Successful connection to jetson_nano with camlink and gopro, but audio is not working. I stopped streaming, connected the cam link to another usb port on jetson_nano, and tried streaming again, and it finally works. But when I tried auto boot mode, I couldn't hear the audio. It's video only, and is there a command I can choose my own audio device?
There's commands to select the audio device, but if its not working only in boot mode, it might be an issue with the service file used. I can't say for sure, but I threw together a sample service file for a rasbperry pi as a demonstration. Maybe something like this, modified for you jetson, would work: github.com/steveseguin/raspberry_ninja/blob/a9d61fb5ccee2d71d7991fe7482afd514c5974f1/raspberry_pi/raspininja.service If the above doesn't work, and you need help, join the discord server at discord.vdo.ninja and when I get some time I can boot up my jetson and try to debug the issue with you
@@steve_seguin Thank you for your answer. I'm not familiar with Linux or computer languages, so it's hard to understand your feedback. But I think this will help me find a solution.
You forget the option to share the internet on the mobiledevice and use wifi from the rpi :) It have to be close but its possible. I am looking for a way to let the raspberry do some type of inteligent bounding using more than one hotspot wifi maybe more shared internet by regular phones. Something like a travel streaming router. Would it be possible to get rtmp from mevo or gopro and then streamout to webrtc maybe encoding is not needed but can it passtru to webrtc ? Speedify is not available in Denmark so alternatives would be nice...
the MEVO would support RTMP out, which is reliable over a bonded connection -- so long as you can keep the bandwidth stable and sufficient. WIFI works fine for RTMP, but for webRTC I would not recommend it. WebRTC is very low latency, and packet loss impacts it. It might not be an issue always, but network congestion (like being in a hotel or at a big event) will cause even a nearby WIFI adapter to be poor.
@@steve_seguin omg, I am just waiting for the feature of rtmp and SRT support on the raspberry ninja. I really love this project. I would like to go directly from the raspberry ninja to twitch or UA-cam over rtmp, or connect it to a multi streaming encoder over SRT
Adding RTMP support is easy enough, with the exception of perhaps automatic recovery/retry -- that could be more complex. Meh, if that's all that's desired, I can slip that in.
SRT would be awesome and i can see a lot of use cases for it and although i love NDI it would have to be NDI HX3 (using the advanced SDK $$) unless a wired connection was done. But SRT would be ideal. Also probably not a big deal because it's not very accurate and that could have been on purpose but you might want to use YT studio to blurr out the GPS you have being shown. Also i logve me my Peplinks but how has the MAXC transit (the dual?) been working out for you. i heard they self throttle a lot due to heat resulting in packet loss excessively if used for constant internet usage. any truth to that?
I haven't looked at belabox in a while, but this last I checked this had Lower latency, doesn't need a server, uses python, supports data messaging, has access to vdo.ninja ecosystem, and has raspberry pi optimizations, but it's more prone to packet loss and does not do h265. Yet. Raspberry ninja images do have FFmpeg installed, so you can use SRT or NDI if you want, but I've not yet made any special efforts to support that. If doing IRL, and don't mind added latency and firewall setup , belabox is fine.
Nice
Successful connection to jetson_nano with camlink and gopro, but audio is not working.
I stopped streaming, connected the cam link to another usb port on jetson_nano, and tried streaming again, and it finally works. But when I tried auto boot mode, I couldn't hear the audio. It's video only, and is there a command I can choose my own audio device?
There's commands to select the audio device, but if its not working only in boot mode, it might be an issue with the service file used.
I can't say for sure, but I threw together a sample service file for a rasbperry pi as a demonstration. Maybe something like this, modified for you jetson, would work:
github.com/steveseguin/raspberry_ninja/blob/a9d61fb5ccee2d71d7991fe7482afd514c5974f1/raspberry_pi/raspininja.service
If the above doesn't work, and you need help, join the discord server at discord.vdo.ninja and when I get some time I can boot up my jetson and try to debug the issue with you
@@steve_seguin Thank you for your answer.
I'm not familiar with Linux or computer languages, so it's hard to understand your feedback. But I think this will help me find a solution.
You forget the option to share the internet on the mobiledevice and use wifi from the rpi :) It have to be close but its possible.
I am looking for a way to let the raspberry do some type of inteligent bounding using more than one hotspot wifi maybe more shared internet by regular phones.
Something like a travel streaming router. Would it be possible to get rtmp from mevo or gopro and then streamout to webrtc maybe encoding is not needed but can it passtru to webrtc ?
Speedify is not available in Denmark so alternatives would be nice...
the MEVO would support RTMP out, which is reliable over a bonded connection -- so long as you can keep the bandwidth stable and sufficient.
WIFI works fine for RTMP, but for webRTC I would not recommend it. WebRTC is very low latency, and packet loss impacts it. It might not be an issue always, but network congestion (like being in a hotel or at a big event) will cause even a nearby WIFI adapter to be poor.
I am tempted to add support for SRT/RTMP to the Raspberry_ninja project, but I haven't had anyone request it yet really.
@@steve_seguin omg, I am just waiting for the feature of rtmp and SRT support on the raspberry ninja. I really love this project. I would like to go directly from the raspberry ninja to twitch or UA-cam over rtmp, or connect it to a multi streaming encoder over SRT
Adding RTMP support is easy enough, with the exception of perhaps automatic recovery/retry -- that could be more complex.
Meh, if that's all that's desired, I can slip that in.
@@steve_seguin ok so at the moment, is there automatic retry with the VDO.ninja webrtc ?
SRT would be awesome and i can see a lot of use cases for it and although i love NDI it would have to be NDI HX3 (using the advanced SDK $$) unless a wired connection was done. But SRT would be ideal. Also probably not a big deal because it's not very accurate and that could have been on purpose but you might want to use YT studio to blurr out the GPS you have being shown.
Also i logve me my Peplinks but how has the MAXC transit (the dual?) been working out for you. i heard they self throttle a lot due to heat resulting in packet loss excessively if used for constant internet usage. any truth to that?
Another thing I did was install Tailscale on n the raspberry ninja. Nice to access it over vpn
Thank you; there's definitely a lot of room for improvement with my setup.
This job is great! Is there any way to make the bitrate automatic?
I'm still working on that. Should be possible to do it tho
@@steve_seguin OK thank you very much! I will be attentive to your publications.
Pros and cons comparing this to BelaBox?
I haven't looked at belabox in a while, but this last I checked this had Lower latency, doesn't need a server, uses python, supports data messaging, has access to vdo.ninja ecosystem, and has raspberry pi optimizations, but it's more prone to packet loss and does not do h265. Yet.
Raspberry ninja images do have FFmpeg installed, so you can use SRT or NDI if you want, but I've not yet made any special efforts to support that.
If doing IRL, and don't mind added latency and firewall setup , belabox is fine.
What kind of mic is that you have clipped on?
I'll probably be doing a video about it soon, but it can be found here:
www.movophoto.com/products/wmx-2
@@steve_seguin you the man! Thanks for your help as always.