First off, this is not working properly. I need someone with a good test setup to build and run this and see if it will work. I am at a point in testing where I need to really dig deeper into it and start simplifying the code get it working. The issue could be with my test setup and I would like to eliminate that before I go any further with it and potentially waste time chasing a setup bug on my end. Thanks. Nevermind. I think I found the issue.
Hi, this is awesome to see and I'm happy to test it out whenever it's ready. My CDVR is an M1 Mac Mini and as I'm sure you know, containers can't use QSV for hardware-accelerated transcoding on these machines. They can use VideoToolbox acceleration natively, but only outside Docker. Because of this right now, Multichannel view works very well for me, and quite seamlessly on this Mac, but attempting the same posed too much of a struggle when using the original Multi4channels. With this new implementation, will I be able to use QSV? Asking because I see a dockerfile as well as a Flask app.
Thank you so much for your great work on this and all that you're sharing so graciously with the community here.
Yes, it can be ran outside Docker. Docker just makes testing simpler for me. It's a snap to build and run when I make changes and switching between test servers. I am having to change the way I am attempting to add dynamic audio track switching as my first try hit a huge wall.
I have tried and tried. I can not figure it out. Sorry but this is hard stuff.
1 Like
All good, thanks for trying. I really appreciate everything you’ve learned and shared thus far. Personally speaking I don’t mind using Channels’ native audio selector to choose the now playing track. Some tracks have two which confuses up the numbering but I’m able to figure out which is which, quickly enough. Hopefully sometime soon this gets easier to control but for now just the introduction of multiview capabilities inside the Channels client app, feels awesome.
I’d love to see some sort of hybrid between multichannels view and multi4channels that allows us to determine which four channels fill the grid on-the-fly. I have some “news grids” and some “sports grids” already set up as channels in the guide, but some days warrant a different recipe of grid channels to keep eyes on, and setting them up in multichannels view is a bit too cumbersome for casual use. Multi4channels has the UI I’m thinking about but it doesn’t perform well on my M1 server because as mentioned above docker can’t take advantage of hardware acceleration.
I’m now wondering if something like ChatGPT is capable of combining these functionalities?
I came up with a simplified plan of changing audio tracks through a middle script. I'm going to give it a try this week and see if I can't get it working. This plan does not involve complicated filtering or ffmpeg. It should also be able to work on any m3u stream.
1 Like
Good news and bad news. The good news is the script actually did dynamically change which audio track was active in the stream being sent to the clients. The bad news is the client player does not care if the audio track changes in the stream before it reaches the client. It will only change audio tracks from within the client player, not from outside the client player without restarting the stream. This is true for VLC and Channels clients. Long story short, it won't work like I had hoped it would and I can not think of a way to do it without an API endpoint within the client that would allow for changing audio tracks.
1 Like
Excellent effort, with some very useful information coming out of it. Hopefully the devs will add an endpoint for this to the clients.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.