DTS-HD MA file size compared to Dolby TrueHD

[Frozen Fractals]

I understand that both tracks are lossless and if encoded using a proper source then both files are identical when decoded and outputted via PCM. However, my question comes to the file size. I did a recent encode of Brave's TrueHD track to DTS-HD MA. The resulting average bitrate for the DTS-HD MA track came out to be 3803 kbps. The original TrueHD track comes out at 8373kbps. File sizes ended up at 2.48GB and 4.21GB (respectively). Why the (massive) discrepancies? I mean, I understand they both use different encoding methods, etc., but shouldn't they come to be in the relative ballpark of each other?

[filler56789]

First of all, your results are very-strange indeed, because DTS Master Audio normally should compress worse than TrueHD (or at least I think so). Accordingly, the maximum bitrate for TrueHD in the Blu-Ray specs is 18.64 Mbps, whereas it's 24.5Mbps for DTS-HD MA.
Anyway, those codecs were chosen and made mandatory not because of technical reasons, but because Dolby and DTS Inc are "big dogs".

[huhn]

the max kbit is meaningless.

dts hd can do 8 channel 192khz and dd truehd only 96 khz at 8 channel so it's totally normal that the maximum from dts hd is higher.
are you sure you have decoded this right?
maybe there was fake 24 bit in the dd truehd track and your encoder removed them?
how big is a flac encode and lpcm?

[Frozen Fractals]

Yes, decoded through eac3to. Fully extracted LPCM (8 channels) come out to be 6.02GB. This is a bit under 2GB more than the TrueHD track and about 3.5GB more than the DTS-HD track. Both tracks play back fine and are 24/48 with 8 channels (standard 7.1 config).

[SeeMoreDigital]

I have quite a few Blu-ray HFPA (High Fidelity Pure Audio) disc's... These disc's offer DTS-HD MA audio, Dolby TrueHD audio and PCM audio by design.

Anyway, long story short. After de-muxing the audio streams from these discs, I can confirm that the DTS-HD MA and Dolby TrueHD audio streams are not much different.

Yesterday I de-muxed the audio streams from The Who's Quadrophenia HFPA disc, which has a duration of 81mins 49.774secs and...

The 6Ch DTS-HD MA (with core) audio stream is 4.69GB (4,802MB)
The 6Ch Dolby TrueHD (with core) audio stream is 4.09GB (4,188MB)

Where things get interesting is when you remove the Dolby Digital 'core' from the Dolby TrueHD stream (something you can't do with a DTS-HD MA stream). Indeed, if I do this to the afore mentioned. The new 'stripped' Dolby TrueHD stream is just 3.73GB (3,820MB).


Cheers

[huhn]

maybe they used terrible bad settings for the dd true HD encode.
i don't know any DTS hd or dd truehd encoder so not easy to judge.
how efficiency is FLAC in comparison?

[SeeMoreDigital]

how efficiency is FLAC in comparison?I fed the 3.73GB (3,820MB) 'stripped' Dolby TrueHD stream into 'UsEac3To v1.0.0' and generated a 6Ch 96KHz/24-bit FLAC encode. Which came out at 4.05GB (4,147MB). About the same as the 6Ch Dolby TrueHD (with core) audio stream.

So encoding to FLAC saves nothing...

[foxyshadis]

Can you post a few seconds of each? Someone else might be able to identify something different in the Dolby track.

[huhn]

So encoding to FLAC saves nothing...

that's why i ask this question is the flac more of the size of truehd or more of the the way smaller DTS HD version. thanks to these information you get an idea what the real problem is.

[Frozen Fractals]

Yeah, I'll post it.

I extracted the TrueHD source from the "Brave" blu-ray using MakeMKV and mkvmerge resulting in an .mka file containing the TrueHD stream. I then ran it through eac3to. It ran ok, but did give some lossless integrity errors (log posted below). I then used the DTS-HD MA suite to create the DTS-HD MA stream and it resulted in the above information I have already stated.


C:\Program Files (x86)\eac3to>eac3to.exe E:\Audio\Brave\Brave.mka E:\Audio\Brave
\Brave_pcm.wavs
MKA, 1 audio track, 1:33:37
1: TrueHD, English, 7.1 channels, 48kHz
Track 1 is used for destination file "Brave_pcm.wavs".
a01 Extracting audio track number 1...
a01 Decoding with libav/ffmpeg...
a01 Writing WAVs...
a01 Creating file "E:\Audio\Brave\Brave_pcm.SL.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.BR.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.SR.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.L.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.BL.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.C.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.LFE.wav"...
a01 Creating file "E:\Audio\Brave\Brave_pcm.R.wav"...
-[truehd @ 00307360] Lossless check failed - expected 00, calculated 15.
[truehd @ 00307360] Lossless check failed - expected 00, calculated e7.
--[truehd @ 00307360] End of stream indicated.
[truehd @ 00307360] Lossless check failed - expected 00, calculated 91.
-[truehd @ 00307360] Lossless check failed - expected c2, calculated 6b.
-------------------------------------------------------------------[truehd @ 003
07360] Lossless check failed - expected 00, calculated 9a.
---[truehd @ 00307360] Lossless check failed - expected 00, calculated 9f.
-----[truehd @ 00307360] End of stream indicated.
[truehd @ 00307360] Lossless check failed - expected 51, calculated 2a.
[truehd @ 00307360] End of stream indicated.
a01 The original audio track has a constant bit depth of 20 bits.
eac3to processing took 4 minutes, 48 seconds.
Done.

C:\Program Files (x86)\eac3to>

EDIT: Here is a sample of the center channel extracted from the TrueHD source using eac3to (see above log). Download (https://drive.google.com/file/d/0BwLArwoHLiZlVzkwZWpQaHhNMUU/edit?usp=sharing)

[SeeMoreDigital]

Can you post a few seconds of each? Someone else might be able to identify something different in the Dolby track.
I've uploaded an (84.2MB) 60 second .m2ts sample, taken from the first track of the Blu-ray (HFPA) disc version of Elton John's - Goodbye Yellow Brick Road.

EDIT: The .m2ts sample includes: AVC video | 2Ch DTS-HD MA audio | 2Ch Dolby TrueHD audio

https://onedrive.live.com/redir?resid=52563BF1B3DCDF3D%21194

Just click on the 'Download' button (top left)...


Cheers

[tebasuna51]

Audios in your sample aren't 6Ch but stereo.

Each sample can be different, here:

File/Size Original Decoded WAV Recoded FLAC
--------- ---------- ----------- ------------
DTS-MA 23.108.944 34.541.636 17.143.502
Thd+Ac3 22.419.148
TrueHD 17.624.268 34.523.588 17.131.687

The TrueHD is 31 ms longer, but identical, to the DTS-MA.
The size of Flac is 50% of WAV, and is a typical value.
The TrueHD compression is good in this sample, not always is so good and many times the Thd+Ac3 is bigger than DTS-MA

[SeeMoreDigital]

Audios in your sample aren't 6Ch but stereo.
Sorry about that. I've corrected my previous post.

Okay, I've uploaded a (228MB) 120 second .m2ts sample, taken from the first track of the Blu-ray (HFPA) disc version of The Who's - Quadraphenia.

The .m2ts sample includes: 6Ch DTS-HD MA audio | 6Ch Dolby TrueHD audio (the 'huge' AVC video stream has been removed to reduce the data size)...

Link: https://onedrive.live.com/?cid=52563BF1B3DCDF3D&id=52563BF1B3DCDF3D%21185


Cheers

[jpsdr]

The issue of THD vs DTSHD, i've noticed, is particularly when multiplexing m2ts for Blu-Ray. THD+AC3 results in a bigger file than DTSHD-MA, so for Blu-Ray, if you want to save/minimalise the final size, DTSHD-MA is a way more efficiant than THD+AC3 ! Even if eventualy the encoded THD+AC3 file is smaller than the DTSHD-MA file.

[tebasuna51]

...
The .m2ts sample includes: 6Ch DTS-HD MA audio | 6Ch Dolby TrueHD audio (the 'huge' AVC video stream has been removed to reduce the data size)...

Whit the new sample:

File/Size Original % Decoded WAV Recoded FLAC
--------- --------------- ----------- --------------
DTS-MA 110.189.472 53% 207.212.612 96.893.098 47%
Thd+Ac3 97.535.422 47%
TrueHD 87.943.102 42% 207.194.468 90.414.362 44%

Here the FLAC is bigger than the THD.
BTW the decoded WAV's (DTS-MA/THD) aren't bit-identical, there are a peak difference of -53 dB (with a delay of 0.021333 ms).

[SquallMX]

The issue of THD vs DTSHD, i've noticed, is particularly when multiplexing m2ts for Blu-Ray. THD+AC3 results in a bigger file than DTSHD-MA, so for Blu-Ray, if you want to save/minimalise the final size, DTSHD-MA is a way more efficiant than THD+AC3 ! Even if eventualy the encoded THD+AC3 file is smaller than the DTSHD-MA file.

Yes, the muxing overhead is higher for THD+AC3 that for DTS-HD MA, sometimes I gain an additional 1 Mbps for the video stream by compressing THD+AC3 to DTS-HD MA (768 Kbps for core DTS), and if I down-sample from 24 to 16 bits the gain its around 2.5 - 3.0 Mbps, great for long movies!!!.

So, if you're encoding for Blu-ray DTS-HD MA is the better choice.

[marcusj0015]

No, TrueHD is the best choice because the files are smaller, and far easier to decode...

[SquallMX]

No, TrueHD is the best choice because the files are smaller, and far easier to decode...

Not for Blu-ray, the muxing overhead is higher so the resulting m2ts file is bigger that the equivalent DTS-HD one, for HTPC storage maybe (MKV has close to zero overhead).

[marcusj0015]

I didn't realize muxing overhead could be affected by codec choice, does TrueHD have smaller packets, needing more of them? is that the cause of the problem?

[SeeMoreDigital]

I didn't realize muxing overhead could be affected by codec choice, does TrueHD have smaller packets, needing more of them? is that the cause of the problem?With respect... If you want to compare a formats file size don't bother with containers, just use the elementary streams...

[marcusj0015]

With respect... If you want to compare a formats file size don't bother with containers, just use the elementary streams...

In general I agree with you, but if there's a gig of overhead or something, you should start counting it. idk where that line should be drawn though.