At our office, we have 8 x TV's each tuned to a different TV station. Due to the office environment, we cannot have the TV speakers turned on.
I want to be able to stream audio from the 8 separate TVs over local wifi. The idea is the end user can open their browser and listen to any of the eight different channels. For example;
192.168.1.100:8081 (TV1 Audio),
192.168.1.100:8082 (TV2 Audio),
192.168.1.100:8083 (TV3 Audio),
and so on...
My current thoughts on a solution is connecting an audio cable to each TV and plug them into a single 8-channel audio hardware device such as this >
M-Audio Fast Track Ultra High-Speed 8x8 USB 2.0 Interface
https://www.amazon.com/M-Audio-Fast-Track-Ultra-High-Speed/dp/B000Z8U0IY
Is it possible to connect this hardware via USB to something like a Raspberry Pi (RPi) and run an open-source audio server that can serve the eight separate audio streams over the local wifi network?
A few points;
1. Wifi is mandatory as most of the users only connect to the network this way.
2. Some of the channels are delivered by cable so DVB TV tuner cards are not an option.
3. The Linux box needs to be something compact hence the RPi reference.
4. Audio quality can be lo-fi - thinking 128kpbs mono - should this keep resources / stream more reliable.
5. This is not for illegal web streams. This will be used in a closed network environment only.
A few questions;
1. Is there a piece of Linux software with a command-line interface that can read the USB data from the hardware soundcard and stream the different audio channels?
2. If the software is available, will the RPi be powerful enough to serve approx 100 concurrent users over wifi?
3. Am I simplifying the setup and missing something else that needs to get this working?
Any help or feedback is greatly appreciated. Thank you
I want to be able to stream audio from the 8 separate TVs over local wifi. The idea is the end user can open their browser and listen to any of the eight different channels. For example;
192.168.1.100:8081 (TV1 Audio),
192.168.1.100:8082 (TV2 Audio),
192.168.1.100:8083 (TV3 Audio),
and so on...
My current thoughts on a solution is connecting an audio cable to each TV and plug them into a single 8-channel audio hardware device such as this >
M-Audio Fast Track Ultra High-Speed 8x8 USB 2.0 Interface
https://www.amazon.com/M-Audio-Fast-Track-Ultra-High-Speed/dp/B000Z8U0IY
Is it possible to connect this hardware via USB to something like a Raspberry Pi (RPi) and run an open-source audio server that can serve the eight separate audio streams over the local wifi network?
A few points;
1. Wifi is mandatory as most of the users only connect to the network this way.
2. Some of the channels are delivered by cable so DVB TV tuner cards are not an option.
3. The Linux box needs to be something compact hence the RPi reference.
4. Audio quality can be lo-fi - thinking 128kpbs mono - should this keep resources / stream more reliable.
5. This is not for illegal web streams. This will be used in a closed network environment only.
A few questions;
1. Is there a piece of Linux software with a command-line interface that can read the USB data from the hardware soundcard and stream the different audio channels?
2. If the software is available, will the RPi be powerful enough to serve approx 100 concurrent users over wifi?
3. Am I simplifying the setup and missing something else that needs to get this working?
Any help or feedback is greatly appreciated. Thank you
Last edited:
You can do that with gstreamer.
Phofman is correct in that there will likely be 100msec of latency or more.
Gstreamer is not well documented. I know because I wrote a streaming application using the command like version of it and it was like pulling teeth. In the end it works great and gstreamer is very nice but the process of trying to figure out what to do in order to get it working was painful. If you decide to go down that road, drop me a line and I can save you some hair pulling.
To get the audio in you could think of using a pro audio interface. There are some for a couple hundred dollars with 8 (mono) inputs. If you keep the sample rate down at 44.1k you can easily read in this many channels to a Raspberry Pi and then send them out at a low bitrate AAC or MP3 stream that is good enough for a noisy bar/restaurant. Gstreamer can do all the conversions and stream multiple streams simultaneously. I know because my own system can take one input and stream out to many different clients. I have never tried 8 inputs however or streamed that many channels simultaneously but I think it is doable.
Finally, as Phofman mentioned, there are modulated RF "analog and video sender" things that you can buy that will do all of this for you with essentially zero latency. If you make sure to get one with as many channels as possible you will prevent leakage from one wireless channel to the next corrupting the audio. You probably need to space out the frequencies used in order to do this, because the filters are not sharp. This type of stuff is not expensive and there are lots of offerings for remote controlled aircraft like drones where people stream back the audio and video signal from a mic/camera mounted to the drone. They are on the order of $20 a pop I think. I am somewhat familiar with these things, because the video channel has just enough bandwidth to send 44.1k SPDIF and I tried that before coming up with my currect gstreamer/wifi solution. The downside of these is the presence of interference thanks to cheap design and implementation of the TX/RX circuits. Not good for hi-fi but for your needs probably perfect.
Phofman is correct in that there will likely be 100msec of latency or more.
Gstreamer is not well documented. I know because I wrote a streaming application using the command like version of it and it was like pulling teeth. In the end it works great and gstreamer is very nice but the process of trying to figure out what to do in order to get it working was painful. If you decide to go down that road, drop me a line and I can save you some hair pulling.
To get the audio in you could think of using a pro audio interface. There are some for a couple hundred dollars with 8 (mono) inputs. If you keep the sample rate down at 44.1k you can easily read in this many channels to a Raspberry Pi and then send them out at a low bitrate AAC or MP3 stream that is good enough for a noisy bar/restaurant. Gstreamer can do all the conversions and stream multiple streams simultaneously. I know because my own system can take one input and stream out to many different clients. I have never tried 8 inputs however or streamed that many channels simultaneously but I think it is doable.
Finally, as Phofman mentioned, there are modulated RF "analog and video sender" things that you can buy that will do all of this for you with essentially zero latency. If you make sure to get one with as many channels as possible you will prevent leakage from one wireless channel to the next corrupting the audio. You probably need to space out the frequencies used in order to do this, because the filters are not sharp. This type of stuff is not expensive and there are lots of offerings for remote controlled aircraft like drones where people stream back the audio and video signal from a mic/camera mounted to the drone. They are on the order of $20 a pop I think. I am somewhat familiar with these things, because the video channel has just enough bandwidth to send 44.1k SPDIF and I tried that before coming up with my currect gstreamer/wifi solution. The downside of these is the presence of interference thanks to cheap design and implementation of the TX/RX circuits. Not good for hi-fi but for your needs probably perfect.
We are looking to only stream mid to lo-fi audio (not video). As far as latency goes, what would one expect in such a scenario? Are we talking m/secs or a few whole seconds?
Afraid not. Something like an FM transmission is a no go.
This is about latency. As Charlie says, you will hardly get under 100ms audio delay behind video which is way too much for reasonable use.
Why? It is a standard solution for this requirement. E.g. every theatre uses multiple wireless microphones. Just let users tune their receiver to the channel carrying their desired TV sound.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.