Questions tagged with AWS Elemental MediaConnect

Within a secure AWS video-on-demand platform, the goal is to authenticate users that can get access to content stored in s3. If there is a different service for authenticating users, how can one integrate that with AWS? Will this be in addition to Cognito or replace Cognito with the other authenticator? Are there specific steps to follow to accomplish this task? This is the architecture diagram for aws service: https://aws.amazon.com/blogs/media/creating-a-secure-video-on-demand-vod-platform-using-aws/lg...

I am using medialive and mediaconnect to restream srt streams to multiple destinations. I realized that the RTMP output is not respecting the bitrate specified and even sometimes is not respecting the resolution. Is there a special setting to enable for it to output with the specified resolution and bitrate? Current setttings: ``` Width : 1920 Height: 1080 Codec: H264 Rate Control: CBR 6 500 000 Frame Rate: Specified 30000 / 1001 Progressive GOP Size : 60 Frames B-Frames 0 Scaling Default AFD Passtrough ``` --> Received stream : 1920x1080 or 1280x720 / 1.3 to 1.6 mbps ## ** EDIT/UPDATE** I tried to create a brand new channel from a workflow, selecting twitch as a destination but replacing the endpoint with my own RTMP destination. The only output option was 1080p@30/4.5mbps output. Didn't change anything except the endpoint. Start the flow. And I'm still receiving 1.5mbps video. I was hoping to avoid rolling my own encoder to forward streams to our clients, but if there's no way to control the output of medialive, I don't see how I can achieve it witihin AWS.lg...

We are looking to deploy this or similar solution for our company we have 4 offices with 700+ cameraslg...

I have been testing livestreaming with MediaLive to YouTube and Facebook and in the course of the stream we occasionally put up still images (e.g. some text for audience to read, or photographs to view) which is causing MediaLive to invoke the Input Loss Behavior - we get black or repeated frame (as configured) and lose audio though we are sending continual audio throughout. On CloudWatch I am able to see InputVideoFrameRate drops that appear to correlate. Changing the channel general settings Global Configuration to Support Low Framerate Inputs = ENABLED seems to have helped greatly. With this the problem is avoided sometimes but other times we still lose audio behind the repeated frame for a few seconds before it resumes. Input path is from OBS Studio via RIST to MediaConnect to MediaLive using SINGLE_PIPELINE rather than STANDARD. Is there any way to avoid MediaLive detecting still images in a stream as input loss?lg...

Hi, I'm struggling creating Media Connect flows using cloudformation where the ingest protocol is not zixi-push. The documentation does state that srt-listener is not supported via cloudformation ( reference https://docs.aws.amazon.com/AWSCloudFormation/latest/UserGuide/aws-resource-mediaconnect-flowsource.html#cfn-mediaconnect-flowsource-maxlatency ) but I'm trying "rtp" and it fails. Additionally, the error message is not very helpful "Error occurred during operation 'AWS::MediaConnect::Flow'." (RequestToken: <redacted>, HandlerErrorCode: GeneralServiceException)" A working template (using Zixi on port 2088) looks like this; ``` `{ "AWSTemplateFormatVersion": "2010-09-09", "Description": "Media Connect Flow Test", "Resources": { "MediaConnectFlowA": { "Type": "AWS::MediaConnect::Flow", "Properties": { "Name": "WIZARDA", "AvailabilityZone": "eu-west-1b", "Source": { "Name": "WIZARDASource", "StreamId": "A", "Description": "Media Connect Flow Test - WIZARDA", "Protocol": "zixi-push", "IngestPort": 2088, "WhitelistCidr": "<redacted>/32" } } } } }` ``` but keeping the protocol as zixi and changing the ingress port results in failure (this could be be design I guess as it's a non-standard Zixi port). Similarly, and more importantly for what I want to do, trying to change the protocol to "rtp" fails e.g. ``` { "AWSTemplateFormatVersion": "2010-09-09", "Description": "Media Connect Flow Test", "Resources": { "MediaConnectFlowA": { "Type": "AWS::MediaConnect::Flow", "Properties": { "Name": "WIZARDA", "AvailabilityZone": "eu-west-1b", "Source": { "Name": "WIZARDASource", "StreamId": "A", "Description": "Media Connect Flow Test - WIZARDA", "Protocol": "rtp", "IngestPort": 2088, "WhitelistCidr": "<redacted>/32" } } } } } ``` Can anyone advise on the right construct to create a flow with RTP source? (also rtp-fec failed) for completeness, the I've run via the console and also using the CLI e.g. ``` aws --profile=aws-course --region=eu-west-1 cloudformation create-stack --stack-name="mediaconnect-rtp" --template-body file://..\MConly.json ```lg...

Hi, I have been using AWS MediaConvert for a while to transcode videos. Since the 25th of May, I notice that the transcoding step takes half an hour when using a sample small video. This in contrast to a couple of seconds of how it used to be before the 25th. The same video is used and there are no changes to the configuration of the MediaConvert job. For example, now I see in the mediaconvert dashboard the submit time of 2022-06-02 13:23:54 and a finish time of 2022-06-02 13:49:58 An example of a long running job id: 1654169034448-uehndv An example of a fast running job id: 1653030543613-wixaa6 What can explain this sudden major increase in transcoding time by AWS Mediaconvert?lg...

Our current understanding is that once a MediaLive channel is started, outputs begin to push immediately. E.g. one RTMP output is YouTube, and once the channel is RUNNING the stream is immediately being pushed to YouTube Is there a workflow that allows a channel to be in a 'preview' state before pushing to the outputs? Where you can view your inputs, before pushing them. Our current workaround is to have a MediaLive 'preview' channel for checking the quality of the feed. And once ready to go live, we start a second MediaLive 'live' channel to send the output to the destinations (and subsequently shutting down the preview channel). This is OK, but obviously, there is a delay in starting up each channel, and we are looking to enable being able to go from 'preview' to 'live' in a matter of seconds vs potentially a matter of minutes.lg...

Hello! We're using MediaLive for livestreaming the output of MediaConnect streams. We have 2 redundant streams coming out of MediaConnect and into standard MediaLive channel. This channel is handling switching if one of the streams is unavailable. We also configured input loss behavior to display when BOTH of the inputs are down. ``` "globalConfiguration": { "supportLowFramerateInputs": "DISABLED", "inputEndAction": "NONE", "outputTimingSource": "INPUT_CLOCK", "inputLossBehavior": { "repeatFrameMsec": 1000, "blackFrameMsec": 10000, "inputLossImageType": "SLATE", "inputLossImageColor": "000000", "inputLossImageSlate": { "uri": "<technical difficulties image.png>" } ``` While testing, we noticed that disabling one of the inputs makes the channel to properly switch to the other one. But when we disable the second one, streaming stops without displaying the slate image (then repeats last 10 seconds of stream). When I was testing the same configuration as above on single pipeline channel, it was working fine. Do you know what might be the issue?lg...

My aws elemental media convert is showing Network failure message failure code networking error time 2022-03-16T14 Retryable true Someone please help melg...

The documentation suggests there is cli support for mediaconnect but it doesn't seem to be there yet. Is there a rollout date?lg...