. It’s all been a wild ride or a chase to get them all playing. I’ve succeeded in most of them and found happiness in playback after some years. My music server has a maybe dozen file types.

I think if you are searching for a neat and tidy world of aligning standards and formats (even from the same companies) you are in for a very rough ride and hard lessens.
 
Last edited:
Back
Top