Need some direction on Capture Cards/Software SDK's

[acrozier]

Hello,

I am looking for some advice on how everyone else would go about meeting a specific need. The situation is described below:


I need a solution that can capture up to 5 streams at the same time on a single machine (multiple cards is ok). These streams will come in as Composite to the card.


On the software side I need to write an app that can capture those streams and encode it to MP4, preferably directly so as to save system resources to store/convert. As far as the quality of the video DVD Quality would be preferable, but I realize this may not be achievable given the amount of streams I am looking to capture at any one time.


We have an old application that runs of Windows Media Encoder 9 for capturing (to WMV) and uses Osprey 440X cards to get the analog signal in.


A few years ago I wrote desktop app leverage Expressions Encoder 4 SP2 and the new PCI-E models of the Osprey cards, but this is where I ran into troubles. Our target output is MP4, but in order to get to MP4 in Expressions Encoder you had to go to IISMV first and then to MP4, which means you not only had the overhead of capturing multiple streams, but then converting them out of IISMV. I also now see that it seems they are dropping support for Expressions Encoder and going all in with Azure which it seems wouldn't provide any capture abilities at all (just stream, convert, store, etc).


What does everyone else do? I suppose Linux & FFMPEG? If anyone has any windows based ideas please let me know, maybe leveraging the Osprey SDK, or another capture card all together.

[usually_quiet]

mp4 is a container file format, so strictly speaking, you can't "encode to mp4", but you can package your already encoded video in an mp4 container.

The mp4 container can hold different kinds of encoded video, but H.264 is the most common. Is that what you want? If so, you probably want a computer capture device with multiple inputs for composite video, plus on-board H.264 hardware encoding to reduce the demands on the CPU. The only type of product I have seen that has those features is used for video surveillance/security. However, like most of this website's members I'm more familiar with capture products that only capture from a single composite source, so I don't know enough about these other products to say whether any will be suitable for your project in every respect.

[acrozier]

My mistake - yes h.264 is what we would be targeting. I suppose that the Capture card to use is less important to me as what software is good to capture it.

The Osprey cards work just find to get the signal in, but outside of Windows Media Encoder 9 which is obviously well past unsupported I can't seem to find a reliable framework to capture multiple streams.


Was just wondering how people normally handle these types of scenarios, I know that Osprey offers an SDK, but I don't know if this would be something that could achieve the functionality needed, or if the SDK just provides methods to grab the video/audio streams, still leaving us with the problem of finding software to capture it.