Ever thought of working with ffmpeg?

Transcoding audio track into mp4(aac) is just about 30kb of hardcore C code.

That’s a small highlight on our progress on the streaming service, we are building a system which accepts user content and allows to create long-lived broadcasting translations containing many tracks in one stream.

For example your stream may start with file ‘introduction’ then ‘adv1’ then part of the file ‘main content’, ‘adv2’, ‘main content’, ‘final part’ and so on.
The only thing you need is to upload your audio/video tracks to our service, and create your stream using our interface. If you prefer, you can setup different audio track for your stream.
We will use adaptive HLS/DASH streaming for different client bandwidths.

We will not concatenate your video files together, instead we are using real-time stream composition on our streaming servers which build video content just out of the files you uploaded.

Here is initial presentation (MPEG-DASH) which muxes 2 video streams (5 seconds each, sample-after-sample) and 2 completely different audio streams: http://video.reverbrain.com/index.html