I'm trying to stream a long video (6 hours) to Youtube using amazon services 24/7.
I followed this guide https://aws.amazon.com/ru/blogs/publicsector/live-streaming-facebook-youtube-aws-elemental-medialive/ and it seems that everything worked. But there is one problem. When the broadcast ends, viewers in YouTube see a black screen. What can I do to start the video from the beginning, and make all this works in a cycle in 24/7 mode.
When i'm trying to configure my Amazon LiveMedia Channel to make my input video looped, i don't see the desired option.
What can i do to make my LiveStream looped?
I found: Edit the channel >> Click on the required Input under 'Input attachments' >> Under 'General input settings' please update the value of 'Source End Behavior' to 'LOOP'
Related
I’m obtaining H.264 video from a DJI drone in an Android library I wrote. From there the video is distributed via WebRTC to many subscribers. This works.
Now one came to the idea if it would be possible to have an RTMP stream aside, so that a parallel publishing of the video to platforms like YT or FB would be possible.
I integrated the code which does H.264 to FLV to RTMP and it works perfectly with at least two open source solutions I have tested: OSS/SRS (https://github.com/ossrs/srs) and node-media-server (https://github.com/illuspas/Node-Media-Server). I publish to instances running here in my LAN and view that by VLC. That works fine.
It doesn’t work if I publish to YT directly. Then I thought I try to insert restream.io into the chain. But it also does not work reliably. Restream at least is a bit more chatty regarding what’s happening, but not chatty enough: What I see is, that I can connect and disconnect - the dashboard window reacts promptly. Same as YT does. I see that the RS dashboard shows bitrate, frame rate and key frame rate and the statistics confirms that. Just - the screen remains black (as with YT) and there is just this spinning wheel.
I can exclude, that I have any kind of weird firewall problem, since I can perfectly uploading H264 as FLV stream using FFMPEG from the command line.
So what is the state: I have two open source RTMP servers, which tell me, all is fine. I have two major public RTMP servers, which don’t say much, but don’t confirm that it works either…
I'm looking for some hints to find out, what is wrong with my stream :)
The simple reason was: YT REQUIRES audio. My stream didn't contain any audio. So I multiplexed a silent fake audio stream into the upload and it worked.
I am streaming video from an IP camera to Youtube using the restreamer software (https://datarhei.github.io/restreamer/) running on a Raspberry Pi 4.
Actually everything works as well. But when the stream is interrupted (for whatever reason), a new livestream is created only when I click on the "Dismiss" button in Youtube Studio.
But since I want to start the new stream automatically, I'm looking for a way to execute the click on the button (or the associated command) via the Youtube API.
How can I execute this?
I will answer my own question because I found the problem description and a solution in another forum: https://obsproject.com/forum/threads/stream-didnt-show-error-or-load-to-youtube.130897/
Something has changed on youtube in the last few days, but the solution to manage a real streaming reconnection is this: from the youtube control room from the manage icon, you need to create a scheduled streaming. Start the encode on this stream and everything will work fine. If the connection drops, youtube will say streaming offline; when the connection is re-established, streaming restarts.
I followed this link with the second answer (two steps) to implement the video upload to Youtube: Upload video on Youtube using curl and api v3
The code works fine, but when I executed, I got response saying:
status:
embeddable: true
license: "youtube"
privacyStatus: "unlisted"
publicStatsViewable: true
uploadStatus: "uploaded"
That means it is working, but when I went to my youtube account to check the video on videos list, it said:
Processing abandoned
The video could not be processed
The video is not playable with a red triangle, so are there any more info about why? and how to fix it? the video is about 30 seconds long, and when I manually upload from my youtube account UI, it works.
also my scope is configured with:
scopes: ['https://www.googleapis.com/auth/youtubepartner-channel-audit',
'https://www.googleapis.com/auth/youtube.upload']
Export your video again and save it in a different format supported by Youtube. at first, youtube kept abandoning the processing of my mp4 video but when I saved it as MOV, I was able to upload it with ease. Hope it helps because uploading with video classic didn't fix my issue.
This video will explain to you all the possible reasons for your video rejection by youtube.
watch till the end so you can understand what is wrong with your video.
https://youtu.be/-iVtTHhfYtA
This is a new video with new reasons behind youtube rejection
I had the same problem with Electron. Solved by moving the API call from rendering process to main process. Apparently the video didn't upload at all using the node.js API in browser context.
Every week I have to broadcast live Private event on YouTube. All settings always is identical.
Is any way to automate process of creating live event on YouTube web page with script or any app?
Ideally would be to start live stream without go to YouTube page every time.
Wirecast and OBS does not have option to create event inside the app. Only XSplit does but is on Windows.
There is lots of settings on multiple pages I have to go through every time before you press "Start Stream" button. Just trying to simplify that process in one click. I am on mac.
Any ideas?
I'd like to upload an audio file to my website, programmatically create an video slideshow using an image already on the server, and then upload that video to YouTube using its API. The website is built on Grails, so a Java solution would be great; Linux command line could also work.
Uploading videos using the YouTube API is working well. I also have no trouble getting avconv or ffmpeg to create videos that will play in, for example, Windows Media Player. This works to make a playable movie:
avconv -i image.jpg -i audio.mp3 video.mpg
However, manually uploading that video to YouTube results in a format error. I can create other files that YouTube will process successfully, but the audio plays back with regular stutters. (A second or so of clear audio, then a brief gap, then some clear audio, ...) For example:
avconv -i image.jpg -i audio.mp3 video.mkv
I have experimented with a variety of output formats and codecs, but all fall into one of those categories: apparent encoding success but YouTube processing failure, or processing success but unacceptable audio playback.
I wonder if anyone can point me in the right direction. I'm open to any solution that will let me solve the basic problem of uploading audio to YouTube via my Grails website without forcing my users to create the video slideshow themselves.
Thanks in advance for any clues.
It is quite easy to create a YouTube slideshow video on iPad. This guide shows the steps.
Adding music, applying transitions, editing photos can all be done there.
I built a pretty similar web service, based on ffmpeg: it's called tovid.io — uploads user-provided mp3 to YouTube as a video, using uploaded image file for a visual track. User uploads mp3 file, an image, then we render a video in the backend, and publish it on YouTube under the user's account.
It is very easy to render an h264 mp4 video following YouTube specs using Linux command line ffmpeg. Let me know if you're interested in our particular command line.