2017-02-19, 22:34
silence
(2017-02-06, 13:00)mambro Wrote: Kody Krypton is out!
I hope someone continue the work on the dual audio patches
(2017-02-25, 08:43)xhbl Wrote: It's been long time I haven't updated this post. Recently I got some time to update to Jarvis but not Krypton since Krypton has rewritten a new video player, which also means new dual audio support code needs to be developed for it. This might need quite some time because first I need to dig into this new video player code. Let's see if I'm lucky to dissect it.Great!
(2017-02-25, 08:43)xhbl Wrote: It's been long time I haven't updated this post. Recently I got some time to update to Jarvis but not Krypton since Krypton has rewritten a new video player, which also means new dual audio support code needs to be developed for it. This might need quite some time because first I need to dig into this new video player code. Let's see if I'm lucky to dissect it.
(2017-02-26, 13:40)Mettbrot Wrote: If you invest time in this maybe you want to work on a permanent solution which can be included into mainline? I bet you can get an overview of what it takes from one of the knowledgable devs in the area, I bet FernetMenta has a take on how the architecture should look like ideally
The way its done on most set top boxes/SAT or cable boxes is without any configuration: Select the Output with the most channels in audio settings and every detected output on the system with less channels automatically gets a downmixed (stereo) signal.
(2017-02-25, 08:43)xhbl Wrote: It's been long time I haven't updated this post. Recently I got some time to update to Jarvis but not Krypton since Krypton has rewritten a new video player, which also means new dual audio support code needs to be developed for it. This might need quite some time because first I need to dig into this new video player code. Let's see if I'm lucky to dissect it.
(2017-02-26, 20:27)methcat Wrote:(2017-02-26, 13:40)Mettbrot Wrote: If you invest time in this maybe you want to work on a permanent solution which can be included into mainline? I bet you can get an overview of what it takes from one of the knowledgable devs in the area, I bet FernetMenta has a take on how the architecture should look like ideally
The way its done on most set top boxes/SAT or cable boxes is without any configuration: Select the Output with the most channels in audio settings and every detected output on the system with less channels automatically gets a downmixed (stereo) signal.
it's been a few years since i've read through this whole thread, but i'm pretty sure there were some reasons this was specifically not accepted in the main branch, things may have changed, but they were fundamental usage decisions, and not based on technical ability to do so.
thanks xhbl for your long time work on this, i've been using these builds exclusively since i can remember.
(2017-02-27, 11:57)Mettbrot Wrote: The reasons lie in the current architecture of this patch (duplicating AE), not in the feature itsself: http://forum.kodi.tv/showthread.php?tid=...pid2074153 and the following 10 posts is a good start.
I bet the offer to help find a better solution still stands...
(2017-02-26, 20:42)methcat Wrote: i saw the update on the main page, thanks! is there any notable difference to using this or the libreelec 7.0.2 that darkangel posted last july? that was the first libre build i've tried, but hasn't really seemed any different to me other than the splash screen.
(2017-02-27, 21:09)xhbl Wrote:(2017-02-27, 11:57)Mettbrot Wrote: The reasons lie in the current architecture of this patch (duplicating AE), not in the feature itsself: http://forum.kodi.tv/showthread.php?tid=...pid2074153 and the following 10 posts is a good start.
I bet the offer to help find a better solution still stands...
Technically AE is NOT duplicated in this project. In terms of Object-Oriented paradigm, just two AE instances are created in the factory. AE itself as an object is kept intact, only some global audio settings and player interface code are duplicated. Also, if it is named as an engine, I bet we've heard of a twin-engine airplane.
The reason is that it will be a significant impact to the architecture if taking dual (or multiple) audio into consideration of current AE. AE handles a lot of things like audio codec processing, decoding, pass-through, recording, DSP etc., as well as device (sink) handling. it will be very hard to process additional audio within one AE, not to mention another big issue - synchronization between these audios plus video. In a word, complexity.