Without this, you one normally checks this out with vendor’s HTTP/CGI reference and/or using network sniffing tools.

We need to broadcast from a network camera using Microsoft Expression Encoder to use its smooth streaming feature.

The 1st problem is that MEE does not accept urls as live source. At this point your app makes the trick as when configured, it exposes new live sources to MEE and works fine.
The test has been done using an AxisP1346 camera and is is ok (http://user:pwd@ip/axis-cgi/mjpg/video.cgi?resolution=640×480&camera=1&compression=25).

The problem is that the production environment has Panasonic WV-SC385 cameras and has not been possible to configure them in a proper way to make the JPEG video source accept its stream. Neither is this stream accepted by the WMP, so I suppose that non standard mjpeg is broadcasted by these cameras. On the other side, VLC does recognizes this stream.
So several alternatives spring to mind:

- One attempt now would be to make VLC broadcast the Panasonic input stream in such a way that could be valid for your JPEG video source. Which settings should we use for this to be ok for your app?
- Another would be using a different filter that accept the Panasonic stream, but how could we expose it to Microsoft Expression Encoder?

Regards

Thanks for the reply :cool:

Where does the program get the .TTF file to process, lets say when specifying the statement “Face=Times New Roman” in the .ini file; what will the program do to find the Times New Roman font file?

It takes a Windows font through Windows API. So it can be any font registered with Windows, such as residing in \Windows\Fonts or available through name mapping.

What does the statement “Weight=1024″ in the .ini file do?

Refer to MSDN http://msdn.microsoft.com/en-us/library/dd145037%28v=VS.85%29.aspx “lfWeight” constants.

Just FYI, I am using windows XP to activate the DOS window in order to activate the rasterizing program.

It is a Windows application, you can run it directly too.

Anyway, because you create a camera and set it up outside of Skype, it might be a good idea to let Skype use it “as is” without modifications in settings. I will check this out when I have a free minute for this.

2. Skype is actually incapable of using high res video, it will anyway cut its 640×480 from the center of the image. With this behavior I don’t see a good reason of tuning the driver just for Skype.

But this is perhaps not the main reason. This is a generic driver, and with M-JPEG URL only we cannot really change camera’s resolution.

To ensure compatibility with applications (including *awfully* made Adobe FMLE) the driver will resize video in code to change resolutions. Any application can take advantage of this in code, but not through UI.

I don’t see a reason why the driver needs to resize video in software without a reason, it is not a good mode of operation. Changing resolution on camera is specific to camera and model. There is no support for this with generic URL. I could do it for specific model, but not in generic driver.

3. Flash won’t process JPEG as is, it recompresses video. This is where tit loses quality.

Later D-Link cameras were OEM models mauufactured by Vivotek – those were the models easy to deal with (with definitely good M-JPEG). At some point later, D-Link again changed their partner and the cameras again were not as good in terms of M-JPEG.