Live streaming with addition of overlay images, text, etc.

[munchoa]

Hi guys,

I would like to ask you for some help.

I am trying to combine videos from several sources (webcams) then add some overlay images and text and stream the result live. The main problem is that text and images can be changed at any time and are not known in advance. So basically I must be able to add unknown staff to the video at any time (something like live sport events when they display results).

I was considering using AviSynth (pretty powerful staff). Problem is as far as I understand it can be used mainly for postproduction and not for live processing. I was hoping to find some plug-in for AviSynth (before trying to code one myself) that will allow me to add images and text "on-the-fly" but no luck so far. If anyone knows such a plug-in or function, please let me know. I'll be most grateful to you.

Can you tell me what other software/hardware I can you to achieve that? I prefer to be able to automate the process (i.e. the software/hardware does not require UI interaction) but will accept anything! I can code if needed (e.g. write a plug-in for existing software).

Any help appreciated!
Thanks in advance!

[NerdWithNoLife]

Sometimes it's possible to use ffdshow in between AviSynth and whatever program you're using to stream the video. You would adjust ffdshow's video decoder settings. There's a part where you can use AviSynth. Then you'd do something like:

Layer(last,ImageSource("overlayimage.png",pixel_ty pe="rgb32"))

But even if you could get it working (probably after much aggravation) it would be tough on the fly.

It's pretty hard to find a good free program that will let you use a custom image for overlays. You'll probably need payware here.

[munchoa]

Thanks NerdWithNoLife,

I came to the realization that what I actually need is a Video Mixer SDK that I can use to program what I need. I'm currently investigating two such SDKs for Windows:

http://www.medialooks.com/products/multimedia_components/video_mixer_object.html
and
http://www.datastead.com/products/tvideograbber/overview.html

Both offers capturing from various sources, adding images, text, effects, etc. and then saving the video to a file or streaming it over the network. Not free, but will allow me to do what I want.

I should also say that using a SDK suits me better, because it allows me to obtain images and texts from any source such as XML file, web service, user input, etc.

Thanks again for your help.
I'll be back with more news about my progress for anyone who is interested.