[WINDOWS][PATCH]Bitstream output of HD audio formats

Awesome work, but will this be integrated into xbmc eden final?

I'm getting a "Audio Device Failed to Initialize" error when trying to play DTS-HD(MA), even though both my devices are supporting it (Win 7 Pro x64, Nvidia GT520, Onkyo TX-NR 905).

My XBMC settings are as described in post 1, using the newest driver for the graphics card.

Any hot tips why that might be?


Awesome work btw!


Edit: Just got it working - using the newest nvidia beta driver did the trick!

DanielaE Wrote:It's more a matter of correctly modeling reality rather than fallback, to be precise. The background is the HDMI transport protocol which carries either exactly one IEC 60958 audio stream in its multiplex, or four of them. This relates to a 2-channel or a 8-channel sample structure at the interface level. The true number of output channels in a given audio stream is determined by out-of-band 'valid' flags in case of LPCM or in-band signaling in case of the various encoded formats according to IEC 61937.

DanielaE - looks like you nailed it this time - thanks great work - all the TrueHD titles that exhibited stuttering and audio sync, are now rock solid and playback perfectly

Excellent work and thanks also to DDDamian for compiling - and including the Audio Delay patch.

Thanks for all the work put into this!

XBMC feels complete to me now.

No more external players!

I opened up a mkv container I had earlier with non-hd audio and merged it into a new mkv with a dtsma track. This time I can go in to the audio screen in xbmc during playback and see that it's hd audio that's being played, but on my reciever it still just says DTS.

If I change "receiver gets other DTS formats" to "decoded (LPCM)" my receiver display show PCM, so I'm assuming something's wrong and it doesn't get the hd audio format from xbmc for some reason, can anybody help me out? Thanks in advance!

edit: nvm, solved it. Even though my receiver supports it, changing "receiver supports" from DTS-HD High Resolution Audio to DTS-HD Master Audio made it work. Thanks a lot for this patch!

reta Wrote:edit: I'm assuming I don't need the w7 codec pack and the shark007 thingie that bluray mentioned somewhere, since that's for external players and this patch makes hd audio playback possible in xbmc itself, right?

W7 codec pack and Shark007 are mainly for W7MCE. For XBMC, you can simply update your GPU latest driver from here- AMD driver or Nvidia driver. Make sure that you select the proper HDMI audio output and configure speaker to "stereo" in W7 control panel, and then select the same audio output in XBMC. If your GPU and AVR supports bitstreaming, the latest "XBMC Beta 3 + HD-Audio + Global AV Delay" should bitstreaming DTS-HD/TrueHD. If you don't feel comfortable with MakeMKV file, you can try ripping it with AnyDVD HD to ISO or BD folder or copy/paste M2TS file to your local storage. You'll have 14 days trials from Slysoft with no string attach.

All you have left to do is enjoy the show!

T800 Wrote:No more external players!

I still need it for most blu-rays to enjoy the original blu-ray menus, previews, chapters, etc and 3d once in a while. For blu-rays that I want to playback with External Player, I simply named it "title_ext.iso" and use this protocol "<rule filename=".*_ext.iso*" protocols="*_ext.iso*" player="ext_Player"/>". The rest of my blu-rays still can playback using the latest XBMC patches.

The latest XBMC patches and external player "TMT5" are excellent combination for XBMC!

I have a TrueHD mkv file made with mkvmerge which had a slight sync issue (few hundred ms), but can confirm the latest build seems to have resolved it.

Thanks very much and great work!

DanielaE Wrote:It's more a matter of correctly modeling reality rather than fallback, to be precise. The background is the HDMI transport protocol which carries either exactly one IEC 60958 audio stream in its multiplex, or four of them. This relates to a 2-channel or a 8-channel sample structure at the interface level. The true number of output channels in a given audio stream is determined by out-of-band 'valid' flags in case of LPCM or in-band signaling in case of the various encoded formats according to IEC 61937.

@Daniela - quite correct - was forgetting that's in the muxer code, not the actual output being forced. Sounds like more folks are ringing back with good results on TrueHD sync. A fantastic piece of work - and it will have a postive impact on AE as these timing issues are not resolved there. Very appreciative of your work!

RavenNL Wrote:Awesome work, but will this be integrated into xbmc eden final?

No.