I would like to distribute the live output (video and audio) of 2 satellite receivers to many TVs in my house.
Most of my TVs are equipped to be DLNA clients. Those that aren't will be equipped with Raspberry Pi XBMC clients, which support being DLNA clients.
I'd like to allow each TV access to either satellite receiver by connecting each satellite receiver to a Hauppauge HD-PVR device and then streaming the output to them.
I'd like to use miniDLNA to provide the stream, possibly with some transcoding via ffmpeg.
The "server" that the HD-PVRs will connect to will run Fedora 19. The HD-PVRs mount as character device files in F19, /dev/video1 and /dev/video2.
I'm proposing to add code in miniDLNA to allow it to accept a char device file as a streamable file instead of a regular file.
Comments, suggestions ?
I have a similar requirement...
I would like to stream live video from my RPI camera to my Sony TV (a dlna client). Minidlna works fine sending video to the TV from disk storage.
I have looked at using vlc, as a dnla server, but although this was mentioned in early docs as being a feature it has not been implemented.
Can using readymedia work? or is there a better application I haven't found yet?
Does MythTV do what you want? At a minimum, you can grab their code for
"Does MythTV do what you want? At a minimum, you can grab their code for
Kinda sorta, but not really.
MythTV does "live" TV viewing via a variety of hardware devices, HD-PVR being one of them. However, I find their implementation unnecessarily complicated (for what I need). Admittedly, MythTV does lots of neat things that my implementation won't do. On the other hand its a bear to set up and isn't optimal for live TV viewing.
Furthermore the MythTV backend does not serve the live TV functionality to non Myth type front ends. (ie it does integrate with XBMC front ends, but not to DLNA clients)
I've just looking for something simple that can be used with any DLNA client.
Reading input from a character device isn't overly complicated. I'm just waiting to hear if anyone has a better idea or if there are any gotchas out there that I should be aware of before I start writing code.
I've been looking at code and playing around with 1.1.1.
What is the easiest way to feed minidlna a device file ? Should I create a process that reads the device file (stream) and put it into an intermediate file (buffer) and have minidlna read that file more or less normally ? Or should I change the minidlna read process itself to work with the device file directly ?
Any hints or advice will be greatly appreciated.