SamyGO

samygo1 wrote:I noticed a couple of people mentioned it and it seems to have been missed, but is there any possibility of a hack to enable DTS passthrough via optical output? My receiver supports DTS and so it would save on CPU usage etc

Yes I was hoping the project was already going in this direction and that implementing actual DTS decoding was only a step on the way.
For most of us it would make extremely much more sense to just skip the decoding and output the DTS signal to our dts-capable surround systems.
Now i reallize that a good number of people do not own a DTS-capable surround system and un the end I guess there would have to be one version for those who do not own a DTS-capable sound system and one version for those who do. The CPU-hogging decoding/downmixing operation would only be needed for those who do not own a DTS-capable sound system.

Right now there is only silence in the DTS capable system (no signal from the optical output) if you try to play movies without the hack (TV just goes "Audio encoding not supported")
If you load the hack I will have sound through the optical output (but it will be downmixed stereo not DTS).

By the way when the stereo is downmixed with the current patch it seems that whatever sound is coded to go through the center speaker (typically the dialogue of the movie) is much weaker-sounding than the environmental sounds (typically mainly heard through left+right front speakers or sometimes by the back surround channels) - so the sound effects are really loud while the dialog sometimes is hard to hear.

Thanks for your wonderful efforts so far! What you already have accomplished is nothing but amazing!
Now please give us passthrough DTS to optical

yes, i noticed the same delay but i don`t really mind it,
i think it has something to do with the increased cpu usage OR audio codec determination by TV.

karamouzas wrote:A few files with "header compression" (or "header stripping") encoding, are not playable.
(checked with MediaInfo)

This is due to the limited implementation of the muxer routines by Samsung staff. They didn't read the spec properly and/or were unable to implement that feature. It's nothing magical. The header compression just stores the repeatedly reccurring bytes (like "0b 77" for AC3 or "7F FE 80 01" for DTS) within the codec header. The demuxer just needs to prepend those bytes befor sending the data chunk to the codecs. Unfortunately this is not done.

At the moment I am analyzing the demuxing stuff. My intention is to replace the bugged Samsung code by ffmpeg. It should be possible, but will take some days...

michalhejl wrote:Have someone noticed an increased time to switch channel while SamygoDCA is loaded?

Actually I did not have noticed that. There is no reason since all hooks do apply for media play only. I'll have a look on that.

Some issues while using v0.3:
1. Tv freezes (LE32?652 non cip) when tryin to load mkv files with those characteristics.
But if do not use DCA, tv plays files corectly
2. Playing mkv file with three audio tracks (mp3, AC3 ? DTS), only the first (mp3) plays audio. No audio on next two tracks.
3. When start PVR 0.10.2 and SamyGO DCA v0.3 with gdbtrick, problem with energy-saver - module does not work at all. No reaction to keypress
4. If DCA v0.3 and NFS client are loaded by gdb on TV start, and there is no NFS on network available, TV freezes and reboots.